This is pretty sad to read. Before Covid, the Amazon Go store experience was phenomenal. All the convenience of a 7-Eleven but with the pricing of a normal grocery store. The food options were really good and the BlueApron style meal-kits were amazing. The Alexa integration was also nice for being able to just verbally ask what's the next step on a recipe while you're busy stirring or chopping things.
When it rolled out to Amazon Fresh stores, it was a breadth of fresh air. The painful clunkiness of self-checkout was gone. The slow and pointless exercise of unloading and reloading your cart was gone. You could just bring your reusable shopping bags, throw stuff in, and walk home. By far the most hassle-free shopping experience to be had.
Scan as you shop is a big step backwards and feels like you've got the annoying self-checkout experience looming over you the entire time you're there.
The selection and operating hours both took a hit during covid and never recovered.
Lots of services would be phenomenal when you can offload much of the cost to run them to cheap offshore labor.
Remember how great Uber and Doordash were when much of the cost of operating was offset by underpaid workers and VC funds? Now that they don't have unlimited piles of money and cities/states are making them pay more fair wages, the cost-benefit of those services has diminished. I was ok paying $5 for $20 of food to be delivered, but now it's more like $10 - $15 in fees/tip, plus fees hidden in menu prices making that $20 food cost $27.50.
From my point of view, Uber and DD were always going to head in that direction because by any objective measure, they are worse than the solution that already existed. The introduction of a middleman created new costs for drivers, customers, and restaurants alike, yet it's actually worse at delivering food than the conventional "drivers work for the restaurant" model. Sure, you get one website where you can see all the delivery options in the area, but the cost is making all of those options shittier. I could rant about how stupid that industry is for hours, but I digress.
What Amazon is doing is different because they're actually doing something that adds value, and they're solving a problem that actually exists. Maybe I'm naive for believing any of this matters.
To be honest, I never thought much about how the whole system worked. I've only been in an Amazon Fresh twice. It had the "just walk out" thing but I don't have a Prime account so I never used it. I just figured it was UHF RFID. Seems like an obvious solution, but I'm not the engineer responsible for figuring out what to do if you put two exits next to each other.
It's true that delivery services add a middle-man, but you don't seem to be considering that they add a lot of value in some important ways:
1. The single website/app is not just useful for discovery, but also to provide a consistent ordering experience. An individual restaurant might take orders over the phone, via their website, or via their own app. Those options will vary from one restaurant to the next. Not great for customers.
2. In the old model, restaurants would have to decide how many drivers to have on hand at any given time. During quiet periods they would have to decide between not offering delivery at all, or having a driver sitting around doing nothing. A delivery service that aggregates demand & supply is very useful for restaurants & drivers. Pooling of resources & layering on surge pricing creates a lot of elasticity.
I would argue that what they should have done is tried to at this value at minimal cost to customers, drivers & restaurants. If they were leaner businesses the overall proposition could be a lot better.
You bring up an interesting point. Just thinking out loud for a moment here. Is the current VC business model potentially the problem area and not the actual businesses? The need for rapid growth/software company like profit margins when these companies need more time and potentially are like non-software companies in terms of profits?
> 1. The single website/app is not just useful for discovery, but also to provide a consistent ordering experience. An individual restaurant might take orders over the phone, via their website, or via their own app. Those options will vary from one restaurant to the next. Not great for customers.
Aye. Arguably this is why old-school streaming Netflix won so well -- all of your options in one place. Likewise, there is one app for all food choices nearby.
Also agree on the second point, though presumably larger chains in built up areas could play the same game; e.g. Domino's in a borough of NYC has an "all-drivers-in-Queens" pool.
I'm curious as to how food delivery worked in your country before Uber et al.
In the UK, outside of corporate chains, it was largely a cash in hand operation for both the restaurants and drivers.
Drivers would typically be paid a small sum for showing up on the night and then keep the delivery charge and any tips. The cost to the restaurant was pretty much negligible.
The middle-men, in addition to their own cut, have likely lead to a larger tax burden on many of the drivers and restaurants.
Honestly I think there's a great role something like Uber could fill, simply as a nationalized, not-for-profit tax funded arrangement that businesses could opt into, maybe even a branch of the post office. Ensure delivery people get good benefits and pay for doing a needed job, and the service brings those people, restaurants and stores, and consumers together. We could make things more convenient for people, provide businesses with more opportunities to sell their products, while providing a lot of jobs to people who need them, and also cutting down on the number of vehicles going around at any given moment.
The main drawbacks to these services are the outrageous fees they charge to make money, and that there's more than one so drivers and businesses have to manage an array of phones or tablets. This would address both of those things.
> Under that model, wouldn't drivers only want to work for the busiest restaurants?
It doesn't take many orders to keep an individual driver busy. Typically the restaurants would manage how many, and which, drivers worked each night. If the night was unexpectedly quiet then a driver might finish early.
> With delivery apps, a restaurant can have just a few orders per night and still be open for delivery.
Yeah, the main benefit of delivery apps is enabling restaurants who primarily serve sit-in diners to deliver also. The overall benefit to delivery focused restaurants is much more tenuous.
> I would bet that there's an order of magnitude more deliveries happening (and drivers working) these days vs pre-Uber.
Perhaps, but if a driver was already busy all night under the previous model, then they're potentially harmed by the current model.
> also to provide a consistent ordering experience
Vendored web POS is widely available, and was emerging before UberEats and DD ever did. See: Toast
> A delivery service that aggregates demand & supply is very useful for restaurants & drivers.
This is only true in theory. You aren't giving enough credit to the logistical skills of small business owners, they are really good at allocating and dispatching drivers. If you're small enough to have no deliveries for several hours, then don't schedule any drivers during that time. If a delivery comes in, tell your cashier the address and have them run the delivery. Besides being quite efficient, there's an added bonus that you're cross-training your staff. That's how we always did it, at least.
It's anecdotal evidence, sure, but every delivery driver I have asked (half a dozen from several different cities and states) has said that their job is harder and pays worse today than 10 years ago.
"Well, just be more lean!" What a concept, I'm sure they haven't thought of that. It's much easier said than done. Logistics software is extremely complicated at scale- there's a reason governments subsidize delivery logistics through services like USPS, Royal Mail, etc. However, at the one-business microcosm, it's a simple enough problem that a seasoned owner/manager can organize an optimal dispatch in their head, on the fly.
Thankfully the local pizza place here still has their own drivers and it’s only a two dollar delivery fee. Totally fine with that. And everything always arrives fresh and fast and I know where and who to complain to and I know the drivers are at least somewhat vetted.
I'm with you. After trying a few of those delivery services and finding them poor at best, I've gone back to only ordering delivery from places that do it themselves.
Did you ever try getting groceries/food (other than pizza) delivered before the apps? Uber didn’t replace a better option, they made it literally possible to get these services in many places where it wasn’t an option at all. And in areas like Manhattan, where city cabs and delivery options abound - the the fact that many (most?) people still use UberEats/DD is instructive.
The value of convenience is significant. I happily pay to avoid the mental overhead of shitty restaurant websites or repeating a paragraph-long group order back and forth across a poor kitchen speakerphone call.
Yes the current system has many flaws but let’s not pretend it was utopia before.
> they made it literally possible to get these services
These companies also made it possible for restaurants to be taken hostage and be registered for delivery services or marked as closed/unavailable on those platforms against their wishes, or pay even 60+% commission on an order for the service. Once only a couple of these companies are left standing and have all restaurants in a chokehold I'm convinced the situation will be even "better".
> the the fact that many (most?) people still use UberEats/DD is instructive
Yes, it says that too many people do what's best for them and them alone. It's not as instructive as you think. Many people had slaves because it was good for the owner. What did you learn from that case of "many people do it"?
> I happily pay
You don't and don't even realize that. You get an overall cheaper service but someone pays. Usually the restaurant owner, employees, and delivery people. Maybe some VCs too, for now, while they don't completely own the market. Just like Amazon's same day delivery means some driver has to pee in a bottle for close to minimum wage but otherwise it's absolutely a huge win for everyone who is you, needing the delivery now.
The value of convenience may be significant but is it really great enough to find the cost of providing that convenience without external subsidy over the long run.
To me it is ridiculous that anybody thought local personal services like food delivery, dog walking, and taxi driving could ever be converted into durable, efficient, profitable national or global monopolies. In the absence of monopoly power such low barrier to entry commodity services invariably see their profit margins squeezed to cost from competition. All this is freshman or highschool level economics.
You're right, it sucked. Here's the thing: we can have the shiny, sleek ordering system without the stupid parts of the Uber/GrubHub/whatever business model. You've probably used Toast before.
> in areas like Manhattan, where city cabs and delivery options abound - the the fact that many (most?) people still use UberEats/DD is instructive
I would have leaned towards "destructive" but I get your point.
UberEats and DoorDash are still money-losing operations that subsidize the actual cost of the service, despite rising prices, barely paying their workers, and abusing their relationship with restaurants.
>What Amazon is doing is different because they're actually doing something that adds value
is scanning your groceries manually really so cumbersome to say the "just walk out" model is adding value?
It's nice, but like you said. it's not really that better than existing solutions, and knowing the underworkings makes it much worse. Maybe the tech will catch up in a decade, but it was definitely unviable with the current system.
The scanning isn't the problem. It's the pedestal the size of a postage stamp upon which I'm expected to balance $250 worth of groceries. And if I fail to do so, the machine freaks out. And even if I do succeed in stacking all my groceries up as demanded, the machine still freaks out. Want a beer? Fuck you, wait god-knows-how-long for an attendant to come appease the machine.
I always go to a human cashier whenever the option presents itself, simply so I don't have to deal with the miscarriage of engineering that is the self-checkout machine. Unfortunately, my local grocery stores don't staff their registers very often anymore.
> they are worse than the solution that already existed.
In my country taxis were (and price wise still are) the worst; Uber is a vast improvement; better cars, nicer drivers, convenient booking (until very recently you had to call 1 day in advance to book a taxi) and a bizarrely lower price. The same trip with a normal taxi costing 130 euros is 30 with Uber. So yeah, no. Oh but the taxi companies made their own app! Which is of course a horrible failure.
Ah yeah, my comment was specifically related to the food delivery services provided by Uber.
Granted, the taxi service hardly any better in many countries. In the US, it's rapidly building a new underclass of indentured cheauffeurs, chained down by vehicle maintenance costs and massive fees from service providers.
> What Amazon is doing is different because they're actually doing something that adds value, and they're solving a problem that actually exists.
This shook a bit of illusion from my eyes, thank you. I never thought of Amazon's efforts in this as solving a real problem or adding real value! This is because for me, it doesn't, and I never thought it through beyond that.
Once again, the human tendency to think that our own situation represents the norm fooled me.
>Dominos had this down pat twenty years ago, how come everything with an app is so much more expensive?
The people making the app need a much bigger cut than Domino's. Domino's cut only needs to be big enough to pay their delivery drivers, they make their money selling pizza and the delivery is just an additional service that drives business.
The app maker needs a cut big enough to pay their delivery driver and also be the main revenue stream for their entire business since they're not making money off the pizza.
While I'm not going to claim that was a good idea, the point is that Dominos sees making and delivering pizza as one and the same business.
> 154 examples of the DXP cars were built. Along with the delivery driver, each DXP car could carry up to 45 pizzas, 12 two-liter sodas, and all the extra dipping sauces one could ever want. Add to that the fact that the car’s still held their five-year 100,000-mile warranty, were cheap on gas, and created a surprising amount of buzz, it is hard not to look at the DXP program as a success, low production aside.
I believe that Dominos would include the delivery as part of its core competencies and business value that distinguishes it from other pizza companies while others do not consider delivery as part of their core business.
The margins of the delivery business could be razor thin because the scale at which the app can operate is much bigger than domino's. There are ~7000 domino's in US, and hundreds of thousand on uber eats. I think as more apps offer the same service we'll see a race to the bottom and have a reduction in cost, but not to the VC subsidized levels we saw five years ago.
Picking up one order from a restaurant takes the same amount of time as picking up a dozen. This is a much more important economy-of-scale than having a large app country-wide install base to spread dev costs over.
So I think a third-party delivery business must seek exclusivity contracts to be competitive with an in-house solution.
So what we need is bundling of orders based on
1) restaurant location
2) recipient location
3) time
And if that could be done, maybe just 2 orders to start with (for example carpool benefits for 2 people), that will be awesome, but if you could pool 4-5 orders, that will make bank !
I guess this model, if scaled purely based on the model, will imply restaurants co-located (same parking lot) will benefit from such an arrangement.
At least in the UK this is the default behaviour from Deliveroo, standard delivery does cost money but might get your driver/biker go to multiple other houses on their way to you, or pay extra to get it direct. It made me stop using them, as there's no way to guess whether the normal option will bring the food fresh or not, and paying the premium each time made it feel too expensive.
Also paying the extra premium (whether Deliveroo, Just Eat or Uber Eats) for first delivery is kinda pointless (at least here in the South West) as the drivers seem to be delivering for all the apps at the same time, so your delivery takes 35 mins for the 10 min premium delivery and sometimes they even turn off GPS after they've picked it up if they're delivering for another app first.
So expensive for a frustrating cold food delivery.
I think that's the real difference under all this - Dominos employs drivers (and they get paid by tips, sure there are issues) but it's all one thing. If they deliver a cold pizza it's on them.
All the delivery app companies are NOT making the food so the blame game starts AND they are not employees so a bunch of gaming is going on.
I know my one experience with Uber eats is such that I'll never use it again, and instead travel myself.
Also food delivery drivers employed directly by restaurants typically have an insulated carrier in their car or on their bike to keep your food warm. Uber Eats drivers don't so the food gets cold way faster.
Guess that depends on your location - in the 3 European countries I've lived in, Deliveroo and Uber drivers/riders almost universally had insulated cuboid bags (usually with one of the companies' branding on) in their car or on their bike in all three countries, while Dominos pizza was the same in one country, in another country Dominos used cars with no insulation except the cardboard pizza box, and I don't know about Dominos in the 3rd.
(And I'm not sure if my experiences were representative of each country, just that they were consistent in each city.)
Uber Eats drivers definitely don't have them where I live. Food almost always arrives cold, even when the delivery time is very short. When I lived in NYC most deliveries were done by bike and all of them had insulated bags so the food was typically warm unless it had been super delayed.
That factor is driving the virtual restaurant (ghost kitchen) trend. There are commercial kitchens set up purely to service delivery orders. They use a single facility but have multiple different restaurant brands with different menus (burgers, Mexican, Italian, vegan, etc). These ghost kitchens have no dining rooms and you probably can't even directly order food outside of a delivery service.
These are the worst quality-wise. Yes, you can write a review if something went really bad. But not having customers right there who can complain and might even get an extra or a replacement gives less incentive to keep quality up. "Specialising" in everything, pizza-burger-kebab-china, is also likely a warning sign.
IMO thats the most important point of delivered food. Just compare the process and you will instantly see the difference. In a restaurant, the waiter will openly serve you your food. You get instant feedback. And see right away if something is off. In the delivery case, you get a closed package at your door, say "Thanks" to the deliverer and close the door again. Only when the guy (who isn't responsible for the quality anyway) has left you open your package and are confronted with whatever you got. Maybe something spilled, maybe something misses, maybe something was wrong. And maybe its already cold... Whenever you use a delivery system, you take all these risks, and apart from a grumpy review, you can't do much about these... Personally, I think this is where the margin hides. Delivered food can generally have lower quality without customer complaints reaching the vendor.
I was thinking about it context of complaing to folks running the ghost kitchen vs say at a pizza hut location. I have gotten free food when they screwed up my order at pizza places, and they did care about resolving my problems.
I wonder if it could be solved with a decentralized system. You'd still need a way to vet drivers and handle refunds. I'm not sure other delivery apps bother with the former until bad reviews come in, and maybe the latter could be just between the customer and the store? Would be a big win for customers and gig workers if you could get past the obstacles.
The VC industry is like Hollywood: both smart purveyors of a popular, highly value added service to begin with, both driven by nepotism and inertia via far more scope for capture than your average line of work, and both with a cost structure mostly about the lifestyle of the executive class.
when it’s just starting out in something, you get instant classics.
when it’s out of ideas but still has the levers of access tilted all the way to “got mine bitches”, you get the late MCU, Uber for X, and OpenAI Larry Summers Edition.
The company Weee which primarily sells Asian products does this sort of batch order, and their prices are the same as grocery stores because they rightly knew that their initial Chinese userbase (and other Asians in general) is extremely price conscious and won't accept such increased fees for delivery. I just listened to a podcast episode about them on How I Built This which is where I got this information, people should give it a listen.
There is some degree of that in these apps already, for example I believe both grubhub and doordash will group up orders so one courier can do multiple deliveries back-to-back in a single trip.
> how come everything with an app is so much more expensive?
After ton of investment investors are finally looking for returns now. And consumers have decided convenience reigns supreme. Right from the SAAS products to daily breakfast customers there is lot of premium on convenience.
> After ton of investment investors are finally looking for returns now.
We need a quippy phrase for this, but unfortunately "Pump and Dump" is already taken for a kind of stock-market manipulation.
Instead, I would like to coin/nominate "Tease and Squeeze".
First consumers are teased by the idea that the disruptive new service has invented a magic new secret to low prices and great value... But after the company as achieved some kind of market strangehold, those consumers are squeezed with price-hikes and quality-drops as the investors--who bankrolled the early predatory pricing [0]--try to recoup their investment plus additional profit.
[0] Note that the term "Predatory Pricing" usually refers to artificially low prices to destroy competitors, although I can totally understand why some may assume it means preying on customers with high prices.
Granted, Tease and Squeeze has a nicer ring. Then again, the enshittification neologism became standard langue in no time and is understood by everybody.
I did consider the term when writing my post, but they only partially overlap.
First, the issue of intent and overall history. Tease and Squeeze implies it was the plan all along, but an "enshittified" product doesn't have to be that way. For example, a company could have been successful without any "Tease"--no predatory pricing or monopolistic behavior--and then taken a nosedive later when the founder retired and sold it to a new owner, meaning it's Squeeze-only.
Second, "Squeeze" doesn't require a quality-drop, it can include purely a price-hike, but "enshittified" products almost always mean a drop in quality, regardless of whether prices rise or not.
"Underpaid" and "fair" is a modern political myth, except in the instances when there is a concerted effort to undermine the semi-natural labor market. Pay will never be enough, as a point of fact. It will always be "underpaid" and "unfair" in discussion. As this is an immortal pressure and negotiating tactic. It is incentivized, and therefore we will get more of it. That works both ways, but at some point there has to be a reckoning that accepts that it is better to have employed than unemployed unskilled workers. In the context of a massive and growing unskilled population base. Yet the same people who demand that everyone earn a living wage, sometimes punctuating their points with riots, tend to be the same crowd who wants to import unskilled workers by the millions.
There is no effectively borderless world in which most are both employed and not "underpaid". It won't happen both ways. In fact, the pay trend is about to tip massively against fairness. In a manner that even will make economic rationalists uncomfortable. These "fair pay" efforts are a distraction before that storm.
A free market without political intervention is at least as much of a myth. If employees don't engage in politics with concepts like "fair wage" and "underpaid", employers certainly will with concepts such as "bottom quantile wages drive inflation". The Nash equilibrium is for both sides to engage in the market but also in the political context the market is embedded in. So complaining about workers demanding higher wages instead of letting the market sort it out is pointless, regardless of your political convictions.
I noted that there were two sides. Explaining it doesn't advance the conversation.
"...in the political context the market is embedded in" is nonsense, especially when appealing to the "market sorting it out" in the next breath.
Which argument are you recruting? Political intervention or the free market? Both? That's nice hand waving, no offense. Seemingly to make it as if your position is so elevated that it satisfies all political ideals. Instead, choose. Unless you mean to imply that the (free) market is also the political market. But your language should be a lot more precise if that's your point, so as to eliminate the look of vague line straddling as a rhetorical tactic.
Your rebuttal over-simplifies my response and therefore seems like a word dump.
Your view, absent your muddled (imo) rationale described above, is included in my response. The complicating factors were a). the common simultaneous advocacy for importing millions of unskilled workers and b.) the implied fact that increasing wages decreases available jobs.
Those factors are included in my total "complaint" (observation). Which is not a "political conviction".
Re-reading your comment I think we actually have the same understanding of the dynamic, I think I misread "immortal" as "immoral" before which gave the comment a quite different vibe.
> "...in the political context the market is embedded in" is nonsense, especially when appealing to the "market sorting it out" in the next breath.
I meant:
complaining about (workers demanding higher wages instead of letting the market sort it out) is pointless
Not:
(complaining about workers demanding higher wages) instead of (letting the market sort it out) is pointless.
Agreed my wording could have been more precise there.
looking forward to a world where everyone anywhere can make a decent living wage. The technology and productivity have improved so much but the wealth inequalities have only grown wider. If AGI arrives and we have abundance, lets hope that it is not only the few elites that control all of that wealth and power.
Maybe that "storm" is consequences of unchecked ML, make the rich richer and the powerful more powerful, push the poor out of jobs. I disagree that fair pay is impossible though.
Yeah you’re leaving us hanging here! To me it seems like people are learning their worth a little bit more lately and are demanding a little bit more/less financial stress.
I took it to mean that whatever amount of money people currently think would be needed to be "fair", say, an increase from $20/hour to $25/hour, would within a year or so become "unfair" again, even if inflation was zero. Employers will always want more work for less money, employees will always want more money for less work. Same as the seller of any product would like more money for that product, and the buyer would always like to pay less for the same product.
So you can say the company is "greedy"...but that implies that the worker is also "greedy" - which is true for both IF you define it the way I described above. Or you could say that the company (owners) are trying to maximize their return on investment (time + money), and the the workers are trying to maximize their return on investment (of time).
>Employers will always want more work for less money, employees will always want more money for less work. Same as the seller of any product would like more money for that product,
Is that true though? I come from a family that works in the crafts, small business mostly (which is like 70% of businesses). Most employees are employed for life, what's being made generally costs a decent amount but quality is high, people are content when they make what's perceived as fair and enough to maintain a middle class lifestyle, and so on.
Nobody actually tries to rip each other off, cuts every penny and leaves for 5% higher salaries. People are around for decades. The MBA business logic isn't some universal truth, in fact it's not even how most people work.
It'll "never be enough" in the sense that there is a non-existent potential of higher wages to stop agitation for more money.
Because there is no disincentive for labor to ever-advance the claim that their wages are enough.
In addition, the incentive to ever-claim that wages are not enough is unable to be quneched due to the fact that there is no risk. That logic is what supports my point.
The point can be proofed to economic mechanisms, as well. The exchange of money for goods naturally sets prices via an auction mechanism that, in turn, naturally creates an economic hierarchy in purchasing power.
Which is a verbose way of saying that raising the wages of the working class, while periodically necessary, will always lead to rising prices to create relatively par pourchasing power over time. When speaking of wages, purchasing power is all that matters (assuming lack of debt).
Artificially and dramatically forcing wages up only creates economic churn in job availability, housing equality, and asset ownership. All negative for the working class. If forces them into unemployment, out of urban centers, and creates firesale auctions for assets that wealthy people take advantage of.
To make matters worse, the "forced equality" economics of communism are only feudalism in disguise (which is hidden on the balance sheet of the State and Feudal Lord). The closer that the working class gets to forced economic equality, the closer that they get to slavery.
The storm that's coming is a cyclical return to various feudalists battling it out for workers who are paid the lowest possible wages and who have no labor protection. But on a wider scale than we are used to seeing.
The problem with wage debates is that there is no solution aside from a delicate balance that is easily corruptable. In fact, its natural state is corrupt imbalance.
This problem is unsolveable because it is inherent in the system of money.
This isn't what they meant apparently but in a very literal sense, pay can never be enough because it is economically necessary to underpay workers relative to what they contribute because profit by definition can only exist from surplus value, i.e. making more money from selling the product of labor than you pay for that labor.
If I pay a worker $20 per hour to make 20 doodads an hour in my doodad factory and then sell those doodads for $5 a piece, I make $5 for every $1 pay my worker. Of course that extra $4 per doodad has to account for the cost of the doodad factory itself (initial investment, maintenance, material) and the literal cost of doing business (taxes, permits, fees, paying an accountant) and of course market fluctuation (warehousing, building a financial buffer to remain solvent if sales lapse or production has to stop) but if the business is meant to be successful, that needs to add up to less than the extra $4 I'm making. And the bigger that difference is, the more profitable my company becomes and the more spare money I have to expand my business and get ahead of my competition up to the point where I can sell shares in my business to other people and pay dividends from my profits to them.
So in other words not only can't you pay workers exactly "enough" (i.e. 100% of the value they contribute to the company) but you're actively incentivized to pay them as little as you can get away with (and perversely, paying them more by reducing your profit margin may actually be worse for them by making your company less competitive and less resilient). This is before we even get into the politics of what work is overvalued or undervalued (e.g. the idea that CEOs should receive massively disproportionate compensation for their labor, which again creates pressure to compensate for this by underpaying lowly workers).
This, by the way, is why collective bargaining matters. No matter how cool and nice your employer is, the market actively incentivizes them to do the least for you they can get away with and punishes them for doing more. So by not taking advantage of collective bargaining (i.e. using the pressure of all employees to your advantage rather than just yours individually) you're leaving money on the table. Note that is still true if your role is proportionally overpaid relative to other workers in your company unless you literally co-own the company.
> pay can never be enough because it is economically necessary to underpay workers relative to what they contribute because profit by definition can only exist from surplus value
That's a silly definition of "enough" to label as a fact.
I feel like the lifestyle you can buy with the pay matters too, but even if we ignore that aspect, I think most people will agree that the worker getting 90% of the surplus value is "enough". (just as an example percent) There's no reason the bar should be the absolute extreme.
> That's a silly definition of "enough" to label as a fact.
There's no "factual" definition of "enough" as "enough" is subjective.
I'm also not sure what you're trying to argue. I already said that it can't be 100% for practical reasons and that the market actively incentivizes companies to aim for as close to 0% as possible instead. You pick 90% as an arbitrary percentage. So it might be 99% or 50% or 1%. But you're still agreeing that the measure for "enough" in this case is in relation to what the product of that labor can be sold for (after deducting the various costs). In reality that percentage is often much closer to 50% than to 90% - in many cases (especially so-called unskilled labor) it is closer to 1% than 90%.
A better rebuke would be that "enough" is normally about need, not revenue. Especially because it's nearly impossible to value the relative contribution to the bottom line for all labor involved (and in fact trying to minimize "cost centers" can lead to a rude awakening about this).
So if we define "need" as "enough to sustain a healthy existence in society and access to moderate leisure activities" (i.e. having a roof over your head, food in the fridge, access to medical care and transportation and enough spare change to spend on things like streaming services, books or the occasional trip to a nearby cinema, amusement park or restaurant) we can define a minimum monthly income and if we assume a 40 hour work week (which should be uncontroversial given that overall productivity has only gone up since its introduction and we need to provide some leisure time to accommodate family formation and the aforementioned leisure activities) we can translate that into a minimum hourly wage. Something tells me that this minimum hourly wage would be significantly higher than the legal minimum wage in the US, even if we don't actually consider the lackluster state of medical care affordability (note that I didn't assume public single-payer healthcare so Medicare/Medicaid is just a crutch to account for medical care being unaffordable otherwise at that level of income).
But of course the system is not set up that way. It's not about needs, it's about paying as little as you can get away with and this means that pay is proportional to power (i.e. C-level execs are paid a lot because they have a lot of control over the company and therefore "deserve" more even if all the value they "create" relies on the actual labor and competence of others) and the floor is theoretically infinitely low (i.e. you could realistically charge people to do labor for you, e.g. as a "training opportunity" to earn a meaningless qualification required to access other jobs) and ultimately the company needs to produce a lot of waste money (profits that are not re-invested in the company) that can be drained by its shareholders/owners (who ideally contribute literally no labor in return).
> There's no "factual" definition of "enough" as "enough" is subjective.
I agree. Which is why "enough" is not an objective threshold of 100% of surplus value.
> I'm also not sure what you're trying to argue.
I'm arguing that "less than 100% means not enough" is an incorrect statement. That's all. The bulk of your post isn't relevant to what I was saying.
(To be extra clear, that's my paraphrase of this line: "in a very literal sense, pay can never be enough because it is economically necessary to underpay workers relative to what they contribute because profit by definition can only exist from surplus value, i.e. making more money from selling the product of labor than you pay for that labor.")
> But you're still agreeing that the measure for "enough" [...] A better rebuke would be that "enough" is normally about need, not revenue.
But I did address that. "I feel like the lifestyle you can buy with the pay matters too, but even if we ignore that aspect,"
That is because I have not given a definition for it. I don't have a specific one in mind.
I just think "100% of surplus value" is a wrong definition, so using that definition to argue that "in a very literal sense, pay can never be enough" is a very flawed argument.
I don't think I need to offer my own definition to claim that, do I? They later said it's subjective, which seems to favor my claim.
Does that clear things up? Is there a reason you thought I was trying to give a definition, and failing? Is there something wrong with my argument?
Luckily we live in a society with at least some degree of freedom, enough to where if a worker is convinced that they are entitled to 100% of their self-percieved value of their contribution then they can and should start their own company. As what you are referring to is ownership compensation. Many workers do this.
Unfortunately, much if not most of the time this calculation by the worker will be in error. As they tend to exclude the company scaffolding that they are not responsible for but that enables them to make their contribution.
Either way, its amazing that there is so much legal opportunity today for anyone to build an asset (eg: business) that they own. Asset ownership being the key to legal and ethical economic success in this world. I feel like more children should be taught as much at an early age, with less emphasis on the "nobility of work" that tends to exclude the importance of assets.
If group ownership is important, workers can certainly begin a business with that in mind.
Last, stock is sometimes sold publicly (direct ownership of the asset).
> Unfortunately, much if not most of the time this calculation by the worker will be in error. As they tend to exclude the company scaffolding that they are not responsible for but that enables them to make their contribution.
I literally addressed that "company scaffolding" in more detail than you do here (second sentence of the second paragraph). There's no error here though as running the company and owning the company are two different things and while there are plenty of business owners who need to work for their own company to keep it running, we're specifically talking about surplus value after the cost of paying all labor (including your own, at market rate), i.e. what ends up either being reinvested at the end of the fiscal year or drained by the owners/shareholders as dividends. You may need someone to run the business side of things, you don't need that person to also own the company, be able to drain any profits and fire you at will.
> If group ownership is important, workers can certainly begin a business with that in mind.
We like to proudly talk about individualism when it comes to seeking praise for personal achievements but times like this demonstrate how it can blind us to systems thinking. Yes, you can create a worker cooperative. And yes, you can be a business owner who tries to be "fair" to their workers. But you exist in competition with other businesses who don't engage in such "waste" and you compete in a system where the "benefits" of such an approach are considered externalities and have no direct value (even if you try to argue it's "positive PR" that will only get you so far). In other words, the system disincentivizes and actively punishes redistributing surplus value to your (co-)workers - because it's literally based on profit-seeking (i.e. maximizing that surplus).
> Asset ownership being the key to legal and ethical economic success in this world.
Yes, and asset ownership is only attainable if you have the necessary capital to buy assets. Being self-employed won't get you far because you're still selling your labor for money (just with fewer middlemen but in practice not really) and if everyone owned their own business that's all they'd accomplish. If you want to be economically successful from asset ownership rather than labor, you want your "asset" to "do the work for you" (i.e. passive income). But in real-world terms this ultimately means you need to get a cut (i.e. profit) from someone else's labor (and they therefore can't take this advice or else they wouldn't be in that position for you).
I'm a business owner. I've been at various stages self-employed, an employer and a business owner using external freelancers and subcontractors. But I have no illusions that essentially the easiest "legal and ethical" economic success I've seen involved little more than either rent seeking (i.e. being paid for continuous usage of something you already built/bought) or scalping (i.e. selling something someone else made/did at an inflated price and pocketing the difference).
I mean it becomes a myth when you define it out of existence; I normally don't like semantic arguments but you're arguing against an idea of "underpaid" that means something different than what the people who say it are trying to convey. The notion of underpaid isn't weighed against the value of one's labor, it's weighed against cost of living and the value of one's time.
We've generally decided it's unreasonable for someone to need to work more than 40 hours a week so the minimum pay for 160 hours of labor had better be enough to live on without being on welfare or other government assistance. It might not be a glamorous life but you at least have to be able to make rent, pay for heat, AC, food, clothes, water, toilet paper, health insurance, a cheap cell phone, transportation to and from work, and some basic possessions like a bed. And if it's not then you're being underpaid. I think that's a pretty fair measure, it roughly represents the cost of a person's undivided labor.
And I think it's fair to say that for almost all Americans that cost has gone up and not insignificantly over the past 5 years. Folks that make good money by can take the hit but people on the bottom rung have nowhere to go. And people will given the opportunity begrudgingly sell their labor for unsustainable prices because it's better than nothing but that road leads to the record scratch in the economic game of musical chairs we're playing when money stops moving.
And that's when Socialist-inspired policies we have kick in and gov't has to forcibly redistribute wealth through higher taxes and even greater government spending to start the motor again. We've done it before and I'm sure we'll do it again. But I'm of the opinion that a little regulation to keep us out of the death spiral and letting the market allocate resources is better than when the government has to step in and do it.
The local Starbucks over the years has had many turnovers in the staff. It usually takes about 3 days for a new worker to learn the job.
That makes it unskilled labor.
Skilled labor is something like welding, where it can take a year to get good at it. Or things that require calculus, which take 4 years of specialized training to even get an entry job at it. Or flying an airplane - you can't just flip through the instruction booklet and fly an airplane.
It usually takes about 3 days for a new worker to learn the job.
Starbucks hires juniors and trains them to be good baristas. That takes a long time. Just because someone is able to work a coffee machine after a few days doesn't mean they're not skilled eventually. Your argument is like saying 'it only takes a few days for junior devs to be onboarded, so there's no real value in senior devs'.
It doesn’t take a phd to make a coffee, no matter how fancy the cinnamon/cocoa heart is.
All you're saying here is that you've failed to see the value in customer service, hospitality, speed, accuracy, politeness, etc that go with a retail coffeeshop job. The coffee is the same in pretty much every Starbucks, but the experience can vary wildly depending on how good the baristas are at everything else besides making coffee. Those skills take years to master.
Yes, that’s what I’m sort of saying. I haven’t failed to see any of that. But it doesn’t take a phd to make coffee politely or less politely. All I want is the coffee. Don’t ask me what my name is. I don’t care about the „experience”.
You do care about the experience though. Everyone does. If you have a bad experience you'll complain, or stop going to that Starbucks, or moan to your friends. To you, by the sounds of things, a good experience consists of asking for your order, paying, and getting out fast. An experienced retail worker will be able to read all that from your cues - not engaging in small talk, having your payment ready, moving along quickly, etc. They'll factor those things into their interactions with you, they'll remember you next time, and they'll make the experience what you want. All that takes time to learn, and few retail workers get good at it. I will concede that it doesn't require a PhD though. Few jobs do.
I don't go to Starbucks anyway because they just keep that stupid ritual of "what's your name" and they can never understand what I'm saying so we are just standing there on the opposite side of the counter exercising our knowledge of the military alphabet. If that's the sort of skill required, sorry, anyone can do it. Tell me - how long does it take to serve your first coffee at Starbucks as a barista?
Just sell me the damn coffee and bugger off with "experience". It doesn't require a "skilled" person to politely serve me a coffee. Just make sure the coffee doesn't taste like shit and I'll come back to you unless you spit at me or offend me. Again - no need to be "skilled" to not do that.
I think people talk past each other because one side takes "unskilled" to mean basically "anyone could do it with some training" (I can't run an espresso machine anymore, but I learned how in a few hours when I needed to). You get better at unskilled labor (anyone knows this, there are baristas who are much faster than others; able to run the entire shop which normally takes two).
The other side takes "unskilled" to be a derogatory term that means the worker isn't a human being with thoughts and wants and needs, and deserving of a wage.
How could you possibly know that it takes “3 days” to get good at being a barista? You should try it out sometime, with a 50 person line. I’m sure it’ll be a cake walk and your labor will be fairly compensated. I’ll even give you a 3 day head start just to be fair!
You can do the basic job after 3 days and after a month rush hour won't be a problem either. I've done similar jobs, the skill cap is not that high.
My labour won't be fairly compensated which is why i don't do jobs like that anymore, but they're not hard to learn.
To be fair, Starbucks purchases machines that are automated to a much higher degree than your standard high end espresso machine you'd find in a local shop. It's a lot more button pushing in the correct sequence, a lot less art. This is standard procedure for a multinational fast food service company: they want things automated and streamlined not only for speed of service and easy training, but consistency of the product.
From what I heard from a former Starbucks barista, the biggest challenge was to not scald the milk during steaming (which is not very difficult, basically just requires holding it at the correct angle to induce a whirlpool effect and keep things moving).
The skilled aspect is dealing with customers efficiently and pleasantly, and all the nuance/craziness that can entail.
They likely meant that onboarding takes three days. As there is no meaningful metric for competence beyond that (or at least no metric that is measured and used) "good" doesn't apply.
There are of course tons of ways a good, seasoned barista distinguishes themselves from a trainee with 3 days of onboarding but those are externalities. The difference between skilled and unskilled labor is not whether competence and experience can make a difference but whether that difference is measured as performance or only affects externalities.
I.e. "unskilled labor" is not about the worker but about the company. Notably these often do involve skill that directly contributes to performance but because the job is considered unskilled, it's instead treated as some nebulous a priori form of intelligence and personal aptitude. What's more, beyond a certain point skill is often actively punished (e.g. by raising quotas to take advantage of higher productivity, resulting in more work for the same pay).
>> I.e. "unskilled labor" is not about the worker but about the company
That part really stood out to me.
While a lot of us can agree that barista is an unskilled job, I heard people call barmen labor a skilled labor.
Which is crazy, because it's essentially the same job (mix stuff up and serve in a cup). But it makes sense if "unskilled labor" is a function of an employer not an employee
Knowing professional welders myself, and trying my hand at welding, it takes a lot more than 30 days to get good at it. A welder also needs to learn some chemistry, all kinds of techniques, the strengths and weaknesses of various kinds of welds, how to prep the weld, etc.
A good weld is a thing of beauty.
Welding is also pretty dangerous. It takes a good welder to do it safely, and do it without ruining very expensive parts.
There's good reason that competent welders get paid a lot of money.
In 30 days, the new welder probably has learned how not to set himself on fire, blind himself, fill his lungs with poisonous gasses, etc.
In any job there is a spectrum from bare minimum to a master. This goes for welding and baristas. Is someone who pushes a button on a Nespresso a barista? If I go get a soldering iron am I a welder?
You overlay the normal distribution of all baristas and welders and look at the difficulty in learning various skills. I'm sure they are not exactly equal.
That sounds like a fun idea in theory, but in practice that distribution would require hours of research per job, or more, and I'm not aware of any source that publishes that sort of information. So it doesn't give me a way to sort jobs into "skilled" and "unskilled". Or to give them a reliable 1-10 rank on how skilled they are.
I can be pretty sure welder beats barista, but by how much, how it relates to other jobs, how we set various bars, that's all a lot more difficult.
I actually doubt there's a neat defineable ceiling for the skill of a welder nor a barista so it also comes down to where you make the cut-off because any improvements beyond that become negligible in terms of ROI. But that would also require being able to actually quantify that vague notion of "skill" which in turn would require an honest assessment of the actual real-world exhaustive job description of both jobs including all implicit, unstated and culturally expected parts (e.g. barista is a service job so customer service can be as big a factor as the actual coffee-making but we don't think of social skills as actual skills or emotional labor even if it's obvious when they're inadequate or badly performed).
So in other words, we could have an objective, empirical look at whether welders or baristas actually need "more skill" and which skillset is easier to master for the average person (which is its own rest nest of considerations again) but if we did the necessary (if not impossible) prep work to even get started on that, we'd have to already agree that "unskilled" is a descriptor for jobs that has very little to do with the actual skill requirements and more with attitudes towards those skills or the people doing those jobs.
Sure but unless you're Tesla, welding quality is more relevant to your bottom line than coffee making quality (yes, I'm aware there's more to being a good barista than just making the coffee).
> You can onboard welding in less than 30 days too.
You can be doing useful work as part of a welding group in 30 days, but you will not be a master. Grinder and paint make me the welder I ain't and all that.
I couldn't reliably tell you the difference between a $30 bottle of wine and a $300 bottle of wine but that's why I don't claim to be an expert on wine.
Also, there are other coffee places than Starbucks. Saying barista is an unskilled job because Starbucks coffee is particularly bad is like saying programmer is an unskilled job because all Squarespace websites are equally bad.
> How could you possibly know that it takes “3 days” to get good at being a barista?
Because I have gone to the same Starbucks every day for many years, and see many new hires as the staff turns over. It takes about 3 days to go from "cannot operate the cash register" to "efficiently fulfills customer orders".
Unskilled labor simply refers to jobs where any specialized skills required can be learned on the job in a short period of time, usually less than 30 days.
It doesn't literally mean the workers don't know how to do anything. It's certainly not a myth - it's a classification of work, at least in the US.
> Unskilled labor simply refers to jobs where any specialized skills required can be learned on the job in a short period of time, usually less than 30 days.
You're not going to learn welding in 30 days. Nor are you going to learn how to design a bridge. Or program in C++. Or diagnose a patient. Or be a lawyer. Or drive a race car. Etc.
I didn't say all jobs, did I? You successfully listed (some) jobs that might take longer than 30 days to learn.
The thing about "unskilled" jobs, everyone seems to look down their noses at, is that they facilite the people doing the lofty jobs of: lawyer, doctor, c++ programmer, etc. Without them getting done for you you couldn't do what you're doing.
Can the brickie work without their "unskilled" laborer? Sure, but good luck getting you house built in a reasonable time.
I don't understand your point. Yes, there's a lot of unskilled jobs, possibly a majority of jobs. Far from everyone looks down on unskilled jobs, and nobody is saying they're not important.
The brickie's "laborer", I literally said it in the post you replied to; the person who moves bricks, makes up the muck, generally keeps the brickie supplied with materials.
Credentialism has certainly gone too far in many fields, but I’d still like my doctors, lawyers, and engineers to have more than 30 days of training in their field.
Where are all these js devs on minimum wage going paycheck-to-paycheck?
I think everything is easy, given the right approach and mindset, or everything is approachable given the right approach and mindset and will to put enough effort to get to a given outcome.
The concept of "underpaid" and "fair" wages is fundamentally subjective. Is a burger flipper underpaid because he cant raise a family a buy a house on that wage, "underpaid"? Or is his labor simply not that valuable? Is a techbro making $300k "underpaid" because his employer is making $500k off his work?
On the other hand skilled vs unskilled labor, as well as the concept of human capital are well recognized concepts in economics.
That's just playing with words. Putting Gordon Ramsay in the same bucket as "burger flipper" makes as much sense as putting Linus Torvalds in the same bucket as "keyboard monkey".
Why are you pivoting to "value"? The original discussion was about skilled vs unskilled labor and whether that assessment is subjective vs objective. You might think that accountants are useless paper pushers whereas burger flippers are Hard Working People That Get Actual Things Done™, but that's orthogonal to how much skill[1] is needed to flip burgers vs be an accountant.
So you're trying to derail the discussion to something about "value", because the page I linked about "skill" has a passage about how skilled laborers were historically important to the economy? What does this have to do with the subjectivity/objectivity of "underpaid", "unfair wages", "unskilled labor", or whether a burger flipper is more "skilled" than an accountant? As I said earlier, even if you think that accountants are useless paper pushers, the fact that they're pushing papers in a very specific way that takes years to learn, makes them more skilled.
First I'm pivoting now I'm trying to derail, when all I've done is answer you question; but to make it plain: yes, it's subjective. How else are certain skills valued more that others? There's no objective measurement to these. Comparing the skills of one disapline to another doesn't work, like apple to oranges.
The objective measurement for "skill" is how much training/experience/talent is required to carry out a particular job. Sure, there might be some fuzziness/ambiguity to this, and there's various degrees of freedom to how you compute a "skill score" or whatever (eg. what's more skilled an accountant or an auditor?), but it's hard to argue that a burger flipper is more skilled than an accountant. You can make the argument that the value of a burger flipper vs an accountant is subjective, but that's irrespective of the skill required.
Whether someone is getting "underpaid" or a "fair wage" on the other hand is entirely subjective, and you can come to whatever conclusion you want depending on your politics. On one side of the spectrum you could argue any sort of situation where the employer is capturing surplus value from the employee is inherently exploitative[1] and therefore "underpaid" and a "unfair wage". On the other end of the spectrum you argue that supply and demand curves are the ultimate arbiter of what's "fair", and any wage that is determined by the free market can't by "underpaid" or "unfair" by definition.
Yeah I don't think we're going to agree about skills in this, especially since I'd consider usefulness as a main objective measurement of skill. Using your "objective" measures could lead to a juggler of knives having the same level of skill as your accountant.
> With the two examples you chose, I‘d say there is not much of a skill gap.
Unless we're talking about really high end burgers, you can take almost anyone off the street and train them to flip burger patties within a day. They might not willingly do it on account of it being boring/tiring/poorly paid work, but it's not exactly hard to learn either. I doubt you can do the same for an accountant, unless your idea of an accountant is something like "manually copying entries into a ledger". Even teaching excel to someone who hasn't used excel ever, to a capacity where they can do meaningful financial reporting probably can't be done within a day.
On a relative scale, I agree with you that the amount of training involved differs for accountants and burger flippers, thus this is a good example.
On an absolute scale, comparing skills of burger flippers, accountants, aeronautical engineers and surgeons, the first two basically lump together.
I look at skill gap more in terms of „how hard is it to completely automate/autonomize this job“. Which is fiercely easy both for the burger flipper and the accountant, yet a bit harder (though not impossible) for the other two.
With all the increased productivity and advances if we can't support decent living for every person then what's the point? Work yourself to death to make some motherfucker rich?
> "Underpaid" and "fair" is a modern political myth, except in the instances when there is a concerted effort to undermine the semi-natural labor market.
No, and I'm guessing you're just under the spell of market fundamentalism. An unfair wage is something like when the worker either
1. can't achieve a socially-acceptable living standard with the job,
2. when there's too-great of a mismatch between the labor-theory-of-value price and the market-theory-of-value price for their labor, or
3. or when the wage is below the actual the market price.
I made those up on the spot, so there may be some holes, bit it gets the gist.
> Pay will never be enough, as a point of fact.
Being underpaid is different than "wants more money."
> It will always be "underpaid" and "unfair" in discussion.
No. I'd never talk about a fortune 500 CEO or a FAANG engineer as being underpaid.
> Yet the same people who demand that everyone earn a living wage, sometimes punctuating their points with riots, tend to be the same crowd who wants to import unskilled workers by the millions.
I see what you're saying, but the real world is a lot more morally complicated that the single-axis world where your comment here makes sense.
Your "guess" before your arguments is a pointless jab and ploy to win a point without having yet said anything.
If you think that accusing me of "being under the spell of market fundamentalism" is a succesful rhetorical strategy, then you neither understood my post nor are you worth arguing with.
My comment was made in the context of mentioned complex factors that you are ommitting.
Your 1 and 2 numbered responses are pure nonsense. Response 1 reveals to me who I am dealing with. This is going to go nowhere. Response 2 is ideological hand waving. 3 is true, but I included points that addressed it in my post.
There is zero political (effective) difference in the labor market between being underpaid and wanting more money, if the ask is for more money.
If they are truly underpaid, then they should ask for more money and get it. If they are not underpaid and ask for more money lest penalty x, then the market effect for employers is the same.
A further problem, as illustrated by your definitions of underpayment with which I disagree, is who is to decide what being underpaid means?
The answer to that question is that what constitutes underpayment will always be murky as far as the market is concerned and therefore, again, my point stands. Pay will never be enough, as a point of fact.
>No. I'd never talk about a fortune 500 CEO or a FAANG engineer as being underpaid.
"Never". That's a perfect example of the self-centered view of much of labor advocacy. Not only do they publicly only take the other side of the market into account when it opportunistically suits them (ie: when bucking for asset ownership or when owning assets themsevles), they have trouble comprehending how their own principles might scale to different skill levels in the labor pool.
> I see what you're saying, but the real world is a lot more morally complicated that the single-axis world where your comment here makes sense.
I disagree with your single-axis categorization. I disagree with the implication that this is largely a moral argument.
I disagree with the assumption you would be able to make moral assertions that had any impact on the economic forces that determine pay and job availability, given the mentioned constraint (mass immigration). My original point is that the labor market will not be able to have it all. It has to choose, and clearly it has made its choice.
But in its defense, the pro-labor market was co-opted generations ago by Business. That dissonance that you percieve and interpret as "moral complexity" is exactly their influence.
Problem is, when you get an economy acclimated to false prices set by adversarial “business” “strategy”, the only logical result is depression when it’s time to actually create money for investors
Fair warning to all who come after - the children comment for a decent distance below are a battle based on correct/ incorrect/ technical/ personal definitions of skilled labor.
Does that warrant a warning? That's just a natural occuring discussion, and selecting threads to follow is basic forum-reading skill. The rare gems of surprising insight are often found in the depths of slightly off-topic discussions.
Previously I've noticed another kind of meta-comment, the top level complaint by a self-appointed moderators about the overall shape of the discussion elsewhere, instead of engaging directly with the actual posts in question.
With intellectual pursuits, you can always find diamonds in the rough. Simply due to the effort YOU are putting in, given your then state of mind, knowledge and experience.
You can wade through a conversation on creationism and evolutionary denial and understand humanity or even deepen your skills.
However that's you - your unique circumstances bringing more to the table.
Others would very much appreciate a warning, because they already have working definitions, have seen this argument before and would prefer spending their time on other pursuits.
With a little forum experience, who has trouble scanning a thread and seeing it meander away from their interest? So I question the need.
Even if there is the need, who should be the judge for others over the quality of a subthread? I would feel presumptuous to pass judgement beyond my vote for each comment.
Aren't those diamonds all that matter anyway? Recently I increasingly find new and curious beginnings dampened and limited by intellectually lazy attitudes, by comments that start with "I mean...", that reiterate the default state rather than entertaining a new thought. Inadvertently, well-meaning soft-moderation like yours might steer even more people away from interesting topics.
wrt DoorDash (and Instacart etc), along with this squeeze I’ve observed the quality of service decrease severely. Could just be that my sample of a now wider employee pool is crappy, or workers are trying to “scale” to make a living wage with many concurrent orders, but I believe it’s part of the race to the bottom. I love the convenience, but the experience is worsening. How can these vendors continue to profit when lower quality leads to increased refunds?
Well in my observation people keep adjusting lower quality of service. I mean it is not going to be the case that people will start shopping in stores and/or cooking in kitchens.
How can you tell that it's wasteful? Seems to me that pointing at activities and saying they're wasteful is exactly the central planning you mentioned. Everything "decadent" looks suspiciously wasteful. Stage shows, exotic fruit, ball games, wedding photography, hats, flowers, songs, pets, tourism, novels, hobbies, diamonds ... it often puzzles me how economic activity makes a country wealthy when only a tiny part of it is efficiently providing health, nutrients and shelter. And in fact maybe 90% of it is wasteful, but nobody has the magic ability to identify exactly which 90%, and that's why central planning wrecks wealth, and VC funding apparently thrown down the toilet may in some mysterious way be beneficial to us.
Yeah, I plan on paying approximately double (100% markup) if I'm ordering delivery on an app. That's why I only do it when I'm in a hotel room on a work trip and I don't have the rental car keys. Hotel food is not cheaper, not tastier, and I have way less options (I am a vegetarian).
> Though it seemed completely automated, Just Walk Out relied on more than 1,000 people in India watching and labeling videos to ensure accurate checkouts. The cashiers were simply moved off-site, and they watched you as you shopped.
Wow - I did not know this. This makes it all a whole lot less impressive and interesting that it was just people off shore watching you.
Apparently this news came out in May 2023. I also was bamboozled into thinking that sophisticated computer vision algos -- that worked -- were doing this.
I'm now picturing the remote workers constantly switching between cameras, studying: Did he put down the can of kidney beans or the can of corn there? Wait, the man picked up the bananas, but then he handed them to the woman in the white shirt. Let's charge them to her account. Wait, she handed them back at 14:42 in aisle 16. Going to switch them back to the man.
All day long, day in day out, for years. I am the last to criticize 'low wage jobs' paternalistically, because I know they may be much better than what the workers might otherwise be doing: perhaps just toiling in the fields for 16 hours a day (or worse: something like 'melting down discarded PCBs to recover trace metals'). But still, I do not think I would want to do this job nor that it was worth it. I get that they were supposedly trying to train an algorithm to do it. I'm glad that they aren't keeping at it any longer though now that it's proven so unworkable.
Edit: On thinking about this tangent, it seems that if global regulation is the only solution, it would make sense to enforce regulation from the ship building side (in western countries), rather than the wrecking side (in developing countries) which will only displace the bad practices to less scrupulous countries. A solution might be sizable amount of money that had to be paid to escrow that could not be released back to the owner until the ship has been disposed of in an environmentally sound manner, incentivizing and funding the proper scrapping of ships. Or perhaps a levy which funds safety practices and equipment for scrapping companies
Uh, really? I was certain that the aftermath of WWII was responsible for establishing a form of global governance (by consensus) via the UN, and then there’s the benefits of free trade agreements that drove economic globalization.
The free trade agreements came after the environmental movements. Claiming environmental movements caused this is too strong a claim. But it is not entirely a coincidence that one followed the other.
Specifically what movements and when? It’s a strange claim that globalization was caused (or even accelerated) by environmental movements. Certainly the UN came before the environmental movements of the 60s/70s. The World Bank and IMF were also established soon after the conclusion of WWII. The wiki article for globalization makes just a couple of off hand mentions of environmental issues. I’ve only ever heard of globalization in an economic context, and I think by convention this is the lens most people view it by
When Americans clamored for rivers that didn’t catch fire and for smog in LA to go away it became clear to corporations that it would be much cheaper and better for profits if they setup factories in poor countries that didn’t require them to stop egregiously polluting. Thus began the momentum for free trade agreements and breaking down of trade barriers.
I used the term “globalization” as a proxy for free trade. That was bad on my part.
I can see how environmental policy could accelerate this phenomena, but surely the asymmetry of labor costs and cheaper/faster shipping is reason enough to offshore labor? Economical shipping seems to be the enabler, and miserly humans, as ever, the cause
The ownership structure for ships used in international shipping is anything but straightforward. For instance, for the the MV Dali (the ship that crashed into the Francis Scott Key Bridge) you might think we can go after Maersk, but in reality they're only chartering it. It was actually built by Hyundai Heavy Industries in South Korea for a Greek company but later sold to a Singaporean company, operated by a different Singaporean company, and crewed by 20 Indians and 1 Sri Lankan. In this complex web of ownership/relationships how do you exactly "enforce regulation from the ship building side "?
Well as I (naively) suggested whoever owns the actual ship (and thus disposes of it) can claim on an escrow that isn't released until the ship has been verifiable disposed of in an environmentally sound fashion. So if the ship ever is sold to another party, that would be built into the purchase cost (that they could claim this money), even to the final purchaser (the wreckers). Probably some huge loophole or perverse incentive that I haven't thought of, but that's at least one suggestion.
Due to heavy regulations on the shipbreaking industry in Western countries, > 50% of ships worldwide are (or were, when the documentary was made) dismantled in Alang, India.
The industry is extremely polluting (to the environment and the workers) and the working conditions (incl. safety) are dire to put it nicely.
If you did some kind of shell game with 5 different people all passing around their items, then you don’t get charged correctly is it theft? Did anyone ever try that?
I worked for Amazon in Seattle when Amazon Go first launched. As you'd expect, lots of SDE teams made games out of trying to fool the thing in various ways.
A few attempts were successful early on (passing items back and forth, one person moving something to the wrong shelf and another person picking it up, people dressing in identical outfits, etc), but the success rate in fooling the system was very very low. No method of trying to trick it that I ever heard of worked consistently, and it definitely seemed to get harder to fool the longer the store was open.
At the time I thought that whatever algos were being run on the camera feeds were getting better. Knowing that it was basically all manual, I'm not sure what the explanation is for the store seemingly getting harder to fool over time. Possibly just placebo, or people lost interest in trying so hard to fool it.
Probably fewer patrons and similar number of employees in the sweatshops. More eyes per patron leads to fewer errors. Or maybe they trained the best grocery cv model around from having a big high quality dataset, and you were fighting it. But then I'd think the tech would've been passed to whole foods instead of packing up shop, so final guess is sweatshop singularity theory.
> If you did some kind of shell game with 5 different people all passing around their items, then you don’t get charged correctly is it theft?
IANAL but I'm confident the legal answer is "yes", the same as if someone was using sleight-of-hand with objects at a cashier-and-conveyor-belt checkout station.
Whether charges are brought and how easily the case can be proved is another matter, but the intent is what makes it a crime.
It sounds different to me because each person can say they were just trying to buy the items they walked out with and didn’t get charged. Like if you scan an item at self checkout and the machine says it’s free, that’s not going to be theft. At least I sure hope not.
I thought one of the marketing lines was that Amazon was so confident/comfortable in their implementation was that they assumed all liability for mistakes (though as I type this I realise they may have worded it to encapsulate only honest mistakes rather than people trying deliberately to break the system).
To get a little pedantic, assuming such a promise existed, it actually doesn't mean as much as most people think.
A merchant saying "we won't sue you in civil court for the missing money" does not prevent the local government from criminally prosecuting that same person for theft.
American TV dramas often show the police asking people "Do you want to press charges?", but the idea that the question matters is a myth, since victims of crime don't get to decide that. At best, it's a terribly misleading shortening of: "Just for my own private curiosity, do you plan to lobby or press upon your local government officials into pressing charges?"
>though as I type this I realise they may have worded it to encapsulate only honest mistakes rather than people trying deliberately to break the system
Exactly. Retailers generally assume liability for honest mistakes made by their system/employees and sometimes even their customers. However, when people knowingly exploit a loophole for financial gain it becomes fraud. Here's a real example - https://www.wfmynews2.com/article/news/local/greensboro-woma...
> the workers might otherwise be doing: perhaps just toiling in the fields for 16 hours a day
That is robber barron propaganda to make you believe their enslavement of people (by creating the right situation where the people have no choice) is actual good for them.
I originally wrote a long reply, but I'll just say that I don't buy your framing, and I don't think the people who depend on that foreign money flowing in to pay them for their work at the prevailing wage where they live agree with you either.
Right... so the line goes. They're free to go to any neighbouring city that will do the same to them now that we've destroyed the village (by poaching young people in various ways. There are many ways to destroy a city if you are strategic and have financing.)
In these countries the cities are created around the factories etc and planned as part of the development.
You're not free once your subsistence relies on this corporation. It's slavery but because some tokens are being transferred (the food supply is probably also corp. owned--especially in the early stages of the development) it is possible to fool the naive / stupid into think this is not slavery.
What if the alternative to these jobs is not other jobs, but no jobs? If working in such a job is considered slavery, then most of the west also employs slaves in the form of "simple, low skilled" jobs, such as cashiers. Many people don't have many options, and losing even such a job could mean unemployment.
Pratchett really got technology, imo. Sometimes it really is high energy magic, but most of the time it's just labourers you can't see being exploited. I especially enjoyed the line "money dangled is far more effective than money given" or something like that... it's true.
> Though it seemed completely automated, Just Walk Out relied on more than 1,000 people in India watching and labeling videos to ensure accurate checkouts. The cashiers were simply moved off-site, and they watched you as you shopped.
Reminds me of the "delivery robots" that weren't really automated after all, they were remotely navigated by cheaper workers on playstation controllers in Brazil and the Philippines driving them on the streets through cameras.
It's just a tech-illiterate journalist who can't seem to understand the difference between "annotators watching and labeling videos to validate the model" vs "people watching the videos live to remotely decide the cost of every user's purchase".
Or maybe they do know the difference, but wanted to bait audience.
If you required a high accuracy like 99.9% to charge a customer, you could have a system that was mostly automatic but still needed human review when the model isn’t confident enough. It’s hard to know exactly what this means without a lot more details, which Amazon is unlikely to provide.
Wonder if that 30% include the cases where someone walks in, strolls around, and walks out without buying something? Which I see people do all the time, interested in the concept.
If it was just a matter of building up a large enough training set and corresponding model do you really think Amazon would be abandoning a technology they spent so much time, money and resources on, in a somewhat embarassing manner?
Sounds like fake-it-till-you-make-it stuff right? This is exactly how I would bootstrap a startup that wanted to do this. Kind of impressed with their scrappiness to be honest.
What I think is funny is that circa 2008 I had a manager who used to work at Amazon who told me that "a surprising amount of Amazon artificial intelligence is artificial artificial intelligence, low paid workers".
I heard this was behind mechanical turk. Sounds like the playbook remained the same.
I'm pretty sure this idea has been kicked around, and is generally seen as not-super-useful because it requires a perfectly reliable low-latency connection.
Phantom Auto did this. They had cars ferrying people around CES 2018 in Vegas driven by remote drivers in LA. Apparently the company folded just a few weeks ago.
Regular grocery stores have really gone downhill as well. Whenever I shop, there’s at most 1 cashier doing checkout and usually 0 (only self-checkout being open). I consider myself pretty proficient about knowing what sets off the machine but still set it off 60% of the time (about some weight imbalance etc) that requires an attendant to come fix manually. I’ve gotten to dread the grocery store trips as they require so much overhead time. I really wish the “just walk out” could’ve been popularized and caught on at more stores.
I live in Western Europe and the self checkouts don’t have scales, except for weighing produce, and even for buying alcohol, they trust you to say you’re over 16. It’s always fast because you wait in one line for 4-8 self checkouts and take the first free one.
On the other hand, using the cashier is often frustrating. You have to self-bag anyway and the cashier won’t start scanning your items until the person in front of you has bagged all of their stuff, had a chat about the weather and counted their change. The bagging area is always way too small, making all of this take a while. A few times I’ve even waited while one cashier counts their float, leaves, and a new one comes and counts their float.
I travel extensively, and I think attitudes towards self-checkout vary heavily depending on what preceded it where you are.
In some places, you had smoke-stained cashiers throwing your groceries to their side without so much as looking at you, except to complain you're not bagging fast enough. I think it's fair to say that in those places, self-checkout is a huge improvement.
In other places though, where cashiers have traditionally been more friendly and expected to bag your groceries, it feels like a less obvious improvement. Whereas before you had chipper high school kids working their first job bagging your groceries, now you have menacing terminals accusing you of being a criminal every time their scales misalign.
Self-checkout is a convergence on mediocrity: some places are going to get dragged up, and others down.
Great point, I’m from Canada and lived in Japan for a decade, in those two places I would say it’s a downgrade. Especially Canada where there is usually a scale.
> the cashier won’t start scanning your items until the person in front of you has bagged all of their stuff
What happens every now and then to me is that the person in front watches open mouthed as their shopping gets scanned and piles up in the bagging area, then they pay and fuck around with their purse (honestly idk what they’re doing) then take forever to pack, while the cashier starts scanning my stuff and mixes it into theirs.
I don’t love self-checkout, it can be annoying when something doesn’t scan and I don’t have the ability to manually input the bar code like a normal cashier does. And sometimes there can be errors (“place the item into the bagging area” - when I already did), but I’m overall glad to have the option for self-checkout.
I've been to quite a few supermarkets where they have like a plank that separates the new shopping from the previous shoppers. Maybe advanced plank technology will spread and deal with this issue.
Honestly I think we’ve all been there at some point - just switching off for a few seconds when we’re stressed or distracted by other things in life. If not at the checkout then somewhere else.
But yeah when you’re the other person in this scenario it’s irritating as hell :-D
Damn I’m in Southern California and y’all have really bizarre checkout experiences to me
- I never have to weigh anything except produce
- I rarely have to wait on an attendant except for something like spray paint
- I rarely have been to a store with absolutely zero cashiers, especially for a grocery store
- Usually the cashier will bag for me
I’ve had similar experiences whether I’m in an urban part of LA or some rich Orange County suburb.
My only two complaints are (1) finding parking at like a Trader’s Joes or Tokyo Central and (2) when the store is super busy, there are a bunch of cashiers, and yet there are still long lines
There is one grocery store near me that always has at least two staff members ready to work the checkout and no self checkout at all. There are virtually never lines more than two persons deep. The only time I ever go anywhere else is when I absolutely have to. Vote with your wallet.
Two of the main grocery stores in my area, one of which has sufficient cashiers, are trying to merge in order to make a worse shopping experience for everyone -- at least I assume that's the goal. The third, the only real alternative, is Walmart. Sucks.
Yeah, the "if anything goes wrong you now stand around with your thumb up your ass while one employee makes their way from broken kiosk to broken kiosk to manually resolve the problems" model has soured me on self-checkout. I find I have a very low "dealing with this bullshit" limit, to the point that if I have trouble I'm likely to just say fuck it and walk out of the store without completing my purchase.
Having used the UK/US type self checkout machines while travelling, I must say it is so nice to live in Northern Europe where the self checkout machines are largely just based on trust, with randomly sampling ~five items of every 100 shoppers. There is no weighing at any of the stores.
If you don't get randomly selected for inspection, there is nothing stopping you from just walking out with groceries you didn't pay for worth a hundred dollars easy. People just don't. Another benefit of having a proper social safety net I guess.
Social safety net is more than just being able to afford it. It's a combination of not being able to afford it and general erosion of social values leading to "take what you can get away with" mentality. So even those that can afford it will still try to get more since it's everyone for themselves.
Instead of cooperation under a system with a social safety net, we get a game theoretic weighting on defection without a social safety net.
> Social safety net is more than just being able to afford it. It's a combination of not being able to afford it and general erosion of social values leading to "take what you can get away with" mentality.
There's a causal arrow in the other direction, IMHO: People like to believe in the 'law of the jungle' mentality and thus vote to cut the safety net. IME it's especially wealthy people for whom it's a kind of game, not trauma, deprivation (of education, health care, food, housing, safety) and survival, and who also get to pay lower taxes, Didn't Jamie Dimon talk about how he liked how 'animal spirits' have been awakened?
> Instead of cooperation under a system with a social safety net, we get a game theoretic weighting on defection without a social safety net.
There is a safety net. Here's how I'd describe the situation:
There will be a number of people, fewer than the total who benefit from the safety net, who almost everyone would say it was good that they were helped.
There will be a number of people, fewer than the total who benefit from the safety net, who almost everyone would say it was not good that they were helped.
Most of politics around this topic is people's differing beliefs in the proportions of the above, and how bought in they are to minimising the latter count because they pay taxes.
> who almost everyone would say it was not good that they were helped
I'm sorry, what? Only in the US helping someone can be sold as a bad thing. It's better to help a few more people that may not really need it than help a few fewer that actually did.
> I'm sorry, what? Only in the US helping someone can be sold as a bad thing. It's better to help a few more people that may not really need it than help a few fewer that actually did.
I think it's worth reading the complete comment and you'll see where what you said fits my wider point.
That's because all of the self checkouts in the US use the weight change to match against the scanned item and any discrepancy leads to having to wait for a self-checkout attendant to manually override to let you continue. That manual override is equivalent to the inspection you're talking about not seeing. And I in the US get blocked needing an attendant 60%+ of the time at some point in the process. Usually up front with the bags I brought from home.
That's happened maybe 4 or 5 times to me in the 200 trips or so (estimated, once a week for years) I've made since I started using the self-checkout every time. It's so reliable and fast for me (store: Jewel-Osco), I just don't have the problems others here are seeing. Usually while waiting in line I don't see anyone else have issues either.
Remove from basket, scan, put in bag. Remove from basket, scan, put in bag.
The scale is the entire bagging area, and it detects when things were removed or added without being scanned, as far as I know that's all it does here. No specific weights.
At the Target next to where I live, the self checkouts don't weigh anything at all. I just scan something and put it in my backpack on the floor most of the time. And this is a Target where the toothpaste is locked up! I've never needed help from an attendant, and I've never been inspected either.
>That's because all of the self checkouts in the US use the weight change to match against the scanned item
Objectively false. Some stores in some areas do that. Most don't.
At Walmart if I only have a few items (most of the time I go to Walmart) I'll just grab the scan gun and scan everything without taking anything out of my cart. I'll bag when I get to the car.
They recently installed a little door at my self checkout. I cannot even walk out of my grocery store without scanning a receipt or flagging an employee to open the door. It’s frustrating.
I know that Krogers has cameras that use some sort of intelligence. I held an object in my hand as I moved other items across the scanner, it set of an alarm and a bunch of still images of my checkout process came up with colored shapes highlighting different on screen items
Might be your neighborhood -- stores with a low shrink rate might be more lax about inspections. I've only encountered it a few times myself but I'm also not the primary shopper in my household.
You can't makes broad, sweeping generalizations like "UK/US type self checkout." It varies so much.
I live in the US and the only place I ever go to where they weigh the items you scanned is Costco. Everywhere else I go (my local Walmart, Target, Aldi, two local grocery stores, and CVS) the self checkout is mostly based on trust with maybe one employee somewhat monitoring some machines. No random sampling.
The only issue I ever run into with the self checkouts is double scanning the product. You do need employee to void items.
Is it a social safety net or stronger legal protections for retailers? In the USA the store has no legal right to detain you and spot check your receipt. Membership stores can revoke your membership for noncompliance and then not let you in anymore without a membership, but that is the closest you’ll find here.
A lot of the waiting is for age restricted items or machine errors rather than trust issues. I have never once been randomly sampled after self surface purchase in either the UK/US.
Without trust issues there would be no machine errors in my experience. Anytime there's an issue it's because I'm scanning too fast for the machine and it can't figure out how the weights work. Or I selected the wrong type of garlic and only the associate is allowed to remove items (why would you not let me remove items I personally added??).
Fair, the inability to remove your own items seems maybe a hang-up from an older view of privileged cashiers. The store I currently use doesn’t have weight checks except for weighed produce. You can leave your stuff in the cart and with the wireless scanner and a bit of forethought everything can stay in there. Different trust models makes things interesting. I have never been checked leaving the store and I don’t think social safety net or otherwise is the driving factor
Yeah, the not being able to remove items grinds my gears.
It's especially annoying when I am waving an item furiously, trying to get the poorly printed barcode to scan, and when it finally accepts it, it double scans the item.
Then when I have to stand there waiting for someone to remove the item is usually where I get disgusted with the process and give up on shopping, especially when it happens multiple times.
The weighing thing seems to be an adjustable feature: in my /extremely/ low-trust Safeway in San Francisco, they started with narrow weight limits that were super-frustrating. Now that misfeature is turned off completely (but they have a little gate you need to wave your receipt at).
I imagine everyone is still playing around to get the optimum ease-of-use vs shoplifting ratio right.
Mandatory item weighting was standard procedure in many western europe self-checkout kiosks for years. The experience sucked and the solution felt over-engineered. They just dropped it recently and moved to a trust-based model overseen by a single employee. Now it's much faster.
I was disappointed when my Waitrose at Kings Cross London got rid of the non weighing machines and replaced them with regular ones. I think the non weighing ones end up costing the store more either through people nicking stuff or having to have staff watch them.
Where I live most stores started off in fort Knox mode but had to tone it down. The bagging area was usually miniscule and trying to fit all the groceries of even a medium grocery list on there without locking the system was nigh impossible
The Whole Foods location nearest to me has self-checkout stations that don't even have a scale under the bagging area. You just scan, put it in your bag and go. I assume they can afford to tank the shrinkage that results from this due to their high profit margins, or they just don't consider the costs associated with the scales to be worth it.
There was a news story a while back about how some chain stores were not going after most shoplifters directly but would keep the recordings on file and press felony charges once a certain total had been exceeded, so instead of getting a slap on the wrist for a few dollars of shrinkage here or there they'd drop the entire weight of the law on shoplifters once those few dollars added up to an amount that qualified for felony theft.
Given that "just walk out" apparently means "replace cashiers with offshore video analysts" I'd wager Whole Foods might do something similar and policing shrinkage in self-checkout is too expensive compared to just keeping track of cheats and throwing the book at them if they make a habit out of it.
That’s how the popular grocery chains in the Netherlands all work. You just scan a few things, put it in your bag, and leave. I almost never interact with anyone.
It uses your rewards card to determine your “risk”, I think. Just a theory. But whenever I’ve gotten a new card, they come check my bag a lot. Then after a dozen successes or so, they stop checking so often. Then hardly at all.
One of my friends forgot to scan an item once when they checked, and he had someone come to check his bag way more often for a while.
And it almost always DOES go wrong. Seriously 75% of the time when I'm done scanning and I press "pay now," it inexplicably says "calling for assistance" and refuses to do anything until an employee comes over. WHY?
And now assholes are intentionally taking alcohol through the self-checkout, deliberately taking up that one employee's time making him or her perform the workaround to check the items out.
And the 20+ years of "unexpected item in the bagging area" bullshit... just get rid of it.
Strangely that issue has almost vasnished for me in the last couple of years, here in New Zealand at least across the several supermarket chains I frequent (although they all seem to use the same self service POS supplier). Sure it was a big issue when these first came into stores but I think they've realised they need to increase the tolerances a bit (e.g. accept even if the weight is <1% off) and it all just works close to 100% of the time for me. I do however generally avoid adding bags to the weight area (as I've found these to trigger false warnings) and just load my shopping directly onto the weighted area and then once I've paid I'll just bag my shopping then. Works like a charm.
I do that, too (scan, pay then bag), but it's a tradeoff - you get smoother checkout experience in exchange for some speed. If everyone was throwing all their scanned stuff in a bag immediately, the checkout would be a lot faster.
What never made sense to me with their go stores was why a store that only needed 1-2 people max to operate had such bad hours. Hearing now that getting the bill is a mainly manual process i guess their hours had to line up with their data entry team in india so people could get their recipes quickly. insane to think about
They didn't even agree with the 700 out of 1000 trips needing review figure
> According to The Information, 700 out of 1,000 Just Walk Out sales required human reviewers as of 2022. This widely missed Amazon’s internal goals of reaching less than 50 reviews per 1,000 sales. Amazon called this characterization inaccurate, and disputes how many purchases require reviews.
Even if the system was fairly accurate, if the vendor is charging Amazon too much for it, it can still be financially worthwhile for Amazon to switch to scanners in their carts.
"inaccurate" can mean lots of things here, from "actually it was 690 out of 1000" to some other minor technicality. Note that Amazon did not provide a figure of its own.
Large corporations tend to tell the truth, but push it as far as they can.
Might be "700 manual reviews per 1000 trips" getting misinterpreted as "700 out of 1000 trips needed manual reviews". If some trips were pathological edge cases that required near-constant reviews and a substantial majority took zero reviews, I could understand why Amazon would keep trying to fix its system.
If my memory serves me right, when I last visited the US right before COVID (SF) I wanted to see an Amazon Go convenience store but it closed down at 16:00 or 17:00.
That was quite unacceptable for a convenience store and it was then I got the feeling Amazon Go was still an experiment rather than a mature technology.
I'm glad it's gone for good, if the process really works like how it's described in the article. Thousands of poor souls doing terrible pointless menial work just so that a few entitled customers can avoid the clunky self checkout (eww- the horror!). The entitlement of the west has no bounds really.
It wasn’t pointless, they were training an AI that never fully worked. It’s ultimately the same kind of thing as people monitoring self driving cars, a boring task that may be pointless or possibly remove a lot of drudgery longer term.
According to The Information, 700 out of 1,000 Just Walk Out sales required human reviewers as of 2022. This widely missed Amazon’s internal goals of reaching less than 50 reviews per 1,000 sales. Amazon called this characterization inaccurate, and disputes how many purchases require reviews.
That sort of thing seems to improve over time. The USPS has had automatic sorting machines for a long time. At first they could only read pre-barcoded mail. Humans had to key in zip codes. Then character recognition got good enough that printed and typed addresses could be read automatically. Some items were still rejected, and they went through manual stations that added a bar-code sticker.
Then manual reading was made remote. There were about 20 USPS remote envelope reading centers at peak. As the vision systems got good enough to read handwriting, then bad handwriting, those were cut back. Now there's only one remote envelope reading center in the US, and what gets there is really bad.
> Thousands of poor souls doing terrible pointless menial work just so that a few entitled customers can avoid the clunky self checkout
Wanting to make your life more convenient and pleasant isn't "entitled".
99% of jobs are things people would rather not be doing (otherwise you wouldn't be getting paid for it). The point is that we can allocate this work in a way that minimizes the amount of time everyone has to spend doing undesirable work.
Are you mad that I sometimes pay "poor souls" to do the "menial work" of cooking me food so I can avoid doing it myself?
> The entitlement of the west has no bounds really.
A very bizarre response to "darn, this was so convenient" - I wonder if this is a troll.
> The point is that we can allocate this work in a way that minimizes the amount of time everyone has to spend doing undesirable work.
1. Not all work is equally undesirable and the way people are paid is not related to that in any way (in fact isn't usually inversely related)
2. Minimising the undesirable work done in total means some people end up doing almost all of it and some basically none (or, if you consider all work undesirable, some people do only the worst and some only the least bad work).
If before, for example, everyone would spend half an hour of their day to cook for themselves, which might be inconvenient but is overall not a big impact on your quality of life, now we have overworked and underpaid restaurant and related staff doing intense work for crazy hours, which is is a devastating hit to their (and their families') quality of life.
The sum of human effort spent on cooking may have gone down in this example, but instead of everyone being a little annoyed by it, some people are living like kings and some are slaving away for their convenience (obviously this wording is exaggerated, but if we look globally, this is basically what's happening).
Sure but you seem to ignore that we don't exist in a socialist society.
Cashiers may have an undesirable job and self-checkout may distribute this undesirable labor among the many instead of burdening a single person with it but those people don't get paid for doing that job whereas the cashier did get paid. Self-checkout is also often slower than checkout at a cashier (it has to be because the cashier is trusted to ring up items correctly whereas the customer has a huge incentive to cheat).
The desirability of being a cashier also isn't inherent to the job. It has more to do with attitudes towards the person doing the job, both from customers and from their employer. Service staff often act as a lightning rod for all frustrations and disdain targeted at their place of work, i.e. they frequently get punished for things completely out of their control. They're also often seen as unskilled labor by their employers so they are treated as easy to replace meaning there is very little incentive to invest in their job satisfaction.
In other words, we have systems that not only require people to work even if the jobs they end up doing are undesirable and undervalued but also actively makes certain jobs undesirable by undervaluing them. If you want to change that, you need to change the system that makes those jobs undesirable, not just do away with the undesirable jobs. People seem to have fewer problems grasping this when talking about the virtues of overseas sweatshops (where changing the system is presumed impossible because the implication is that the system arises from a lack of economic development or cultural inferiority).
This is a great additional perspective, but we seem to be mostly in agreement.
I am quite aware (and rather disappointed) we don't live in a more socialist system, but as we seem to be pretty much stuck under particularly nasty version of capitalism at least for the near future. So unless someone finds a way to turn the whole system around any time soon, I still think at least trying to make some of the worst parts of capitalism somewhat less unbearable is a good thing.
If you don't like going to the grocery store pay someone to go get them for you. Thousands of people remotely "following" you around and scanning things for you, just so that you can avoid using self checkout sounds ridiculous and excessive.
Is it more ridiculous than having someone standing at a cash register in the store doing the same task, while also having to bag for you and pretend to smile?
I don't see where you're getting the impression that 1,000 workers are monitoring every single individual that walks into the store. The figure used was the total amount for the whole program (many locations), only one person was reviewing each case, and they were only even reviewing 70% of cases. Your argument is no different than saying that every time you go to McDonald's, you're entitled for expecting the hundreds of thousands of McDonald's employees globally to band together to make you your burger.
So does having somebody sitting next to the checkout machine watching me (directly or through security camera) while I finish my groceries ?
I fail to see the major difference with somebody working physically in the shop that makes it ridiculous and excessive.
One may even argue that remotely it could opens the job to more people (one could do that from home in their wheelchair while their baby is taking a nap )
There is a difference. In a typical grocery store, people only have to tally what you bought only once at the end during checkout. Where as in this "walk out" system someone has to follow you around "virtually" and keep track of everything you do in the store.
I've adressed that point with the security camera (or adequatly positionned mirror in old-fashion store)
And the workers just sit idle when there is no customer and you have to physically go there. So I dont see the clear advantage, i see it more as a "choose your poison"
An important question to ask is how many minutes of camera-viewing there were per customer. Let's not assume too hard before we judge the level of wastefulness.
The point of the store was to provide the appearance of conventional retail while being exclusive to "members". This would theoretically cut down on shoplifting and boost profits at the expense of shutting out marginalized people.
My first thought is that this was an attempt at satire, but now I realize it's a privileged Western person who has never traveled to a poor country and thinks that if these workers weren't labeling data they'd be self-actualized DJs or crypto investors.
I think your first instinct is the correct one. But I don't buy the argument that these jobs are somehow empowering them.
My criticism is towards the entitlement of shoppers who think self checkout is too clunky and having to interact with a cashier is too annoying but expect an army of people to work behind the scenes (be it anywhere in the world) to make it easy for them.
"Amazon Go stores will still use Just Walk Out technology, and the company will continue to license it to other retailers. Smaller stores in the UK also will keep using the system."
Compared to simply walking out, packing everything into your bags right away without needing to scan with either a hand terminal or "repacking" at self checkout is a lot more friction. Didn't really think it would be until I tried a couple of times in a row and it was incredible.
> The painful clunkiness of self-checkout was gone.
It could be worse. Imagine a smart gate that refuses to open for wheelchair users, or claims the child in your arms is an unpaid item - something that is getting rolled out in many Australian supermarkets.
Shopping there always felt supremely weird to me. Scanning in, getting stuff, and walking out, but I always felt incomplete without a receipt being right there in my email, wondering if I'd done something wrong and had just inadvertently shoplifted.
I shop more often at Walmart, which has recently increased the number of manned checkout lanes and restricted their self-checkout to 15 items or less.
The local Walmart expanded self-checkout and it works surprisingly well (the bag scale seems to be very loosely calibrated or off). I wonder if they're doing things differently depending on how much product walks out the door.
Just in the past month or two, in my (low-crime suburban) area, Walmart appears to have closed the two large self-checkout areas entirely and replaced them with a (relative to before) army of cashiers. I have always found their machines to be very hassle-free, but the shockingly-adequate level of cashier staffing made my last visit surprisingly quick and convenient. I could live with this.
> Scan as you shop is a big step backwards and feels like you've got the annoying self-checkout experience looming over you the entire time you're there.
How do the Amazon "Dash Carts" actually work though? If it were a traditional bar code scanner, I'd agree with you. But reading up on the tech makes it sound like you just place the item in your cart and that's it. If it worked like that I don't see at all what the problem would be.
But this was an obvious sham and a fraud that couldn't scale. I don't think its demise is sad at all.
What's sad is putting cashiers out of work, and even worse is replacing them with outrageously piss-poor alternatives. Self-checkout is a clinic on incompetent system design, and has been for DECADES now. It's mind-bogglingly bad, all to take jobs away from people. Fuck that.
Why is it a fraud? Did they ever tell you it was entirely computer-driven? I thought their value proposition was you could just ... walk out. This feels pretty inescapably the future. Why have people do bad jobs like cashier work when computers could do them instead? This is why UBI is going to be critical. We're moving to a world where we'll have more people than work, and that's not just ok - it's fantastic.
They're not wrong, they're just early. Which you could argue is the same thing on a micro scale, but not on a macro.
> Our checkout-free shopping experience is made possible by the same types of technologies used in self-driving cars: computer vision, sensor fusion, and deep learning. Our Just Walk Out Technology automatically detects when products are taken from or returned to the shelves and keeps track of them in a virtual cart. When you're done shopping, you can just leave the store. A little later, your receipt will be available and we will charge your payment method.
there's no way they were turning a profit. they definitely missed stuff all the time even with a ton of sensors. and sensors aren't the only cost. annotation is by far the most costly operational expense. new product? needs several annotated photos and recalibrated weight sensors. merchant decides to put Christmas branding on the same UPC? now all your vision models are poisoned for that product. it needs to be re-annotated for the month and a half it exists and the models need to be swapped out once inventory changes over again. as long as merchants are redesigning products (always) your datasets will be in a constant state of decay. even if your vision sensors are stationary and know the modular design up front, you still need to be able to somewhat generalize in case things get misplaced (big problem for weight sensors) or the camera gets bumped.
between dataset management, technology costs, research costs, rote operational costs, etc this is a very expensive problem to solve. and large models with a ton of parameters are little help; they may lower annotation costs a bit but will increase the cost of compute.
once i really dug into this problem i saw Amazon Go's Just Walk Out for what it really was: a marketing stunt
the biggest cost is not annotators at the scale you're imagining. it is labor costs.
Amazon bet that the federal govt would raise labor costs to $20/hr and all their competitors (besides themselves with this tech) would get wiped out. They even publicly campaigned and lobbied. That didn't come to fruition as the election promises turned to fluff, and the populists simply chose to empower unions instead.
i mean, labor cost (as in in-store labor) is the target for this cost optimization. unfortunately for the time being labor cost is not as significant as the other costs associated with annotation and dataset curation. technology costs are not really significant if this can be pulled off at scale.
in-store employees know where things are supposed to be and why, if at all, items are "misplaced" according to the modular design
I think this failing, and tesla failing to actually ship "self driving" is clear : machine learning definitely has complexity limits, that we are a long way off from perfecting or even getting beyond some reasonable threshold.
> The company’s senior vice president of grocery stores says they’re moving away from Just Walk Out, which relied on cameras and sensors to track what people were leaving the store with. [emphasis mine]
> Though it seemed completely automated, Just Walk Out relied on more than 1,000 people in India watching and labeling videos to ensure accurate checkouts.
> According to The Information, 700 out of 1,000 Just Walk Out sales required human reviewers as of 2022. This widely missed Amazon’s internal goals of reaching less than 50 reviews per 1,000 sales. Amazon called this characterization inaccurate, and disputes how many purchases require reviews.
> “The primary role of our Machine Learning data associates is to annotate video images, which is necessary for continuously improving the underlying machine learning model powering,” said an Amazon spokesperson to Gizmodo. However, the spokesperson acknowledged these associates validate “a small minority” of shopping visits when AI can’t determine a purchase.
The article is kind of all over the place, but it sounds like there were lots of sensors and also lots of human intervention.
How well does simulated data work in this space? My first stab at doing this scalably would be as follows: given a new product, physically obtain a single instance of the product (or ideally a 3d model, but seems like a big ask from manufacturers at this stage), capture images of it from every conceivable angle and a variety of lighting conditions (seems like you could automate this data capture pretty well with a robotic arm to rotate the object and some kind of lighting rig), get an instance mask for each image (using either human annotator or a 3d reconstruction method or a FG-BG segmentation model), paste those instances on random background images (e.g. from any large image dataset), add distractor objects and other augmentations, and finally train a model on the resulting dataset. Helps that many grocery items are relatively rigid (boxes, bottles, etc). I guess this would only work for e.g. boxes and bottles, which always look the same, you'd need a lot more variety for things like fruit and veg that are non rigid and have a lot of variety in their appearance, and we'd need to take into account changing packaging as well.
as mentioned in another comment, "scale" is not just horizontal, it's vertical as well. with millions of products (UPCs) across different visual tolerances it's hard to generalize. your annotation method is indeed more efficient than a multistep "go take a bunch of pictures and upload them to our severs for annotators" but is still costly in terms of stakeholder buy-in, R&D, hardware costs, and indeed labor. if you can scope your verticals such that you only have, say, 1000 products the problem become feasible, but once you start to scale to an actual grocery store or bodega with ever-shifting visual data requirements the problem doesn't scale well. add in the detail that every store moves merchandise at different rates or has localized merchandise then the problem becomes even more complex.
the simulated data also becomes an issue of cost. we have to produce a realistic (at least according to the model) digital twin that doesn't interfere too much with real data, and measuring that difference is important when you're measuring the difference between Lay's and Lay's Low Sodium.
i'm not saying it's unsolvable. it's just a difficult problem
I want to know why these places don't simply dramatically drop the accuracy requirement.
Rather than giving itemized receipts, give just a total dollar value. Then just make sure that most customers are charged within 10% of the correct amount.
Only if the customer requests and itemized receipt, then go watch the video and generate them one. But after a while most customers won't request it, and that means you can just guess at a dollar amount and as long as you're close-ish (which should be easy based on weight and past purchase history), that's fine.
The problem is taxes. Different items can be taxed differently. Itemized receipts are a must from the get-go unless you want to explain to the government that you underpaid them because you charge within a 10% error-margin.
Come on, it isn’t anything to do with being a “marketing stunt.” Often products like this are expected to lose money at first, but they hope with enough R&D and scale that they can make it successful eventually.
For example, you are pointing out that annotating is costly, but that’s an expense that scales independently of the number of stores. So with enough scale it wouldn’t be as big a deal. Or if they figured out some R&D that could improve it too.
right, that's how it starts. but the improvements in methodology simply aren't there as the ML sector has been laser focused on generality in modeling (GenAI as it's affectionately known). "at scale" doesn't just mean more stores; it means more products and thus more annotation. how many UPCs do you figure there are in a given Target or Whole Foods? i assure you it's in the millions.
one advantages of the Amazon Go initiative is a smaller scope of products.
a big problem in the space is that products that look very similar will be clustered in the same section. large models are very good at generalizing, so they may be more attuned to "this is a Christmas thing", but they won't know that it should be classified as the same UPC as the thing that was in that spot yesterday without you specifically telling it to. how would it know it's not a misplaced product or a random piece of trash? (you won't believe the things you find on store shelves once you start looking) you can definitely speed up your annotation time with something like SAM[1], but it will never know without training or context that it's the same product but Christmas (ie it resolves to the same UPC).
Uniqlo (in Japan and at least in SF) has a cool checkout method.
You drop your clothes into a big bin (I’ve always done it one piece at a time, didn’t think to do all at once) and it adds up all your items. I’ve used this maybe 8 times, 100% accurate so far.
Not sure how it works but I guessed RFID and quick Google appears to confirm:
Self checkout systems based on RFID are very convenient, quick and quite accurate compared to self checkout on grocery stores'l. The main problem on groceries is that the tags are expensive compared to the product (tags prices are in the order of tens of cents depending on manufacturer and size)
Disclaimer: I work on a RFID reader manufacturing company
It’s not very expensive to being with, but it’s magnitudes more expensive than a print tag. Print is either free (a package being printed anyways) or extremely cheap (sticker).
I don't want rfid tags in my apples, and putting them on the outside of each one seems costly.
And things like onions have a loose outer skin, which falls off. And grocery store clerks peel the outside (wilting) leaves off of things like local lettuce daily.
Seems like rfid on groceries would be hard for fresh produce, and we don't need everything wrapped in plastic.
You won't have it anyway. An rfid tag tag placed on any conductor (apples are conductors because of their water content) 'detunes' the tag and makes it impossible to read
Some of my super smart colleagues solved real time inventory with big data within a small enough margin of error that Walmart scrapped the shelf scanning robots that were being deployed.
EAS is a purely psychological deterrent, for liability reasons stores can't really do much when it goes off. As more and more shrink is the result of ORC, RFID will have a place but money is better spent on alternative technologies like facial recognition and ALPR.
If we had a better Blue Bin system it might work. Remove tags at home and place in compartment of Blue Bin, and at the recycling centre they get sorted out and sent back to the retailers.
But if we really could scale RFID tag production down to 1 cent each, then we’d likely just throw them out. Not that I’m a fan of throwing away silicon…
Orders of magnitude more than existing barcodes which are practically free due to being printed on the packaging. A 10c RFID tag would be 10% of the cost of many products, which im sure is unacceptable to grocery stores which I can't imagine have great margins.
I obviously didn’t mean all employees ,only the employees who are doing the check out .
I am not saying RFID will eliminate employees in retail , just that there is cost to not having RFID too, we should only see the diff as the cost of the tech not absolute numbers
I understand, and what I am saying is that it is rare to have retail employees that _only_ do checkout. Nearly everyone performs multiple roles. Eliminating checkout will reduce the overall workload, but won't necessarily reduce headcount, or enough headcount to offset the cost of RFID tags.
> it is rare to have retail employees that _only_ do checkout
Why is that relevant?
If 5 employees spend 20% of their combined time on checkout, and you get rid of checkout, you can go down to 4 employees, can't you? Why does the exact distribution matter? You'll still have a healthy variety of tasks to spread among those employees.
> A 10c RFID tag would be 10% of the cost of many products, which im sure is unacceptable to grocery stores which I can't imagine have great margins.
If the store decides to eat the cost of those? Sure, I can buy that logic.
However, I don't see stores being unable to just raise the price of items by 10c without much pushback, which would eliminate that issue with the profit margin entirely (since the cost of it is offloaded to the customer).
I, personally, would love to be able to pay just 10c extra per item, if it meant that I can have a checkout experience similar to that of Uniqlo.
The tech in store is powered by RADAR (https://goradar.com) - next gen RFID sensor that can work with lower cost antennas. Still unlikely to ever get to unit economics that work for groceries, but really exciting what this team has created.
As far as I can tell, every Uniqlo has this. And you’re right, this is by far the best checkout experience. It’s what Amazon wishes they’d built.
I can’t imagine there’s that much untapped profit in the grocery business that Amazon could turn a profit with such an expensive and unreliable mess like Just Walk Out and I find it’s so typical Amazon to find out that the man behind the curtain is actually a bunch of offshore workers in India.
Decathlon does it but it's a little harder to implement in a grocery store.
You need to basically have custom packaging so the tags are not easily swappable and the RFID tag is not free too, it could really add up for small price items.
The maybe e-waste problem?
I'm not really familiar with how much of a problem this really is but on a grocery store size it could add up
Uniqlo checkout experience is really how self-checkout should be look like, but unfortunately relies on controlling the manufacturing of all products to implement efficiently.
The closest thing I've seen in a convenience store, was a fully automated 7-Eleven that I ran across in Taipei last year. It seems to rely on smart shelves[1], rather than just product video recognition. This could work for most types of groceries.
This exists in most UNIQLOs it's the same mechanism through which the alarm goes off when someone attempts to steal things with the magnetic RFID attached.
It's borderline ancient tech. So why is it not applied everywhere? Because stores would rather have you line up and buy stuff from the "impulse aisle" and tip or donate at their PoS (higher rates with an employee glaring at your selection)
RFID tags are expensive compared to printing barcodes. The volume of products moved in a grocery store is far larger than that of a clothing store like Uniqlo.
I've had a few times when I bought lots of small items and it missed a few of them. Once, they were recognized when I moved stuff in the bin, another time I made sure I didn't move anything before paying and it worked in my favor.
> Uniqlo (in Japan and at least in SF) has a cool checkout method.
Saw this in the Honolulu location and was similarly impressed. It's got to be pretty close to a theoretical minimum in check out speed, while requiring no labor and no accounts.
Some thrift stores let you buy clothes by the pound here in the US. There are always the same people picking out stuff I swear. Probably picking stuff to resell.
"Ginger Market" on the SJSU campus has attempted to use a similar "Just Walk Out" approach. It has not worked well. I've been double charged, charged for items I did not take, not charged for items I've taken, etc. The refund process was also a pain. It was so bad that they had to stop using it last spring, although they claim they are going to give it another go.
Why even bother? Self checkout or a cashier work so much better and I have a hard time believing they are less expensive. The store is plastered with cameras. Seriously, there must be 100 cameras in the place. That's a lot of video to process, which has got to be costly, whether it is a machine or people who are reviewing it.
There was a startup, Standard Cognition, that offered the same experience, but I checked their website (https://standard.ai/) and it seems they have given up on it too.
Edit:
Looking over their marketing videos now and taking a more optimistic approach, is "just walk out" technology all that useful? It seems they pivoted to a product where is much clearer what the value add is (Predictive analytics, loss prevention, context-aware marketing). I imagine "just walk out" technology was likely pretty expensive to implement, but wouldn't have saved much more that self checkout.
Maybe the lesson here isn't that "it didn't work", and more so "it wasn't economically efficient"
Well from the article it's clear that this camera+ai based detection of purchases never worked for Amazon. They had to rely on Indian contractors watching people remotely. It never technically worked, and even if it did, then yeah I agree that it wouldn't make economic sense. Cameras with that level of fidelity and with 100% coverage, tracking N customers at once, are probably a huge capital expense. And all for slightly faster shopping.
I tried the store in Seattle in 2021 and it was a shitty experience. Overpriced, bad stock, and since few are going to actually trust Amazon to get it right, you still find yourself with the Amazon app open the whole time
Wanted to comment on this part -- Prime itself was nothing more than faster shipping, though yes the difference between 4-6 weeks standard vs 2-days was massive (not just slightly faster). But the point is, Amazon excels at identifying friction points that others have just accepted as industry norms, but which, if unblocked, could actually meaningfully shift consumer behaviors.
Go might be a failed experiment, but "slightly faster shopping" is probably an unfair trivialization of what the experiment hypothesis was really about. A core thesis of Amazon in general is basically, to fanatically remove any unnecessary extra steps/friction/bureaucracy/etc., between a consumer and their act of purchasing.
For another example, think back to why Amazon cared so much about the 1-click patent -- legal validity aside, the idea that you can have 1-click checkout, was pretty revolutionarily customer-obsessed, compared to the average online shopping experience of the early 2000s.
And in fact, the natural progression from 1-click, is of course going down to "0 clicks", which is what things like, subscribe-and-save, memberships/subscriptions, or Alexa/Echo styling clothes for you, are meant to do -- they are meant to shift the consumer's mental model of shopping away from emphasizing the build-up from browsing into the climax moment of then clicking to purchase, to instead make the actual purchase decision more of a hidden-in-the-background/automatic thing, instead of a foreground conscious choice.
The issue is that speed is not the only thing necessary. You need speed, and trust.
A cashier-less, grab and go store that works as advertised would be hugely convenient and valuable. But where we are now, people don't think it works, so they double check and get nervous.
And that goes doubly for Amazon. Amazon doesn't care about doing things right. They've shown that over and over via their (lack of) protections against counterfeit items.
Ask anyone in the eastern US, would you trust getting your eclipse glasses on Amazon?
Why would you trust them to get your grocery order right?
> But where we are now, people don't think it works, so they double check and get nervous.
Who does this? This wasn't in the article, do you mean yourself or did you read that somewhere else? I wouldn't get nervous about it.
> Ask anyone in the eastern US, would you trust getting your eclipse glasses on Amazon?
That's a real issue but it's an entirely different concept. I expect them to handle charging for the eclipse glasses just fine. Similarly, I'm not very worried about their checkout gimmick.
I don't interpret that skit as worrying about whether it's going to get the count wrong. It's about the entire concept of grabbing and walking out, in a way that exists even if the system has perfect tracking.
> But the point is, Amazon excels at identifying friction points that others have just accepted as industry norms, but which, if unblocked, could actually meaningfully shift consumer behaviors
I would pay 5% more for my groceries if I never had to wait in the checkout line. I’m a kind of customer that has tried grocery delivery, and didn’t reject it because of the cost, but instead for other reasons like I want to pick out my own food, I don’t want it left sitting when they deliver it, I can get it faster myself etc. I would definitely pay money to make grocery shopping faster and more convenient. I suspect there are a lot like me.
I recently finished a book called "The Secret Life of Groceries" by Benjamin Lorr. Amazing book. I learned that grocery and supermarket stores of the past, seemed to have worked very hard to give the best customer experience possible. Today it feels that customer experience has been replaced by cost cutting.
Self checkout has been rife with problems, internally and socially, but supermarkets keep pushing them. I can only conclude it is cost cutting.
I hate it, especially for large trips to the store.
The scale freaks out too often after I scan an item and put it on the bed, because the weight is slightly off for whatever reason. Then I have to stop scanning and wait for an employee to come over and scan their badge to authorize the sale.
I've gone back to just waiting in line at the standard checkout unless I have < 5 items in my cart.
At least where I live it's a nightmare. Instead of having someone work as a cashier and weight my veggies I now have to weight my veggies myself under the supervision of a worker who's only job is to supervise a machine that is buggy enough to need human intervention for every other client.
It only serves the shop owner because he can have a single drone operating a few self checkout machines instead of a single line, but is that a net positive for society ? doubt, going to a supermarket is alienating enough, we didn't have to make it worse.
And of course they had to add a gate so now you can't even leave the shop without showing the receipt, when you don't find what you want to buy you have to call and wait for an adult to allow you to leave the premise like you're a fucking toddler getting out of the playground area.
I've never experienced any problems with the ones that don't weigh anything. The weighing ones get it wrong often enough to be annoying. If my closest grocery store used weighing self checkouts, I'd be much more annoyed on a regular basis.
Home depot is so overengineered. They are using 24” high res touchscreen pcs with an i7 to just run self checkout. Whoever locked in that contract must be pretty pleased.
A couple hundred dollars is nothing compared to the costs surrounding that PC. Compared to how often systems like that get underbuilt and laggy, using an i7 is a good choice.
Yeah but its self checkout. Self checkout had no lag when it came out like 15 years ago. What did the engineers do to their codebase to require a modern i7, I wonder?
Every other checkout I’ve used has lag time after the scan and until the scale settles. Some even have visual feedback that indicates a busy status during this repeating part of the checkout process. Not to mention the human lag time of moving the product to the scale.
There is no scale on the Home Depot units. You can scan as fast as the scan gun accepts scans. I’ve scanned 2-3 items per second with these.
Using an I7 is a move to future proof. These units will likely be in service 10, maybe 15 years from now.
I am sure they are on a three year replacement cycle like every other piece of enterprise compute. Home depot is also not the only place to do fast scanning. Grocers have always been able to scan instantly as long as I’ve been alive at least.
This surprises me. I regularly get in line and fully complete self-checkout before anyone who was there when I started scanning, even though I'm typically checking out with more items. What's causing the struggle? Every one of these works the same at every store - scan barcode, place item on the weight plate. Repeat. Press pay button, pay with card, whatever, tap/insert.
> scan barcode, place item on the weight plate. Repeat. Press pay button, pay with card, whatever, tap/insert.
scan barcode, place item on plate mostly works except when barcode is crumpled or something.
without barcodes it's a more of a crapshoot. Usually more than 1/2 of my basket doesn't have a barcode (fresh produce, etc.)
Maybe the ones near me are just too sensitive to weight variations.
Oh and one of them really loves to scan a points card 2x, then try an apply the second read as a product and fail.
Every time the machine gets confused, you have to wait for an attendant. All things being equal, the cashiers are faster than self scan. If the lineup for cashiers is longer, it can still be faster, but otherwise it's slower in my experience.
Well, have you tried the one at Amazon Go stores? I've been using those once every month or so, since the year the very first one opened in Seattle, and not once have I had any issues at all.
It was truly fantastic, especially since the app would notify me afterwards how long I spent at the store, and seeing that number always made me feel good. TLDR on this: my average shopping experience from the moment I entered the store to the moment I left took around 3-4 minutes, with plenty of them being under a minute long.
> Today it feels that customer experience has been replaced by cost cutting
This seems to be true of all businesses, and I really hate it. Flying sucks, trying to reach a human for any reason sucks, lines are huge everywhere. Life just seems a lot more stressful purely to put a few extra bucks in shareholders' pockets.
One of the various jobs I had during the 90s to put myself through college was stocking store shelves in a large grocery store. During my time there, I placed every item on the shelf by hand, rotated forward, so that the shelves were easy to front and look respectable by the time the morning came. After a few months, it was fairly easy to eyeball an item in the case, lift it while rotating, and place it on the shelf facing forward. It also became reflex to hear something fall and catch it without looking. We also mopped the entire store by hand, twice, before getting down to the business of fronting the shelves one final time.
As time went on, they hired a cleaning crew to clean the floors with a machine. They'd miss things. After my time, they started having their employees cut boxes and just plop entire boxes onto store shelves, ripping the fronts off of them. I'm sure it saves a little bit of time, but you end up with a store that is a complete mess full of discarded cardboard. All of that adds up to a terrible customer experience.
There is a possibly unexpected advantage to stocking the shelves with boxed items: if the customer wants to purchase items in bulk, they can retrieve the boxes just as easily as they were put there. As a customer, I do that a lot for unperishable goods.
Well, the stock issue is sort of a manpower issue. For example, back home, the shelf with canned green beans, corn, and pre-cooked red beans was 4 cans high, and each section of those items was about 12 cans across. It'd be nice if we could simply spend the time to rotate and stock the items, but that just isn't feasible. It's far easier and more cost effective for the store to periodically check for expired product than it is to rotate them. It's even harder with tiny items like seasoning bottles. It would be nice if someone could design a shelving system that could be moved and stocked from the back, but that'd be quite an engineering feat given the sheer weight of the shelves and the proclivity for items to fall over when they're moved.
All of these stores have budgeted in losses. My managers were almost fired for the sheer amount of negligence they gave the store when it came to ordering product. Each location has a certain cadence to them when it comes to what people are buying. My managers didn't understand it, so they ended up with 25 rolling carts packed solid with backstock (these are 8 feet tall and have two shelves), and another 20-30 pallets. If you're unfamiliar, for a normal sized grocery store, you should really only have backstock on the top sides of those carts. Instead of filling the shelves with the product they'd already bought for the store, they were buying entirely new product! I managed to wrangle the order gun from them for the grocery department and they went from losing $30K a quarter in inventory to gaining about $15K. I kept an inventory in my head of what we had in the back and made sure not to order it.
How? Because when pallets inevitably tumble in the distribution centers, the centers can't waste time sorting through inventory to determine what's damaged. So they count it as a loss and toss the product into crates so that store employees can sort through them. Good stuff goes on the shelf, damaged stuff goes in the trash. Apparently the distribution centers damaged a lot of product, lol. My managers got cash bonuses for turning things around, and I got a 6 pack of beer.
> Today it feels that customer experience has been replaced by cost cutting.
I feel like it's the case for literally every single industry, everything is getting worse for the end user point of view. Public services, jobs, hospitals, cars, new houses, online services, mobile apps...
I think globalisation allowed things to get better and cheaper up to a point and now that we can't delocalize to cheaper places we have to cut corners somewhere else and the entire stack is suffering from it
When the savings are not passed on to the customer (and they usually are not, as in this case) it's more accurate and descriptive to call it "profit maximizing."
I think that the scan-as-you-shop style systems are definitely the next realistic step for grocery stores. I really like using them, as they allow me to put stuff directly into my bags as I go.
I think a further improvement on a system like that would be to use the cheap RFID tech (UHF EPC) so that when you walk through the anti-theft barriers, the system knows exactly what you got, and then there's no need to scan anything.
> I think that the scan-as-you-shop style systems are definitely the next realistic step for grocery stores.
Scan-as-you-shop has been around for at least a decade at this point I think, at least in some parts of the world. ICA (supermarket in Sweden) been doing it for as long as I can remember, and I came across this image of the scanner being used at the store on Mediawiki: https://commons.wikimedia.org/wiki/File:Self_checkout_scanne... (Image uploaded 2011)
Surely this exists elsewhere too, or been judged to only work in certain contexts (like a high-trust environment like Sweden) and won't be the next realistic step.
Jup it has been available for a very long time in the Netherlands as well.
When I moved to the UK, what surprised me is that Tesco self check out is so much more cumbersome. It weighs your stuff, stops working when something isn't exactly the right weight. Super annoying to use if you're used to the system used in Sweden and The Netherlands. But that was in stores in the middle of central London, it may be very different in more suburban "big" stores which is where it was first rolled out in NL as well.
I'm not a fan of the data tracking that I know those loyalty programs are for, but at least in the US most stores overcharge you if you don't use them (they present it as giving a "discount" if you use them but really you are just getting the fair price denied to non-members) so you would need to have deep pockets to resist them out of principle.
The latest evolution in these are really devious: A shelf tag advertises a price in large print: $7.99 with smaller print "with digital coupon" And then underneath, in the color usually used to show the "non-sale price," you see "$12.99." A lot of people just glance at the tag which looks a lot like the normal 'sale price' tag and thinks they're paying $8. Of course, you have to get in the sluggish, stupid app, and best case scenario scan the tag or the item and get to the right part of the app to "Clip" the coupon. Assuming you can get a good enough signal in that part of the store, and that you haven't become logged out of the app... etc. etc. And assuming you don't make a mental note to clip it in a minute when you get out of this crowded aisle and then forget.
I think they just really want to display one price and charge a much higher one to most people.
In my part of the US, anyway, there are several grocery stores that don't use loyalty cards at all, so you can avoid the extortion by shopping at those.
apps are a new thing. before, when it was just a phone number, 867-5309 in the us would usually work to get the discount while being usefully trackable
You maybe haven't actually spent any time in Sweden and compared to how people act towards each other there compared to other countries? Or, you haven't spent enough time in other countries than Sweden, making it hard to compare.
I lived there for my first 20 years, and have since lived in Spain. The amount of trust between people in the Nordics is definitely higher than most other places in the world.
Yeah, I remember it as a kid. But it surprised me when they brought it back because people rarely used it. I still don't really get the point of it tbh, scanning items as you pick them up just spreads out the faff of scanning things and slows people down in the aisles.
Yes. In Switzerland Migros you have this as well, at least in the bigger ones. But they also added scanning with the phone. So you don't need a special scanner, just your phone.
One of the "perks" of my Walmart+ membership is scan-as-you-go. I think I've used it twice. Both times, I only had a few things and the store was packed solid with spring breakers. It was a decent experience but I still like the normal self-checkout the best.
So does Walmart+ (the membership program which also gives a better deal for home delivery). I am curious whether this is going to stay or not, now that in some Walmarts the self-checkouts (where you're still obliged to stop and scan a code to complete checkout) aren't open normally.
I see a couple people using it at the shops near me in the UK (but it's not that popular here either). It's pretty universally available in the bigger shops, but rare in the little metro shops. I really like just slinging my bag in the trolley and running around the store with a scanner. It's dramatically easier than unpacking a trolley for self-checkout or the old school checkouts. It's annoying when there are security tagged items though, because it defeats the purpose of the self scanning.
> We've had that in France for years, and noone uses that as far as I can tell.
Which part of France are you in? Where I am, Paris suburbs, I see it used all the time in Carrefour.
There is a learning curve, which I think inhibits wider usage. I used it a couple of times, and every time I forgot to scan stuff, or forgot whether I scanned stuff, and had to go over the last scans to check.
Yes, same in the UK. Most stores have scan as you shop, but most people do self checkout. I get the impression scan as you shop is seen as being a bit prissy.
As a German it is always a culture shock to see how popular self checkout is in the UK, even though I see it used more often recently in Germany too. I think Germany is about a decade behind...
There's a simple formula that makes you chose it: they keep as few manned checkouts open as possible, as long as the wait for the manned checkouts is inconvenient enough the majority of people will use the self-checkouts.
Wait until you see how popular it is in the US! I can't remember the last time I saw someone use a manned checkout line; my local grocery store has 1 manned lane and 12 self checkout lanes, and the manned one is almost always empty.
I actually prefer manned check-out. It tends to be much faster. The checkout people tend to be wicked fast at swiping groceries in front of those scanners, much faster than I can dig into the cart, find something, fumble around scanning it, then place it in that little post-scan cubbie---ughhhhh. And manned checkout is already parallelized: Next customer queues up groceries while current customer is being serviced.
> 1 manned lane and 12 self checkout lanes, and the manned one is almost always empty
its empty because it’s considered rude to burden the poorly paid checkout person when you can self-service. Once all of the self-service lanes are occupied, the sentiment shifts and customers uneasily queue up in the manned lane, frantically watching if the self-service lane clears up so they can jump out of the manned lane.
One of those freak cases in behavioral econ where the causality is very straightforward and explicit.
I've certainly observed what looks like that kind of behavior, but I don't personally feel rude at all for using the services of a cashier.
In my own shopping, I just use whatever seems likely to be fastest or less hassle -- for me.
If I've got a bunch of stuff, I'm heading to the cashier because they're better-equipped to handle a volume of stuff than I am at self-checkout.
But if it's just a couple of small items, then the self-checkout seems fastest: Scan, plonk, scan, plonk, invoke the incantation so that it can take my money[1], and pay it.
(Unless one of those small items involves something like beer or something else requiring an ID check, wherein: It's back to the cashier.)
[1]: In my neck of the woods, Wal-Mart gets this best, with as few as one button-pushes required to pay and leave (and there was a time when it was zero button pushes to use a debit card at self-checkout there). Dollar General gets it worst, requiring at least 8 button pushes (with three different input methods! it requires input on two different touchscreens and one physical keypad) to pay them and get on my way.
It's also a step towards individualized pricing, which many people, myself included, absolutely hate. This can be done with self-checkouts but it'll work much better with handheld scanners. Normal price: 3x, member price:2x, individualized price: scan to find out.
Big chains plan to gamblify the prices in the next few years using their "member cards" points, some chains are already up to this with specific penetration targets before they move forward. And this is not insider knowledge they're upfront about it.
We have it in Germany for years too (in densely populated areas) and my subjective impression is that next to none was using it for a long time but that it's picking up slowly.
For one thing I think these things take years for broad acceptance and for another the current scanners with their bright and large displays are just what was needed to make it attractive for the young and elderly.
Give a retailer a large display, and they will fill it with ads, with the important info in tiny type. Or even a small display. I've seen user-facing credit card terminals where the small screen is 3/4 ads, 1/4 transaction.
If they'd just put up "Place card against screen here" marker that was 1) where the antenna is, and 2) fully synchronized with when the system is ready for a card read...
Similar in my area (in the US). We've had scan-and-go at the local grocery/everything store for many years, but I really don't see that many people actually using it.
Because the trust level is under zero. In France you have to put items one by one and the weight of your shopping bag has to add up before you can scan the next article. It's the prison feeling with the speed of an 90 year old in any other country's self checkout. 0/10 would not recommend.
Yeah that's basically the same issue. I do think the self scanning is even worse than the self checkout because at least at the self checkout you scan everything once at the end when you're done.
The fact that you have to remember to scan every time you put anything in your basket is just worse for the mental load I find.
If manned checkout is a dishwasher and self-checkout is handwashing dishes in the sink after the meal; scan-as-you-shop is eating next to a little basin where you wash each utensil as soon as you're done using it, before you're allowed to return to the meal.
The 'little basin' is apt because the quality/usefulness of equipment is a huge part of why it always sucks. I used a scan/pay/go system in what seemed to be a trial rollout at a major supermarket. They used your choice of a handheld scanner that lagged tremendously compared to a real scanner gun, or a smartphone app (which of course is laggy AND clunky, since you have to keep unlocking your phone and tapping some button, and waiting for it to register)... I pretty much hated it. I only used it while COVID was a big deal. They ended up pulling it and not rolling it out.
I don’t know if it’s really unpopular, but it’s only used at self checkout in Decathlon which is a chain of supermarkets selling sports equipment and clothes.
> I think a further improvement on a system like that would be to use the cheap RFID tech (UHF EPC) so that when you walk through the anti-theft barriers, the system knows exactly what you got, and then there's no need to scan anything.
Isn't that basically what Amazon had claimed they were doing here, except apparently maybe they weren't?
It always seemed so needlessly complex, relative to well-established RFID tech -- Uniqlo can count the number and type of garments in a pile in a big bucket, and compute an invoice from that.
Even if you couldn't do this at the exit, seems like it would have been a far easier lift to incorporate the same idea into shopping carts or baskets all along.
The cost-per-unit on RFID product tags (EPCs) has tended to limit them to products with a relatively high margin and a relatively high theft potential---clothing being the most obvious example and the most common application of EPCs, with retailers all the way down to WalMart using them for apparel.
You'll note that WalMart doesn't even use the EPCs at POS, which is telling: for most retailers, the main advantage of EPCs is far more actionable alarms at the exit. So they're limited to items where loss rates make the added cost worth it.
The problem is that the grocery industry has notoriously low margins, and the unit price of EPC tags can be the entire margin on a lot of products. On the one hand, Amazon may have been trying to work around the need for higher-cost tags to roll out this kind of automation. On the other hand, I have heard anecdotally that Amazon Fresh pricing was relatively high, so maybe EPCs would have been a wiser use of their extra revenue.
Amazon Fresh was definitely not the cheapest option. It was on par with Whole Foods for groceries, IME.
But that said, interesting point. Didn't know the cost of an RFID tag was that high -- after all, Uniqlo is putting them on some pretty cheap clothing items!
Rfid tags are really cheap, some sellers advertise prices as low as 3c per tag but realistically 10c is probably the cost of a finished product (print, glue, maybe extra protection of the plastic)
That's the problem though - $0.05 is a LOT to a grocer, that's all they keep on a lot of products. Barcodes are free since they mostly go on labels that are being printed anyway. They could push for source tagging, but the vendor would pass on the cost, and grocery is a very price sensitive industry.
For apparel, on the other hand, source tagging is common - even before EPC on higher end goods, Calvin Klein used to sew magnetostriction tags into clothing. Apparel just has so much more price elasticity and loss prevention is a huge part of that industry. Tools are another industry where EPC and source tagging are common, once again, high dollar items with a lot of theft.
Albert Heijn does this in the Netherlands, tied to your loyalty card. You can even import your shopping list and show it on the little hand scanners. At the end, you scan your loyalty card and it shows up on the self checkout.
I think it’s great, I get to go around the shop with a little laser gun.
An even nicer thing about the Albert Heijn self checkout is that I can use my own phone from start to finish. Connect to store wifi, scan items using their app (using the phone camera), pay with my phone (contactless payments) on the self checkout and use my phone to scan an exit barcode at the turnstile. My visits to the store usually last only a few minutes and I don't mind popping in multiple times a week.
Yes it's great. Put everything straight in your bag, no packing at the end.
The only thing I don't like is that they have added the option to the machines to scan all your things there instead of using the portable scanner. Which in my area has resulted in queues at the self checkout... Which used to be very quick when it was only used for the portable scanners.
Scanning with your phone camera sucks its very slow and scanning as you pass the anti theft barriers normalizes running the barriers which will make catching thieves basically impossible. It also runs contra to shoppers expectations. Raises the question of charging the wrong person when multiple phones are nearby as when multiple people are shopping together or just adjacent to one another or when carts are just too close to the barriers. It also assumes that reading is instant and faultless when its not and fails to support anyone who wanders in from the street without an account set up expecting normal payment options. This makes it suitable for a membership club which expects to set up payment as part of setting up membership and useless for the 99% of stores that don't work like that.
Why not wait a few years until even a can of beans has an rfid tag, put a rfid tag in 4 corners of the cart. You should be able to compute which tags are within the bounds of the cart and allow you to pay.
The hardware to read the tags is still too expensive to put in the carts and not liable to be available as part of everyone's phones that soon so the easiest thing to do is retain the existing self checkouts and just skip the part where you scan anything. Roll up to the front. Tap your prominently displayed cart number or numbers tap your card done.
Reducing checkout to 6 seconds from multiple minutes will obviate the new loathsome situation where you actually wait in line to self checkout and will be close enough to normal procedures that it should be easy to transition to.
The scan as you shop systems that I am referring to are where you are given a little barcode scanner with a display, and you go around the shop scanning your items as you pack them. At the end of the shopping trip, you still go to the self checkout machine, but you just transfer your scan list from the portable barcode scanner to the till. At no point are you using a phone camera to scan items. Usually the trolleys have a little mount point for the scanner so you don't even need to touch anything.
I agree that having to scan out via the anti theft barriers is idiotic.
Walmart is demoing a scan as you shop system that I think relies on user cameras and the app. At least that was my impression.
With American hardware at SCO the scanners are just peripherals like a keyboard and a person walking around the store with them would tie up an entire POS computer while items were being scanned with it.
If the portable scanner were itself a portable device this would be tractable however how do you keep people from stealing or destroying them?
> Walmart is demoing a scan as you shop system that I think relies on user cameras and the app. At least that was my impression.
Walmart and Sam’s Club have both had this for years. It is only available at Walmart if you have Walmart+. I tried it when I had a Sam’s Club membership and it was ok but occasionally I’d have account issues, alcohol purchasing issues, and someone still wants to see your app receipt when you walk out the door issues.
It's like a gift-registry, but you take all the stuff with you when you're done instead of letting other people buy it for you later.
We've got Barcode guns at self-checkout here, and I've begun to arrange my items barcode-up in the cart, so I can just get everything rapidly. I often don't bring my bags with me, because I'm going to have to load everything into the car anyway.
That's become a thing at Home Depot, and I do exactly the same. I wish all the self-checkouts would use guns. Once you get used to just putting things in the cart code-up, it makes checkout pretty painless.
Heck, I even do the 'leave the bags in the car' trick too. I'm going to be using the cart anyway, so why do the extra step at the checkout instead of just loading it directly into the bags in the car.
Glad to know I'm not the only person who thinks this is a good way to do it.
>I often don't bring my bags with me, because I'm going to have to load everything into the car anyway.
Thats an interesting optimization. Don't bag your stuff in the store just walk out with a cart of loose groceries. Especially for someone who simply MUST bring all the groceries inside with one trip, I can just load up a crate.
Back in the days when we got plastic shopping bags, I considered it an art form to see if I could get an entire trunk load of groceries into the house in a single trip. Without squishing the bread.
You can get a lot of bags on your arms, and one or two jugs of milk in each hand depending on how long your fingers are.
It just isn't as fun with the reusable folding bags. Or the paper bags some retailers use -- the ones with handles are the worst, they rip a good part of the time. Actually easier to carry a normal paper bag than one with handles.
Plastic bags were great, I could hook as many as I could lift on a padded carabiner.
With mandatory paper bags they seem to have made them as flimsy as possible. I can't go grocery shopping when its raining any more, the bags just turn into mush at the slightest drizzle.
I've pretty much given up on reusable bags. I almost always do grocery shopping at random times, and never bring my bags with me when I leave the house.
That's why if you use a car you have more bags that you need on one run in the trunk. Then you keep bringing them in the house by twos or threes and at some point you drown in reusable bags and put them back in the trunk.
Having developed a hate for all those one use bags clogging your trash also helps. I've started doing this before it was the eco thing to do just because I hated accumulating bags.
Walking to the corner store(tm) if that option is available to you requires a bit more discipline, but I mostly remember to pick a bag when leaving home.
Interesting, it works the other way around here. You get a portable barcode gun that you carry around the store or put in a sort of cup holder on the cart.
So I always bring a bag or a foldable crate, throw everything in while shopping and then put that in the car.
I'm in the US, and we have both. It varies by store. Some have scan-as-you-go, some have self-checkout with fixed scanners, some have gun scanners. My personal preference is gun scanners at self checkout.
peel off the RFID tags from the expensive vine ripened tomatoes and replace them with tags from the really cheap ones (people already do this with the plastic stickers)
In the US (mid-Atlantic states) the supermarkets have tried this for years with very little success. Giant (Ahold) in particular persists with this, perhaps because they have a unionized workforce and would like nothing more than to slash their employee count.
I would credit it much less if it corresponded less well with my observations over more than 20 years shopping weekly at grocery stores, mostly Giant and its predecessors, in Maryland.
My current local grocery store has been my local grocery store for most of those 20-plus years. I remember when it was still a Super Fresh. It was a better place to shop then, and had about twice the current staff, despite at that time being quite a bit smaller.
That Giant has hollowed out the workforce is not in question. The question has been why; I didn't know grocery workers here were unionized, but if the union here is as effective as the one I hoped to join in my own high-school summer job days bagging groceries, then this frankly makes a lot of sense as a reason.
Stop and Shop (in New England) had hand scanners you could do this with 15 years ago. I have been shocked that the technology didn't spread. It was so easy to scan and bag then just plug the scanner in at checkout and pay.
Stop and Shop now has self-checkout where you have to scan and move your items one at a time into your bag, because it's doing continuous weight-based reconciliation.
So if you have six of the same yoghurt, you have to scan them and then place them in your bag one at a time. And if you have a 24 pack of sodas, you need to haul it out of your cart onto the scales as you're checking out. And if anything goes wrong (item didn't weight what the system expected; you moved the item too fast onto the scales etc) then you get a "please wait while someone helps you", which involves an employee having to come and clear the error.
There should be a hand scanner and a "skip bagging" option that allows you to keep heavy things on your cart. I'm pretty sure I have seen it at stop and shop before too.
I refuse to use those (although I will probably need to use it, eventually), because they do this annoying "BING!" when I pass whatever item they are trying to push.
2+3=5. In general, the weight idea is impossible. For any two items, you can find the LCM in units of the precision of the scale, and then you won't know if they got LCM/a of one or LCM/b of the other.
I think there's another technical issue with using RFID. If you have a basket of a few dozen different RFID chips, there's going to be a lot of collisions in the data they transmit, further reducing reliability.
At Meijer stores I've shopped at, there are fancy scales in the produce section.
They work like this: Scan the barcode on the produce if it has one (or pick it from a picture-list, or search by name, or just key in the 4-digit PLU if you're cool like that), put it on the scale, and it spits out a barcode label that identifies the product and the weight.
At self-checkout, one just scans the generated barcode label and puts the produce in the bagging area like any other item.
Unlike the self-checkout kiosks themselves: There's never any wait to use these machines, so it's an easy process that doesn't involve making other people wait.
This process would work the same, I think, for applications where portable scanners are in-use.
They have the same at my local Wegman's but there's really no incentive to use it. I'm not any faster than the cashier so there's no time savings.
And it has several problems as well with respect to loss (e.g. adding more items to the bag). To deal with that, you'd have to weigh it or count again to verify.
The advantage comes during shopping trips when one is not planning to use a cashier to check out, wherein: You're going to have the weigh the produce yourself, anyway, so you might as well get that done in advance.
1. The PLUs are posted on the price signs in the produce section, so it's easy to remember them accurately when I weigh stuff.
Entering the 4-digit PLU for starfruit is faster than navigating a list of pictures or searching by name, and is easy when I just saw it a moment ago: Short-term memory is a blessing and a curse, and this process makes good use of it. (I probably won't remember the PLU by the time I finish shopping and reach the self-checkout, except for bananas: That PLU has made it all the way to long-term memory and is 4011.)
2. Stress. The whole time I'm weighing out my produce at one of these fancy scales (I have to weigh it eventually anyway), I am at ease and am free to work at whatever pace I wish. This is not necessarily the case at self-checkout when there is a line of people behind me who simply want me to get done and be out of their way, so they can be done and out of there themselves.
3. The self-checkout line moves faster when my produce is already weighed. This offers me do direct advantage. But unlike some people, I am not a sociopath: Given the choice to make a thing easier or harder for someone else, at about equal cost to me, I try to do the thing that makes it easier for them. So I prefer that the line move faster, even if that doesn't actually save me any of my own time.
tl;Dr, I could behave more like crabs in a bucket, but instead I choose a more-orderly path and weigh my produce in advance.
Was going to say, this process sounds a lot more complicated than just putting stuff in my bag and letting a cashier deal with it. If they want to automate the cashier away, fine, but I'm not doing their job for them.
The Wegmans I shopped at in Rochester had scales that would produce a special barcode you could scan into the app that would record both the item and the weight.
If customers are willing to use an app, that makes things a lot easier to implement. Safeway/Vons in the US only have self-checkout machines. Seems like it's shifted towards per-item produce pricing, but the stickers often fall off.
See also the RFID tags Uniqlo and Decathlon use for self-checkout. You just place your shopping bag in a bin and it reads all the items within in a second.
My guess is that right now this isn't viable for lower margin food products. Uniqlo and decathlon probably don't sell that much for less than £10. It's going to be a while before you can reasonably do that for a pint of milk.
These would be 290 yen ($1.90) 790 yen ($5.20) in Japan respectively.
I still don't think an RFID will eat that much into their profit margin, but Uniqlo is in the unique position of having stricter control over their supply chain. All the products they sell are Uniqlo-branded products, manufactured specifically for Uniqlo, and they can just have RFID tags embedded in the price tags for all the products with little extra cost.
Grocery stores sells cans, plastic jugs, glass bottles, shrink-wrapped fresh produce and even plain vegetable and fruit or grain and legumes by weight. All of these comes from a wide variety of manufacturers that don't produce directly for you (or even have any kind of direct relationship with you).
The wide variety of form factors makes attaching RFID to everything unrealistic, and even for the stuff you could attach RFID to, it's not like manufacturers would do that for you.
When I worked there back in 2005, Best Buy had an RFID test setup in the basement of their HQ that could tell what you had in your cart without scanning.
Walmart was also working on similar technology, but from what I heard they couldn’t convince suppliers to include RFID tags on all of their products.
The problem with your second suggestion is that it gives the customer no recourse for improperly tagged/scanned items. Interaction with a human or checkout system before leaving the store is necessary, given how ripe for abuse a totally automated checkout system would be.
Agreed - I think there's also the issue of people needing to be pre-registered with such a system, and not being able to accept cash. I would expect that in the future, systems like that will be an "express lane" for the particularly prepared, but not necessarily the default.
Nothing, the same as current self-checkout systems. You can already easily just put stuff in your pocket or ring expensive items up as incorrect, cheaper things. I suspect shops have done a risk analysis and decided that they'd prefer to have more shoplifting and fewer staff.
At least in Walmart I've had workers come over to me multiple times to verify I correctly scanned things. There's definitely more scrutiny involved when you have a centralized self-checkout (this only happens in the more sketchy Walmart, the nicer one seems to trust people more).
> You can already easily just put stuff in your pocket or ring expensive items up as incorrect, cheaper things
This is actually already somewhat solved. The other day I was at a Safeway and one carton of milk had an unreadable label so I scanned a different carton and put the unreadable one in the bag instead (same product). The system showed me a video of me doing that, highlighting the fact that I didn't put the item I scanned on the bag (!) and asked me to wait for an assistant. Pretty impressive.
I doubt the relationship between the rate of shoplifting and ease of shoplifting is anywhere close to linear. Even before self-checkouts existed, people still shoplifted. And in some places, the honor system even works much of the time.
> I doubt the relationship between the rate of shoplifting and ease of shoplifting is anywhere close to linear.
I'm sure you're right. I suspect that there are massive variations from shop location to shop location, even within a region. I also suspect there are different kinds of shoplifting. I remember hearing a friend who is not the kind of person to just pocket an item and leave, bragging about ringing up protein powder from a bulk dispenser as flour to save money. That kind of behaviour is definitely going to be hard to model.
I use this regularly at the local stop & shop -- it's been in place for easily 10 years.
You scan your loyalty card (in my case a picture of it on my phone) and it releases a wireless barcode scanner with an android touchscreen phone-like thing in it.
Walk around the store, scan things and put them directly in my bags.
My favorite part is checking out without scanning anything and getting bemused looks (and occasional challenges) from other customers -- sometimes even from store staff.
Definitely saves time in my opinion over any other option. I only need to touch things twice vs 3 or 4 times and the wireless barcode scanner is much more reliable than the cacophony of things yelling at me with the self checkout. Weights don't always match, vision sensors think i'm doing the wrong thing.
I remember the one time I went to a just walk out store, I bought a chocolate and got the bill the next day which felt weird but now makes complete sense if they couldn’t get the Computer Vision to work but just relief on humans to identify this stuff.
I’d contend that even if a much more capable company like OpenAI tried this, they’d get much closer, maybe good enough for a product but this a really ill defined problem where even the best results may not be worth it.
Putting cameras on the roof doesn’t let you see everything a person grabs, there will always be weird occluded angles. On top of that the cameras need to be super high resolution and zoomed in to detect some of the smaller stuff in shelves. This is just for the trustworthy customers, now consider the thieves, they’re gonna have a field day with this, it’s not going to be easy to prevent a determined thief from not fooling the systems. A self checkout cart honestly makes much more sense, still prone to thieves but it at least lets you skip the line much more reliably.
amazon is notorious for paying late night sketch comedy shows to make fun of them in order to promote their niche projects that only exist in a few cities.
Not India, but my favorite example was the Kiwi food delivery robot fleet in Berkeley, CA. They were controlled manually from Colombia, and from the looks of it, seems like one person was trying to drive 20 robots at once.
A lot of these robot delivery things need to go out with a human handler 100 yards back lest people try and kick them open it seems. Whats the point at that point? Just have the person walk it over.
These ones didn't have a human handler. I think a lot of them just got broken into. I've also seen one get run over by a truck before because it wandered into the road.
Also, even if nothing went wrong, they were so slow that I can't imagine people actually used them. Wonder if they were just pretending to deliver stuff.
Expensify was a pretty well known case of this several years ago — their marketing was all about their advanced scanning technology, and it turned out they were using Mechanical Turk in many cases with little concern for PII (or corporate security) concerns.
(I have no idea if this is still the case, for the record.)
while I agree with the sentiment, as an Indian, I hope this doesn't happen in India. countries which typically do this mechanical turk-like work typically don't raise themselves out of poverty (esp. Philippines, Indonesia, etc.). If anyone wants a specific example, I lead an aspect of web crawling for a FAANG and then other public companies. Over the last 10 years we heavily used those offshore teams, aforementioned, to do sanity checks/labeling, etc. Now, we have initiatives with GPT APIs which perform just as well for pennies on the dollar we spent offshore - and the offshore team that's been loyal for years? They're getting cut.
I know companies that operate in that space and they pay incredibly well, between $20 to $50/hour.
> GPT APIs which perform just as well
That's because they were also trained by exploiting third world groups, paying about $2/hour.
The problem here isn't offering work to developing countries, the problem here is major corporations squeezing them for every cent and not allowing it to be used as a means of getting out of poverty. And that's also why the workers end up performing half-assed work by using automated classifiers and faking their credentials. It's not hard to see where this goes for both.
if you don't think FAANGs (and most companies) participate in "exploitative business" you should find out how your iphone was made (hint: lots of exploited workers).
Never said it wasn't. Amazon's antics especially are well known. The point here is that data labelling itself isn't fundamentally exploitative, even when leveraging developing countries.
I wonder if GPT-4's performance has degraded in recent months because there are less human data contractors on standby to answer questions GPT flags as low-confidence. GPT might be "refusing to answer questions" because it's not able to escalate tough queries to a human.
To be clear: ChatGPT-4 is in general both far too fast and far too stupid for humans to be answering any more than a tiny fraction (<< 1%) of the queries.
But last year I repeatedly saw ChatGPT-4 respond token-by-token much more slowly than a human would! E.g. several seconds between words. It was clearly not a human responding: at least a few times I was testing on preschool counting questions and GPT-4 was not able to answer them. I interpreted the slowness as GPT's poor quantitative reasoning. But what you're saying is simply not true, sometimes ChatGPT-4 is (or was) extremely slow.
Regardless, if OpenAI was running this con it probably wouldn't have been real-time humans writing. First of all it might be enough to have a human in the "mixture of experts" who decides the best of multiple responses when GPT-4 is unable to come to an automated conclusion. But humans could be writing ChatGPT responses due to a quirk in their UX:
- ChatGPT errors out on a certain question and asks you to try again later, as it does (or used to do) frequently
- the response is prepared by the human contractor while the user waits patiently for ChatGPT to resolve its technical difficulty
- when the user asks again ChatGPT can largely read off the answer, using its (genuine) language-processing abilities to handle variations in phrasing/etc
I suspect some/many of the online exam proctoring services are that way. Even some that market themselves as fully AI driven might be really AI = Actually Indians
Mechanical Turk suffers from coordinated fraud by people who want to be paid for doing a task without actually doing it [1]. The company I work for had to spend more engineering effort on building an internal reviewing-the-reviewers system to make it useful than we spent on the original Mechanical Turk integration. I'm not surprised that Amazon would avoid Mechanical Turk for higher consequence applications.
I wish Target would add computers to their carts. For most people, it'd be a nice alternative towards the self checkout area with limited space to place things.
For the more unsavory customers, it could offer video evidence that they're stealing stuff, lock the wheels at the door if more than $X isn't paid for, and create GPS reports of where all the stolen carts end up.
Target allegedly has such good video capabilities they can recognize repeat thieves, keep a record of what they steal, and wait until it reaches felony level after multiple trips before they flag a person to be detained. All in super duper high definition.
I think it’s deeper than that. Someone with a sheisty is gonna attract attention no matter what. The main idea is to combat organized shoplifting rings rather than - for example - addicts stealing small amounts to get more drugs.
They don’t convict based on the video or whatever. They learn their patterns and catch them the next time they come in and steal. Loss prevention associates know most of the usual suspects on sight.
Learning patterns doesn't get you a court-viable record of what they steal over multiple visits to push them over a specific money threshold.
So the most the gait-tracking computer can do is alert that a flagged person might have entered and needs to be strongly identified in some other way. But if loss prevention already knows them on sight then that system is minimally useful.
I expect the thieves are in and out before loss prevention even shows up, especially if they are masked up. Think about it logically: if loss prevention was able to actually work, they wouldn’t have to put so much merchandise under lock and key in high theft areas.
This is another thing Japan has implemented in certain areas. There's a supermarket near us called Trial that has this method of checkout: https://www.trial-net.co.jp/prepaid/regi-cart/
I've shopped at Amazon Fresh's London stores occasionally, mainly for the convenience and novelty. And sometimes I've had Amazon orders shipped there for collection. But never really felt the desire to do regular daily shops there because the prices seemed higher than Tesco etc for roughly the same sort of products?
Aldi has also been trialing the same "just walk out" shopping model at their Greenwich store in London[1]. I always suspected that theirs had some manual intervention involved because of the variable time it takes to get the bill. Sometimes I'd be charged nearly instantly after walking out of the store, other times it would be hours later. Aldi's one is more compelling, IMO, because you get the regular, competitive Aldi prices.
So they are offloading checkout onto the consumer via dash carts. There's no value proposition for the consumer unless somehow they are able to lower prices by not hiring cashiers.
You can also do this at Walmart (with Walmart+), though you still need to stop at the self checkout on the way out to get a receipt. It's pretty handy. I wish Costco would adopt this too.
I am just a few minutes walk from my local grocery store, so I go frequently. Rarely use a cart, just my canvas bag. Will there be portable scanners, or will I have to drag a cart around to benefit from real time scanning?
I don't feel like checking out is any faster than it used to be. They replace 10 human checkout lanes with maybe 15 self-checkout lanes and 2 human checkout. Maybe the self-checkout is a cheaper so they can have more, but people are so much slower on those that it doesn't win in the end. Dunno if the store is even really incentivized to make checkout fast, given that they sell high-margin stuff in the lines.
Maybe one advantage is it finally stops customers from asking to pay half with a check and half with cash for a $2 item in certain places. There's an entire Walgreens I avoid because that's happened 3 times already.
My local grocery store replaced about half the human checkout lanes with 3-per-lane self-checkout, and about 90% of the time I go shopping I don't have any line to wait for. The problem arises when they try to do a 1:1 (or nearly so) replacement.
self-checkout (at least in Hungary) is just clunky compared to the regular conveyor belt and chute setup, so no wonder the pros with the pro gear are faster :)
But I don't want to interact with the cashier. The store near me is almost the thing I actually want. I walk into the store, I wander about picking items I want off shelves, and scanning them with my phone, I put them in a bag I happen to have with me, or my pockets, or my backpack, or whatever, it doesn't matter, then I walk to the big graphical terminals near the exit and scan that too.
That's the only place where it has a step I don't need, it wants payment because capitalism, so there's a bunch of touch, tap, touch again sequences to get from it knowing what I took to being satisfied that I have exchanged money for goods and then I can go.
One thing I barely noticed at all until after it had happened was the removal of cashier operated checking out from most of the store. Ten years ago there were maybe 20 such checkouts, then sixteen after they added some self-checkout features - obviously rarely all manned but they existed. Today it's four, and often none of them are open, because almost nobody stands in a line to have somebody else scan your shopping, why would you?
Instead there are three distinct self-checkout zones. First, the one I use almost always, for people who have scanned their purchases already. There's nowhere to put anything, because these are just vertical terminals for taking payment. Then, short order terminals, a small built-in scale is provided so that you can buy say, lunch, but there's nowhere to load and unload a whole trolley of groceries. Finally the big shop terminals with lots of dedicated space for you to transfer from a trolley onto a scanner and weigh scales, then to your bags, this takes up most of the space previously occupied by cashiers.
If you've scanned everything you can actually use any of these areas, because the terminal understands that you've already done the whole job, but unless it's super-busy it will be easier to use the dedicated area for people who've done all the scanning.
> because almost nobody stands in a line to have somebody else scan your shopping, why would you?
A decently trained human can scan & bag far faster than I can at the self-checkouts.
The self-checkout machines are also sometimes buggy. Our local Safeway's self-checkout was extremely particular: if you scanned the next item prior to the computer registering and processing the weight from the previous item being set down, it would error out and require an attendant. That processing was quite slow — so you needed to lag yourself / rate limit yourself, so as to allow the machine time to think, before proceeding. I eventually figured out the cadence to go, such that I could avoid the error … but humans just don't have this issue.
Items too light to be detected just always error on some of these machines.
Woolworths in Australia has a solution called Scan and Go. You use the app on your phone to scan barcodes as you take things off the shelves. When you’re done, the app charges your stored credit card and you scan a qr receipt at the door as you leave which opens a little boom gate.
I believe there’s still some AI based visual monitoring in place but it’s much simpler. Instead of needing to identify exactly what you’re picking up, it just needs to flag when it thinks you might have picked something up without scanning it. If that happens, it triggers a staff member to come have look in your bags when you scan your receipt to leave.
It seems like a much more pragmatic way to realise 80% of the value you get with “just walk out”.
I feel like Sams Club has this down pretty well. You self-scan as you shop. You checkout as you walk towards the door. You flash a QR code to them and then they quick scan spot check items to make sure you weren't missing anything.
FWIW, this was the fourth attempt by Walmart to get Scan'n'Go working well (each one a complete reboot). It was majority designed and built by two talented developers who were acqui-hired and managed at at distance by another acqui-hire developer.
I think this is funny. I worked on this team at Amazon for two years after successfully building an app similar to Dash Cart for another company using Elixir (which is still going). I repeatedly tried to convince Amazon to do Just Walk Out differently, but their toxic culture wouldn't hear of it. I got bored with the project (and a little disgruntled, to be honest) and left after my signing bonus ended.
I guess it is good that they figured it out in the long run, but they could have fixed this in 2020 if they had just listened.
I used these in London and they were magical and extremely convenient. I would show them to visiting friends as if they were one of London's main tourist attractions!
I knew they relied on some manual labor (bills were delayed), but assumed this was only an intermediate step to collect more real-world training data (until they sorted out the AI).
I still think this is the future though. They could reduce staffing costs and free-up floor space that would otherwise be used for self-checkouts. All the while actually drastically improving the customer experience too.
If someone cracks it, it may make it impossible for other supermarkets to complete economically.
As a fun aside, all the stores were incredibly over-staffed and helpful. If you couldn't install the app they had a browser-based fallback (type "fresh code" into the amazon searchbar on a mobile). They could also lend you a battery bank if you were low on battery. I think they even had phones to lend out to those without smartphones. It was all very optimized around customer experience first and foremost and with no real attempt for the stores to be profitable.
I’ve used “Just Walk Out” at Amazon Fresh regularly over the past few years and it has been both convenient and reliable. But the machine learning is clearly stuck at a level where the need for ML data associates is cost-prohibitive: 70% requiring human verification is an order of magnitude off from their target of 5%.
Notably, they are continuing “Just Walk Out” at a few locations in case future advances make the technology more viable.
> Though it seemed completely automated, Just Walk Out relied on more than 1,000 people in India watching and labeling videos to ensure accurate checkouts.
I know this is ripe to use as a way to dunk on Amazon, but I sincerely doubt that it was as inane as it sounds. They likely had this workforce to train a system that they hoped would eventually operate autonomously. That being said, it sounds like it never panned out.
> They likely had this workforce to train a system that they hoped would eventually operate autonomously. That being said, it sounds like it never panned out.
And this is why I think current “AI” technologies are a dead end. Once the rest of the market realizes that maybe we’ll finally get past this “AI” hype bubble.
That's an interesting point. Amazon tried an AI application where errors were a cost to Amazon, rather than to the consumer. They failed to make it work well enough to use.
What AI applications are in use where the error cost is imposed on the company using them, not the consumer? Reading paper checks and mail envelopes come to mind. Those are automated and work well. Waymo's automatic driving, but not Tesla's. What else?
This is a useful metric for gauging real progress.
> Those are automated and work well. Waymo's automatic driving, but not Tesla's.
If Tesla geofenced their FSD applications as aggressively as Waymo, I suspect the two would be near peers. You can hide an incredible lack of a capability by artificially limiting the scope of your application.
> This is a useful metric for gauging real progress.
Sure, cost / benefit analysis is always useful. I would propose a broader analysis for grading “AI” disruption though.
1. Are humans involved at any step?
2. What is the probability of success?
3. What is the consequence of failure? (Profit loss / property loss / personal injury, etc…)
4. How many variations on a domain (think winter driving, summer driving, driving in a tropical storm, driving dirt roads in a jungle, etc) can the application work in while maintaining consistent performance?
5. Can the model leverage its “learnings” to solve problems / reason about new domains that it hasn’t been trained on?
That’s just a few metrics I can think of off the top of my head that would help to distinguish a true disruptor from the hype.
Feedback: I'm a Prime customer and I've never even got into one of the stores. Having the on-site staff waiting at the front of the store like bouncers just put me off as I walked by! Hahah ...
I also kind of like the idea of people having jobs, so that's also a factor. But I didn't even get a chance to try these out ... and I would have but for the henchmen and women waiting to welcome me, LOL
Heh, Read the instructional sign outside the store. Opened up my Amazon App and got the QR Code on my screen. Could not figure out where on the turnstile to scan it, and I'm a barcode app guy. There was a screen prompting me for my palm print. Um, no. You don't get my money AND my biometrics. We turned around and walked out.
To top that off, the associate carting out in the parking lot ridiculed us and gave us a "buh bye". So, Amazon Fresh is off the list for me. I don't even go to Whole Foods after that experience.
I don't live in a country with anything like this sort of retail experience, but I feel like there is a middle-ground approach which is a fast self-checkout with async payment. Maybe even just dumping your cart on a conveyer and letting the products be scanned in a more controlled environment, or even just QR code stickers.
Uniqlo has something similar to this - they use RFID's on their stock so you can dump it into a bucket at the self-checkout, and it scans everything immediately [1]. You still have someone pop over to check if you are OK, but it is a lot quicker than self-scan or usually waiting for someone.
Is it intentional that so many people are confusing/conflating teams in India labeling data for input to a machine learning algorithm vs. teams in India watching you shop in real time? It's the former. That's how machine learning works. You need labeled training data as input. The article clearly states it's only a small minority of all customers that had to be corrected in real time by actual people.
I've done all checkout styles and I think the best is scan as you shop with a little handheld scanner is best by far, you can pack as you go along and I think I've had a random check like 3 times in years of doing a weekly shops.
I like having a running total and being able to price check, it's mostly fantastic. Only downside is fruit and veg which you have to weigh and print a sticker for, that's about it.
In my mind the perfect experience would be a trolley/cart/basket with a screen on it that detects what was inside.
There was a really neat company called Digimarc[0] a while back that promised to bake the barcode into the packaging such that you could scan any part of the label and receive a barcode successfully. I have no idea what happened to them because it seemed like a slam-dunk idea that would have greatly improved the purchasing process virtually everywhere (and you could just have a traditional fallback barcode in case the baked-in one fails).
The problem with that is the same problem with self-checkout and Amazon Dash. It's solving a problem the retailer has -- the cost of employees -- by turning shoppers into store employees. But it doesn't really solve any problems that the shopper has because scanning all of your own stuff and laying it out on the scale is so inconvenient that it ruins everything else. The only place where I've seen self-checkout be as good or better than having a checker is Lowes or Home Depot, because they don't require you to weigh each item in turn to unlock the scanner so you can move as fast as a checker can. But I think grocery shopping will be the last place self-checkout moves in in a big way because the grocers insist on weighing everything you buy before unlocking the scanner.
Most of the retailers in my area have turned off the bag scale security. But I live in a pretty small town, so maybe theft is at an acceptable level where they’re willing to make that trade off. They’re also using more and more computer vision to ensure what passes into the cart is scanned (Everseen and others), so perhaps they’re finding better success with that than the bag scale.
The scale sensitivity was a bit off initially, but it seems okay nowadays. (If I dropped something on the bagging area it didn't wait for the item to settle.)
I wonder about putting some sort of camera-readable code or even a wireless signaling device (passive/reactive?) on the packaging, even as stickers, a modern equivalent of barcodes for their readers.
Amazon would have to add the stickers for now, but it could revolutionize retail. IIRC I read that a bookstore chain revolutionized their retail industry by putting barcode stickers on all their books, allowing computerized inventory, and now publishers print them by default.
This always made a whole lot less sense to me then say, solving the logistics of grocery delivery.
Which isn't solved despite availability, because it's "solved" by just having someone else walk through a store, maybe doing a few customers orders all at once.
Like it would seem to me that for, at least packaged goods, an entire city could and should be serviced out of a couple of warehouses with better stocking and fulfillment systems?
Say goodbye to the future! The one that IBM promised us 18 years ago in this advertisement: https://vimeo.com/29120357
In fact, that ad also shows how awkward the concept is (or maybe the fact that we're still not ready for it). It's a really fun well-made video but a thoroughly unconvincing advertisement.
> . Though it seemed completely automated, Just Walk Out relied on more than 1,000 people in India watching and labeling videos to ensure accurate checkouts. The cashiers were simply moved off-site, and they watched you as you shopped
There was one in London I used to walk past all the time. It was a neat experience to shop there once, but the selection was very slim and ocado was even more convenient.
I'd always peek through the windows and it never had any shoppers.
> I'd always peek through the windows and it never had any shoppers.
My experience using these locations is that I entered, grabbed the items I wanted and just left. Because they were closer to convenience stores, I was basically speedrunning it. A drink and a sandwich? Maybe 30 seconds tops. There wasn't any point in lingering. I wonder how much of that changed the dynamics of how many people were inside?
My experience with London, Canary Wharf - I'm hungry but the darn Amazon app has auto offloaded. Spend 3 minutes re downloading with bad data signal. Buy sandwich. Meh - it was ok but normal check out easier for me.
Chancery lane Tesco who also had one - darn I don't even have the Tesco app, I guess I'll wait for a normal shop.
For me the system would be better if you could just tap your card like everywhere else rather than needing an app.
Wait what? When did they reveal it's offloaded to workers in India? I remember this being touted as fully automated when it was announced, surely being watched by strangers from creepy angles would discourage a large portion of customers.
ChatGPT is also trained by the efforts of thousands of people in low income countries getting paid a pittance. We don't hear about it much because it's easier to credit a few suits at the top as harbingers of the future.
There is usually a mention of "humans in the loop" when articles discussed Amazon Go[1]. Amazon just always refused to give actual numbers. The best "give" that humans were watching you was that Amazon Go receipts often arrived in your email many hours after you left.
Anyone who works in computer vision immediately knew this wasn't possible beyond a cool proof of concept in a tiny room with <300 SKUs. Even in 2024
Less creepy being watched by actual people I think. They'll usually just forget about you after a few days instead of cataloging every time you lingered at a shelf selling hemorrhoid cream and storing it in a database forever.
There was just a Shark Tank episode where the entrepreneurs pitched essentially this. Nobody questioned the tech side of it, but they still didn't get a deal.
Cool. Now can I pay with cash? Because I'm not going to let anyone build a profile on whether I like warming lube or just regular lube, how much sugar and alcohol I consume, what kind of romance novels I read, how often I floss, and whether I feel I can afford organic fruit.
Hmm you make a good point, but have you considered that your health insurance provider really wants to know these things, and will pay up to $0.35 to learn them? It would be criminally negligent for Amazon not to build and sell that profile!
no, no, because you can just grow your own, or just buy guns with cash and use them to exchange them to food. (and if it's not interstate no need for paperwork, right? - in freedom states)
Some of the questionable "grocery store tech" that I have seen Amazon push out:
- "just walk out" - what a joke. just another way to capture more data points on your "shopping habits" and build more meaningful profiles against you. Oh John picked up the "Radical Juice X", let's start pushing out the "GOP" package of ads when John searches on Amazon for products. Prefer products with "red" color.
- "pay with your palm" [at Whole Foods] -- yea, right. I am going to give out my handprint to amazon. piss off
Surely the data of what products you picked up is much less valuable than the data of what products you actually bought, which would be trivial to collect with a normal checkout system.
> Though it seemed completely automated, Just Walk Out relied on more than 1,000 people in India watching and labeling videos to ensure accurate checkouts. The cashiers were simply moved off-site, and they watched you as you shopped.
This is both sad and pathetic. Not to mention devastates the local community. Instead of kids being able to get their first part-time job in a local shop stocking shelves or doing checkout these opportunities were completely offshored abroad.
My concerns about this were theft, mistakes, cost to operate, and opportunity costs.
Theft is an obvious one. It happens in self checkouts in many ways. We had people come into our store to just leave with baskets of stuff, load their pockets, take stuff from one box to put in a cheaper one they bought, or even put raw eggs and meat in their pockets. That was with a ton of workers in the store looking out for thieves.
Errors could result in us being charged for what we didn’t buy. The amount could be lower or higher. Some people go through their whole receipt double checking it. Many of us just watch it as they ring it up, deal with it there, and stop thinking about it once we leave. Could this impact peace of mind?
Cost to operate seemed like it might be high. Aldi and Costco show a traditional self-checkout can be cheap to operate on top of serving high-end customers. (See theft/shrink, though.) I figured the machine learning alternatives would cost thousands to tens of thousands a unit for combinations of cameras, servers, and licensed software. They’d also need to be monitored and/or trained by humans which this article just confirmed.
I was also worried about fraud. My company bought a solution that monitored customers with infrared cameras, monitored checkouts, and predicted how many registers we needed. That was a terrible system that hurt us more than it helped. We were required to act like it worked to please those who spent hundreds of millions on it. I wondered how often such dishonest referrals helped the supplier sell them. Made me skeptical about more AI.
Finally, the human touch that comes with good cashiers increases customer satisfaction. Our customers often looked forward to seeing us. For some, we were the only or best social contact they got that week. Others liked getting out of the house for a long period of time. Sometimes they’d have us push products and services which, normally a negative for all, is largely how they drove up credit card use and feedback on the customer tracker.
For those reasons, I advised against using AI for anything other than helping employees do their jobs more easily. If for that, that it would be a support tool instead of trying to replace skilled, decision makers. Then, simply invest those millions to hundreds of millions in the employees. Worked for Costco and Publix. :)
Many thieves are cowards, too. They steal less when they think people are watching. Their risk analysis is based on if people are there, if cameras are there, if they think people really watch thieves on camera, and if the store prosecutes thieves. Every box you tick can reduce theft. I think Target was in the lead on that stuff.
Re scales. The scale detects a weight difference between items. If the items are similar weight, the scale or cashier will ignore it. The scales also had a high, false alarm rate that often caused cashiers to ignore them. The thieves also use scanning methods that cause higher, false alarms to encourage that. None of that works with human cashiers and a bascart exchange.
Re registers needed. You would think. The company was top-down, micro-managed, low staff, and blame the worker. The supervisor had to open registers in time, maybe watch self-checkout while the attendant does price checks, and often work a register themselves for most of the shift. Good service was impossible in those conditions.
Instead of having cashiers, the executives decided an AI that thought for us would fix the problem. Of course, we had no cashiers to open registers. They marked us as failures on that, too.
Later, they had us force more people to use self-checkout, closed more registers to facilitate that process, and installed more self checkouts. I’ve been gone a few years. It seems the customers have grudgingly accepted the status quo because both major companies, the cheapest with most products, were intentionally making checkout terrible. Oligopolies in action…
JWO can work in cultures where most people are honest e.g. Japan.
JWO stands no chance in cultures where grift is the norm e.g. NYC where people jump turnstiles let alone snatch groceries. Even self checkout struggles. In these cultures, checkout staff invite their friends to shop then "accidentally" forget to ring up an item or two. If you can keep shrinkage to a few percent you're doing great.
Source: longtime in the industry and owned/operated 3 retail shops myself.
JWO as Amazon does it has your identity to begin with, and as per TFA, the problem wasn't keeping people honest but the amount of human verification needed.
If anything, self-checkouts are a larger target to grift and yet they seem to do fine.
> Though it seemed completely automated, Just Walk Out relied on more than 1,000 people in India watching and labeling videos to ensure accurate checkouts.
More of the same. I wonder if the incredible use of labor in India and other countries will cause their society to bypass their own growth so that the citizens of those countries become pure consumers, the majority of which can only ever get menial or gig jobs due to AI taking over most of innovation. In such countries, the only jobs left may be to aid in the extraction of larger and large amounts of natural resources required for the technological expansion of the west.
In this way, developing countries will never be able to gain independence from the few global powers at the top: instead, they will be forced to follow the folly of the United States and Europe: developing more technology for the sake of technology, and in return they will get a free Unlimited Streaming plan to watch the garbage that we have come to call entertainment.
> I wonder if the incredible use of labor in India and other countries will cause their society to bypass their own growth so that the citizens of those countries become pure consumers
Never gonna happen in India, by the sheer number of populace. You can see the thing you are talking about in some countries in Europe where a small population and a lack of heavy manufacturing drives the whole economy to the servicing sector. It's a question if you can label these countries as developing, though.
Stores in earlier times in Norway³ were geared towards convenience¹
for the shopper.
You went in, went up to the counter, told the clerk what you wanted,
a series of clarifications, advice and suggestions, then clerk² would
fetch your goods, wrap it up all nice for you, handle payment and hand
you the goods.
All you did was request what you needed, got advice, paid for it
and off you go⁴
Then came the supermarkets.
Suddenly every customer was a mouse running in a maze trying to find
the cheese, milk, washing powder and whatever else.
Endlessly walking up and down every isle desperately seeking what
they needed.
Then the customer had to accumulate all the products into bags or carts,
Push or carry it all to the cash register.
Where one person (per register) would sit on their ass, add it all up,
you paid, you bagged, and you were on your way.
A lot of customers did not like that at all.
It wasn't rational.
In the old system the clerk at a store knew where everything was.
It was not necessary for ever customer to memorize and learn every
store they went to in order to go shopping.
It was a massive inconvenience for everyone, except for the person
at the register who was sitting comfortably.
These days, not only must you hunt in the maze,
push it to the counter, now you must also do the
checking out yourself.
Finally, the companies have outsourced all the labor to
the customers themselves and they are happy about it.
There are of course still stores that operate the old way.
but they are in the minority in most places.
When I read about it the first time it felt so backwards.
It more convenient to be able to grab what you need and
be on your way.
Then as I was running up and down isles trying to find what I
need i started noticing and thinking about it.
I dont really think it is rational that everyone has to do all that
walking and hunting.
Perhaps with robots we can be back to the old ways.
But really the old system is back in places where you
order things online and go pick it up, all picked
by clerks, all you need to do is pay and be on your way.
and I love it.
When it comes to groceries and many other products
they will even deliver it to your house.
Its how I get most of my groceries as well as other stuff.
Now I can sit on my ass, and I have outsourced all of that
back to the store.
The old becomes new sometimes.
In fact often.
I would be interested in a study that calculated what amount
of our lives on average spent walking around in stores trying
desperately to find what you need.
¹ It may not match what we think of as convenience today.
² The clerk may work on his own, or there might be more than one,
there might be people working in the back fetching things from storage.
But your point of contact would be the clerk.
Of course often the owner in smaller stores.
³ Probably in most other places as well.
⁴
> Though it seemed completely automated, Just Walk Out relied on more than 1,000 people in India watching and labeling videos to ensure accurate checkouts.
That line was unintentionally hilarious to me. Like looking inside a TV and finding a bunch of fast-working little people drawing images.
> Though it seemed completely automated, Just Walk Out relied on more than 1,000 people in India watching and labeling videos to ensure accurate checkouts.
Yup.
I got the chance to go shopping at one in 2018 (so it's been a while). You could tell there was some "reconciliation" happening, because your receipt didn't show up until 20-30 minutes after you left. My guess is that this was some person via mechanical turk crawling thru the data and indexing what you -really- bought.
Of course, my colleague and I (who were working on computer vision at the time) did stuff like take our bags off, put them in the middle of the floor, and roll a can of soda into the bag, just to see what would happen.
"Of course, my colleague and I (who were working on computer vision at the time) did stuff like take our bags off, put them in the middle of the floor, and roll a can of soda into the bag, just to see what would happen."
I'm many jurisdictions this could be enough to get you prosecuted for shoplifting if it evaded payment.
Yeah, the author is trying to see what would happen if they roll a can into the bag. This is meant to conceal the act from the computer. The act of concealing merchandise is all that is required in many jurisdictions. If they walked out without the concealing act, then I would think the error should be on the store.
There have been some cases of self checkout prosecution too. They don't need any proof of intent. Even regular errors can be prosecuted.
> They don't need any proof of intent. Even regular errors can be prosecuted.
In Common Law jurisdictions, there's no intent required for prosecution of alleged theft, but there absolutely is intent required for conviction! Anything else is a miscarriage of justice.
Except statute and common law can define intent using prima facia definitions. In PA it's considered prima facia evidence as defined in statute that you're hunting if you're walking in the woods with a gun. Doesn't matter your true intent. If you removed an item from a store without paying for it, that's prima facia evidence of shoplifting defined in common law. Again, true intent doesn't matter. The courts will even create president that goes against statute just to make conviction easier - the statute for dog control says it must be "under reasonable control". Instead the courts have ruled that it is a strict liability offense and not subject to the reasonable standard.
So yeah, there's a lot of miscarriage of justice, or at least potential for it. Nobody really cares until it happens to them.
(This article is not the AWS service by the same name, though by the time you conclude reading the Wikipedia page, it should be obvious where the AWS service got its name, and what it does.)
Imagine that next time you are drunk, you can hire a driver who will drive you back home remotely (along with some AI to stop the car in case connection goes away).
I disagree - at a basic level if people are able to tell which products you're picking up the resolution is necessarily high enough that there's a huge creep factor. I am sometimes amazed at what companies will pursue without a glance towards common sense.
I'm a huge privacy advocate and don't think stores should track what you buy at all, so don't confuse what I say next with the is/ought fallacy.
They already have perfect resolution and data retention of everything you buy at checkout time when it's scanned, plus they can verify your identity rather than have to rely on facial recognition or other things. I don't think this is any creepier than what they already do so from their perspective it is "common sense."
The scandal where Target data scientists bragged to reporters about knowing when teenage girls are pregnant before their fathers broke in 2012, and they said they were doing it since 2002. It was based entirely on data mining purchase histories with rewards cards at the register. No fancy AI or facial recognition needed.
That's fair as far as tracking purchases go, but high resolution cameras while You're in the store could also read your phone screen or other things on your person right? That seems more concerning than normal cashiers.
Yes very fair point. I just had a dystopic thought of dynamic prices (electronic price tags?) that update in real time based on information. Is the person undecided? Brand A bids to lower their price slightly. Is the person looking at the same item on Amazon? Read the price and beat it by 10 cents or whatever. Those are situations where it might benefit the shopper, but I could easily see it going the other direction also. not looking at the item on Amazon? Now you're gonna pay too much. And of course, store the complete history of what strategy works with which people so "the algorithm" can tune for your individual weaknesses.
One creepy thing existing doesn't mean that another thing can't also be creepy. Getting rid of one of many creepy things is still nice even if it's not the only one.
Except all these trillion dollar valuations in the AI bubble are based on the belief that AI is replacing humans, not just outsourcing to cheaper humans.
Of course outsourcing to cheaper humans can be great. But that's not what tech is shilling to the world as "AI" right now.
The grocery industry generally has pretty low margins. Wal-Mart's profit margin is 2.39%.
If you go into a grocery shop to grab lunch, spending $8 on 4 items, and they make 5 cents of profit per item? They need to run an extremely lean operation.
they simply need to run a good ordering/inventory system. If they sell every item (on average) in each store every week, that's 2.39% return on the value of the inventory investment each week, or 52*.0239 or over 100% annual return on money they borrow for free because the store pays its grocery bills to suppliers in arrears, net 10 days, etc
What I mean is: If you're making 5 cents of profit per item, and your workers cost $10/hour then the difference between a profit and a loss is 18 employee seconds per item.
> Imagine that next time you are drunk, you can hire a driver who will drive you back home remotely (along with some AI to stop the car in case connection goes away).
Is this satire? It doesn't seem like a fantastic idea to allow someone to remotely pilot a car over a transoceanic Internet link.
The article they link for that isn't readable without subscription. Were they actually labeling data live? That sounds like effectively a secret concierge following you around and writing down what you're getting & like it would be pretty difficult to keep up with. Otherwise, labeled data and machine learning kind of go hand and hand. It's not wild that Amazon would dogfood its own product while creating training data to improve it.
It does sound like a person was watching the shopper live:
> Though it seemed completely automated, Just Walk Out relied on more than 1,000 people in India watching and labeling videos to ensure accurate checkouts. The cashiers were simply moved off-site, and they watched you as you shopped.
But who knows, tech articles are so inaccurate that it might as well have been some misunderstanding regarding how the labeling/training was done.
Yeah, that's line was what got me trying to look at the article they cite, which is unfortunately paywalled. It'd be one thing to do that at like... several stores to generate the training data. But scaling it to dozens while it still requires that? Baffling.
It sounds like the receipts weren't necessarily "live", sometimes taking hours to appear, so I'm guessing they did it as live as possible and when they got backed up they would just revisit recorded video. Something tells me the expectations for throughput were also pretty overwhelming to the individuals who worked there.
Kind of reminds me of how back when speech-to-text started to get good enough to use (in the form of Google Assistant/Siri), my parents wondered if there was someone on the other side doing the transcription.
i've written about this multiple times on HN and always got downvoted
I work in the CV industry and this was a very badly kept secret. There is absolutely no financially feasible way to run something like this with current CV
I work with ML CV industrial systems, and they can certainly do accurate and detailed analysis and identification very quickly. The systems that do this are also necessarily very, very expensive -- much more expensive than any grocery store could possibly justify.
Of course, costs change with time, but right now I don't see how this sort of application could approach being financially feasible.
Seattle Target was the worst shopping experience I've ever had. No cashiers because the employees are tasked with opening locked cabinets instead, even for sunscreen. Security guards with bulletproof vests add to the ambience. Gave up on buying oranges cause the checkout machine wouldn't scan them.
Target, Walmart and all other big chain stores just gave up after the pandemic in 2022. I saw people came and just grabbed whatever they see fit, including a TV or a PS5 out of the door. I understand employees were trained not to intervene but even police were told by city council as well.
Target now puts every valuables behind steel doors, not even a cage anymore. That's fucked up to think about how our society has turned out.
.jpg){kind=link}
When it rolled out to Amazon Fresh stores, it was a breadth of fresh air. The painful clunkiness of self-checkout was gone. The slow and pointless exercise of unloading and reloading your cart was gone. You could just bring your reusable shopping bags, throw stuff in, and walk home. By far the most hassle-free shopping experience to be had.
Scan as you shop is a big step backwards and feels like you've got the annoying self-checkout experience looming over you the entire time you're there.
The selection and operating hours both took a hit during covid and never recovered.