Why do you think those things? Logic & theorem provers massively predate LLMs, and getting LLMs to use them is as easy as asking the LLM to write a proof in the language of your theorem prover of choice, which you can then copy-paste into that theorem prover and execute. And if/when it doesn't work, the error message itself helps with a significant fraction of most other programming problems, so my guess is they would also help here.
Also, there have been substantial new developments and discoveries about what transformer models do (both internally and in terms of capacity) every week or two for most of this year, so why do you think there's no known path forward?
Thinking of logic, I just tried the following with gpt-3.5, gpt-4, and gpt-4-1106-preview. The newest model spotted the trick (and then still got it wrong), the older two didn't even spot the trick. Can you spot the trick?
-
A person is in Nairobi. They board a plane, fly 9000 km north, then 1000 km east, then 9000 km south, then 1000 km west. Where are they now?
Also, there have been substantial new developments and discoveries about what transformer models do (both internally and in terms of capacity) every week or two for most of this year, so why do you think there's no known path forward?
Thinking of logic, I just tried the following with gpt-3.5, gpt-4, and gpt-4-1106-preview. The newest model spotted the trick (and then still got it wrong), the older two didn't even spot the trick. Can you spot the trick?
-
A person is in Nairobi. They board a plane, fly 9000 km north, then 1000 km east, then 9000 km south, then 1000 km west. Where are they now?