Depends on how you interpret the mission statement of building ai for all of humanity. It’s questionable that humanity is better off if ai only accrues to one or a few centralised entities?