This is really, really clearly incestuous tech media stuff as part of a pressure campaign. Sam is the darlin of tech media and he's clearly instigated this reporting because they're reporting his thoughts and not the Board's in an article that purports to know what the Board is thinking, the investors who aren't happy (the point of a non-profit is that they are allowed to make investors unhappy in pursuit of the greater mission!) have an obvious incentive to join him in this pressure campaign, and then all he needs for "journalism" is one senior employee who's willing to leave for Sam to instead say to the Verge that the Board is reconsidering. Boom, massive pressure campaign and perception of the Board flip flopping without them doing any such thing. If they had done any such thing and there was proof of that, the Verge could have quoted the thoughts of anyone on the Board, stated it had reviewed communications and verified they were genuine, etc.
But the board seems to have a weak hand. It can decide to disappoint the for profit investors. But it doesn’t own Sam, or the vast majority of the workers, and maybe not much of the know how. And they can walk if the board disappoints them.
The board’s altruism might be great, but it lacks the legal tools to do what it wants, against organized labor backed by unlimited capital.
The supply bottlenecks have been around commercializing the ChatGPT product at scale.
But pretraining the underlying model I don't think was on the same order of magnitude, right?