One tantalising statement in there is that AGI-level system is not bound by licensing agreements that a sub-AGI system would be (ostensibly to Microsoft).
This phase-shift places a pressure on management to not declare reaching a AGI level threshold. But have they?
Of course, it could be an ordinary everyday scandal but given how well they are doing, I'd imagine censure/sanctions would be how that is handled.
But.. what are the responsibilities of the board that may be hindered? I studied https://openai.com/our-structure
One tantalising statement in there is that AGI-level system is not bound by licensing agreements that a sub-AGI system would be (ostensibly to Microsoft).
This phase-shift places a pressure on management to not declare reaching a AGI level threshold. But have they?
Of course, it could be an ordinary everyday scandal but given how well they are doing, I'd imagine censure/sanctions would be how that is handled.