Some important questions from an EA perspective I don’t know the answer to:
How much of an improvement is it on previous models, and how was it made [i.e. is it all just scaffolding/​ longer reasoning, or was this a bigger model]? Is scaling still working? How has it changed AGI predictions?
How much safety testing was done on this release? What were the results?
Is this being received by the market the way OpenAI hopes /​ how is the AI bubble doing?
Some important questions from an EA perspective I don’t know the answer to:
How much of an improvement is it on previous models, and how was it made [i.e. is it all just scaffolding/​ longer reasoning, or was this a bigger model]? Is scaling still working? How has it changed AGI predictions?
How much safety testing was done on this release? What were the results?
Is this being received by the market the way OpenAI hopes /​ how is the AI bubble doing?