zlacker

Or Sam was the driving force behind increasingly closed research and that went against the board's commitment to "benefit all humanity"?

Maybe the closed GPT-4 details were promised by him to be a one time temporary thing at the time and then he has been continuing to stonewall releasing details later on?

>>kromem+(OP)
Possibly. But that doesnt sound serious enough to constitute "hindering [the board's] ability to exercise its responsibilities".

Maybe its the off-the-books Weapons Division with all those factories in obscure eastern European countries. Or the secret lab with the agi that almost escaped its containment. /s

Money or power. I guess someone will eventually talk, and then we'll know.

>>kromem+(OP)
Since the beginning of OpenAI, haven't we been slowly surprised by the progressive closedness what it was becoming. I think there were multiple threads on HN about this, and the irony in the name. Maybe this has been going on for much longer and reached a tipping point.

>>kromem+(OP)
Following this argument, perhaps the line about Sam being "not consistently candid" is an indirect reference to his preferring the closed approach...i.e. they wanted him to be more candid, not in his reports to the board, but with the public, regarding the research itself.