zlacker

There's a prediction market here about why he was fired: https://manifold.markets/sophiawisdom/why-was-sam-altman-fir...

>>g-w1+(OP)
Still too much in flux to even copy top contenders. Top 6 at this point:

Fundamental disagreement about OpenAI's safety approach

Negligence in Addressing AI Safety Concerns

Sexual misconduct

Conflict of Interest with Other Ventures

Defrauded OpenAI

Cash Flow problems

>>oezi+46
The fact that AI safety is the top two shows how delusional AI safety people are

>>g-w1+(OP)
That's interesting. More like a prediction market about whatever becomes publicly told as the reason why he was fired. If e.g. this was military-related, it's unlikely it would be publicly known, so betting on "Military" is almost pointless

>>airstr+jn
eh, military contracts would become public. it became public that Google was building a computer vision system for targeting drone strikes for the Pentagon, and a number of executives resigned over it. If Altman were in talks to build a version of GPT-4 for the NSA, even if it were a black project, they'd have to account for the revenue somehow.

>>sterli+Br
> t became public that Google was building a computer vision system for targeting drone strikes for the Pentagon, and a number of executives resigned over it.

er what? source?

>>typon+k7
And/or, how active they are on manifold

>>g-w1+(OP)
lol is this some kind of joke haha.

>>banana+jt
This was a huge deal and widely reported. You can easily find copious reporting with a web search if you don't like this link:

https://www.theverge.com/2018/3/6/17086276/google-ai-militar...

>>typon+k7
It sounds like one of the board members is an "AI safety" person, so it's not that crazy.

>>typon+k7
OpenAI is an ideologically driven, political organisation, around the political and technological question of how to deal with AI safety. They only hire people who are AI safety people using "culture fit", and their board was constructed before they were for profit + doesn't hold any equity in the company and only owes a fiduciary duty to the OpenAI charter (which is entirely about AI safety). The board of OpenAI kicking out the CEO of OpenAI is much more likely to be related to AI safety than any other company in a similar situation, this isn't AI safety people projecting their own concerns and what's important to them onto an entity that doesn't care. OpenAI very obviously does care about AI safety, as an organisation, even if they're more towards the "build it" side of the debate in AI safety.

>>airstr+jn
Military contracts are posted and solicited publicly. There's no "dark" acquisition of the type that you are suggesting. You can look up if OpenAI has any contracts with the DoD at [0]. They do not.

[0] https://www.usaspending.gov/

>>g-w1+(OP)
That list of reasons is a mess with tons that overlap. Sexual misconduct is there like three different ways, plus one that’s explicitly abuse of his sister.