How does this resolve to NO?

Looks like just "If no such info has come out by deadline, resolves NO."

Additionally, Toner says she was personally targeted by the CEO after she published a research paper that angered him. “Sam started lying to other board members in order to try and push me off the board,” she says.

https://www.theverge.com/2024/5/28/24166713/openai-helen-toner-explains-why-sam-altman-was-fired

The description states that being mean in pursuit of business or policy is not evidence for YES. The article claims he was doing this based on an ai safety article she published makes it less evidence for YES. It was a conflict over the company's policy re:safety and things got violent.

Work power conflicts around work issues are not personal.

personal issue for conduct outside of work, or personal-relationship violations within work.

This doesn't seem like what the article is about

@Ernie two former OAI board members wrote a recent article, including:

Last November, in an effort to salvage this self-regulatory structure, the OpenAI board dismissed its ceo, Sam Altman. The board’s ability to uphold the company’s mission had become increasingly constrained due to long-standing patterns of behaviour exhibited by Mr Altman, which, among other things, we believe undermined the board’s oversight of key decisions and internal safety protocols. Multiple senior leaders had privately shared grave concerns with the board, saying they believed that Mr Altman cultivated “a toxic culture of lying” and engaged in “behaviour [that] can be characterised as psychological abuse”.

clearly Altman's behaviour and the effect on company culture was a factor, even if these factors were also perceived to have on-effects related to AI/safety

@shankypanky As long as the issues they're referring to were AI/business/AI safetly related, this sounds similar to the item which is listed as meaning NO:

> Things that would not count

• > being mean at work in order to change the priorities of the company

• > being very stubborn about things, or too opinionated, or keeping secrets at work to manipulate the board into making decisions he wants
  

In writing the description, I tried as hard as I could to clarify that YES are for private, non-work decision related issues, OR items which are related to him personally and unethically benefiting such as embezzlement, or preferentially hiring/firing people for personal benefit etc.

If they had a difference of opinion about balancing AI safety vs the growth, power, and wealth of the company, that is NOT a personal issue. If the allegation is that he was just a general pathological liar for personal psychosis/selfish/unethical/"sinful" etc reasons, that would lean to YES. But afaict nobody is alleging that. Everyone is saying it's about AI (kicking out AI safety guys cause they wanted to block company growth and power) not personal. I would like to see related evidence if there are things I'm not including in the above, though.

@shankypanky The lead-in to the article is so well-done, but then you get to the specifics of the concerns, and rather than them being things which violated the mission, it's some vague stuff about they don't like the way he talks or emotes. That type of allegation is trivial to make against anyone, and incredibly hard to prove or disprove. i.e. it's not worth much to me.

X (with less power) says Y (with more power) is a bad guy cause they don't like his tone/attitude.

This is super weak. If the claim is true why don't they say something like: "In a meeting where we discussed making sure we didn't violate our charter to protect and benefit humanity, he lied about his intentions to get us to agree to an edit of a document, which he then never produced, and then denied the whole thing. <and here are the documents>" which at least an arbitrator can look at if they're not public. i.e. make a specific claim about something bad he did, rather than basically saying "bad vibes" and hints.

Anyway, then the article derails itself talking about whether "such behavior" (being mean or using social means to suppress dissent, including firing people who repeatedly go against what the CEO wants) is enough to dismiss the CEO. And the answer must be no. There will always be ppl with less power who resist the CEO and criticize him in vague ways. Such people need to provide proof of actual bad actions by the CEO, or quit, or be fired.

Given the authors, I really hoped for a sentence like "Sam did X which was bad" where X is a real decision. There must be some budget allocation, firing/hiring, manipulation of safety data, hiding of negative reports, bad performance review, bad allocation of resources etc that they don't like? How exactly is his bad behavior causing... risk of world destruction? Is it too much to ask for them to specifically say what is bad?

If they've signed the NDA, will they move to Brazil and then speak freely? This allegedly is the world on the line here. Can they please just speak freely about what bad stuff, specifically, he did? I assume they... hit print screen occasionally? Will they leak those docs?

I really would like to know what the claims against him are.

This feels like it should have time decayed, hard.

26days later, this still feels like a clear NO for me. The issues were about his professional role and behavior not some outside-work personal misbehavior

@Ernie Resolves YES

The ChatGPT maker tapped the law firm WilmerHale to look into what led the company to abruptly fire Altman in November, only to rehire him days later. After months of investigation, it found that Altman’s ouster was a “consequence of a breakdown in the relationship and loss of trust” between him and the prior board, OpenAI said Friday

https://apnews.com/article/openai-sam-altman-chatgpt-2f3303dff2280478947e9bfb04863537

@Shump Loss of trust itself is not decisive either way. the description makes it rather clear, that personal issues are limited to non-AI related decision-making.

I will go over all the sentences and evaluate them:

something like "lying about business issues" doesn't count.

i.e. this says that lying about business issues is NOT a personal issue. I think it's reasonable to interpret the allegations against him to be basically that he attempted to replace board members who held certain views about AI risk, so that OpenAI and himself would be more free and unrestricted in their development. Yes, this would clearly benefit him personally - but, he is the CEO of OpenAI so all pro-business, or business-beneficial actions would; that doesn't mean they're personal.

Things that would count as YES

• behaving improperly towards someone in/out of work would count.

i.e. sexual harrasment, emotional torture. No one alleges this.

• stealing candy bars from walgreens

No one alleges this

• stealing money from work

No one alleges this; in fact, if his motivations were to free OpenAI from non-profit bounds, his actions would be very pro-business.

• using his influence to try to get a friend/partner hired or someone fired

I did not mention this explicitly, but the implication is that this would be an IMPROPER example of this behavior. Proof: imagine he is friends with a world-class brilliant new engineer who is looking for a job, and he attempts to use his influence to get the person to be hired at OpenAI. Is it reasonable to interpret that as a wrong, personal action? Of course not; if you do so, it means that nearly every exec is basically continuously guilty of taking actions (i.e. trying to hire smart people who would be great for the company, but who are also their friends). Similar for fired - based on his public statements, he gives a very reasonable explanation why he may have used some pressure, persuasion, or even things which might be considered manipulation to have someone removed from power. That job, too, is the duty of an exec, if he or she is doing it for a reason which is aligned with business goals.

Things that would not count

• being mean at work in order to change the priorities of the company

The description calls out that using emotions or interpersonal behavior, which is conventionally, or in the real of sin & moral virtue, considered wrong, even THAT would not be considered to be any type of personal misbehavior, since such tactics are common at executive level.

• being very stubborn about things, or too opinionated, or keeping secrets at work to manipulate the board into making decisions he wants

This section clearly anticipates him possibly using secrecy (which involves lying) and specifically defines that as NOT a personal matter.

The hard cases are where he is alleged to have used his authority to change company direction, but also in a way which is alleged that he would profit from personally.

Again, obviously, profit him personally in ways which are conventionally considered wrong - such as using one's CEO power to give contracts to organizations which the CEO him or herself secretly contains an ownership stake in, and keeping that from the board.

@traders FYI, my current views, as explained above, seem rather different than ones which would lead a price of 97% to be reasonable based on what is known public. If, however, certain parties were in possession of private info, I would have no way to evaluate the reasonableness of bets at all.

As an additional check, I provided all related information to Anthropic's new LLM, Claude 3 opus, asking for his opinion on this. Overall, it was basically in agreement with my prior comment. A copy of the interaction is available here: https://docs.google.com/document/d/1k--oZALrG4TgmWPU4fzko7WiFMZtIKHJ9uTRRPy-T1w/edit?usp=sharing

Feel free to comment.

https://twitter.com/axios/status/1725938914018439383

‘We can say definitively that the board's decision was not made in response to malfeasance or anything related to our financial, business, safety, or security/privacy practices. This was a breakdown in communication between Sam and the board.'

Would worldcoin self-dealing count as personal?

@HastingsGreer yes

limit at 69