[sambuccid]

I think sometimes you have to look at the patterns rather than at the single claim. If a large amount of people, that are completely unrelated, tell you very similar experiences they had with Altman, you can take that as a good indicator of his general character.

And if this tendency to misunderstand/be misunderstood always results it Altman gaining more power, even if we give him the reason of the doubt and say that doesn't do it on purpose, it's still a big problem, given the responsibility he has.

The article also mentions many moments where apparently Altman straight out lied, as opposed to being "very persuasive, if you believe those sources then I don't think it's also possible to think he's sincere. I cannot open the article again to get the exact quotes, but the few I remember were: - one time he was claiming he didn't send a message, while people were literally showing him the message he sent, with the confirmation of another OpenAI employee - another time when he accused people of organising a coup, and that someone from the board informed him, and after the person from the board was called in the meeting Altman claimed he never said those words and never accused anyone

These cases can't be put to persuasion, that Altman changed their view, or that someone misremembered, they either happened or they didn't

[Lerc]

>I think sometimes you have to look at the patterns rather than at the single claim. If a large amount of people, that are completely unrelated, tell you very similar experiences they had with Altman, you can take that as a good indicator of his general character.

Yes, but that doesn't work if you look for patterns selectively. There are large amount of people who will tell you vastly different experiences that they had with Altman. If you pick the right grouping, within it, you can find universal praise or condemnation. The article itself acknowledges that.

>The article also mentions many moments where apparently Altman straight out lied.

Does it? It has people saying he lied, and a few things he disputes that he said. If the lies were clearly apparent, I think his position would not be tenable. Which points in the article do they show statements that it clear that he has said them, that they were false, and that he knew they were false when he said them?

The points you list are not clearly apparent lies. At most they are allegations of lies. They might just be different interpretations of the same events. I have seen instances in my own life where someone has said "You said X" the other person says "No I didn't", The first then pulls up the minutes, and says "See you said X", the other responds with "That's not what that says". You see rage bait posts about terms and conditions that take that form all the time. Someone misreads a legal term as meaning something different to what it means in a legal sense and then refuses to acknowledge the commonly accepted definition.

Please respond to this, because I really am interested in the answer, but I did read the article and I didn't see what you appear to have seen.

I have made no claim to the merits of Sam Altman, I just don't like the idea of condemning someone on hearsay and insinuations. There are videos on YouTube claiming he's had people killed. At some point you have to point at something that everyone can agree on is an actual thing that happened and that it actually matters. At most what I have seen is people being able to provide one of those two points on any particular allegation.

I don't feel this should be that contentious. If it were clear there would be demands from all around saying "You did this bad thing, you must resign". Do you think that everyone dealing with OpenAI acknowledges some dark truth and is complicit?

[sambuccid]

Sorry I didn't mean that the artice has proofs he lied, just that some of the situations presented cannot be simple misunderstandings.

The pieces in the article I was referring to are:

> Amodei’s notes describe escalating tense encounters, including one, months later, in which Altman summoned him and his sister, Daniela, who worked in safety and policy at the company, to tell them that he had it on “good authority” from a senior executive that they had been plotting a coup. Daniela, the notes continue, “lost it,” and brought in that executive, who denied having said anything. As one person briefed on the exchange recalled, Altman then denied having made the claim. “I didn’t even say that,” he said. “You just said that,” Daniela replied. (Altman said that this was not quite his recollection, and that he had accused the Amodeis only of “political behavior.”)

> Amodei discovered that a provision granting Microsoft the power to block OpenAI from any mergers had been added. “Eighty per cent of the charter was just betrayed,” Amodei recalled. He confronted Altman, who denied that the provision existed. Amodei read it aloud, pointing to the text, and ultimately forced another colleague to confirm its existence to Altman directly. (Altman doesn’t remember this.)

I agree it's very easy for 2 different people to understand or to remember something differently, and that meeting minutes are not always a reliable source, but for me in the 2 scenarios above is almost impossible for 2 people in good faith to disagree:

In the first case, if you say something, and a big deal is made of it, and 5 minutes later the other person claims that you said some specific words and you deny it, then someone is lying, either you or the other person.

In the second case, if there is something written in a contract, and someone presents that contract to you, reads it out loud, and asks a collegue to confirm, either that person made up the provision, or you are lying, there is little room for misunderstanding.

Given there are no proofs, I can't say he's 100% culprit, and I appreciate your rigor on this because we don't want to result judging everyone by a sort of "trial by public opinion".

However, outside of trials, the judjment can be more nuanced than a boolean "culprit/innocent", and to me the reasons below(*) are enough to distrust Altman and to prefer he wasn't the person at the head of a revolutionary technology that could have huge negative consequences on the society, or on human kind as a whole.

(*) the reasons being:

- amount of people interviewed and their very similar experiences

- the author and the type of journalism he does

- the professionalism he shown in calling out in his article the not-backed allegations other rivals made(for example of murder and sexual assault)

- the power dynamic that is usually in place between someone with enormous power and whealth, and a journalist that could be intimidated by being sued multiple times

Of course the amount/type of reasons needed to distrust someone is very personal, so we might need to "agree to disagree" on this

[Lerc]

That's interesting, I'm sure when I read the article it didn't specifically attribute those claims to Amodei.

To be frank, While I tend to think that Dario has good intentions, I'm not so sure about his judgement. He's made a lot of claims that haven't panned out. I haven't felt that it was due to dishonesty, but more because of hyperbole.

The phrasing "Altman then denied having made the claim. “I didn’t even say that,” he said. “You just said that,” Daniela replied." is very close to the pattern I described above where someone interprets a claim as something different from what was actually said and refuses to back down. Furthermore this was prefaced with "As one person briefed on the exchange recalled" so it isn't even a first hand account. We don't know who the person doing the briefing was, but if it was one of the participants of the exchange, they would have been afforded the opportunity to reframe it to put themselves in a better light.

The second claim is potentially even more of a match for the example I gave regarding people misreading legal documentation. Was this a denial about the existence of words in a document, or was it a denial that the words represented the provision that was claimed. I have seen people do this, they take the existence of the words as proof of their interpretation and take dismissal of the interpretation as a claim that the words do not exist. I don't rightly know why people do this, but I have seen it happen. I suspect you could find an abundant supply of cases like this from the records of the worlds town council meetings.

It is difficult to assess the reliability of claims made by the current administration (understating it somewhat), but one of the things that was said about the Government negotiations with Anthropic was that he wanted a gate to some AI abilities in national security circumstance by requiring a personal phone call to Amodei to clear it. No sane government on earth would agree to something like that. It would be an invitation to providing a corporate interest a massive point of leverage in a time of crisis.

But again I am in a similar position with Amodei. I don't have any direct knowledge of the person so I will reserve judgement. I generally like the approach Anthropic is taking but the exposure I have had to the statements made by Amodei himself has given me pause. I would not condemn him either, but I also wouldn't place a lot of stock in what he says unless I see more to create a more complete view of his character.

You note amount of people interviewed and their very similar experiences but it's the nature of how those claims are similar that concerns me. So many of the claims seem to fall into the pattern that requires the person reporting the claim to judge the sole meaning of what was said. How many confirmed direct quotes have been confirmed to be untrue? I'm open to the evidence, perhaps this article will draw some out, but right now I see people convincing themselves of a pattern and then interpreting their own experiences in terms of that pattern.

The thing is, if you were to ask, I think Altman would agree that he shouldn't be in charge of the world's AI. I don't think any one person should, and I would treat anyone who claimed that they were the right person for that job with massive suspicion. To say that's where he sits is to buy into the premise that whoever is the head of OpenAI controls our future. OpenAI is but one of many enterprises working on this, there are a lot of people claiming they already have lost too much ground, but then there have been many predicting their imminent collapse, like a doomsday cult rolling forward the calendar whenever it doesn't happen.

[sambuccid]

> That's interesting, I'm sure when I read the article it didn't specifically attribute those claims to Amodei.

Apologies, I didn't mean to highlight Amodei in those quotes, I just selected the sentence to have enough context but not be too long, it was a coincidence that they both started with Amodei. I'm not sure if those claims came from Amodei or not, nor I have any specific feeling about him.

> Furthermore this was prefaced with "As one person briefed on the exchange recalled" so it isn't even a first hand account

I'll admit I somehow missed that part, but we don't know how much of this event was in "Amodei's notes" and how much was from the "person briefed on the exchange"

> The phrasing "..." is very close to the pattern I described above where someone interprets a claim as something different

> The second claim is potentially even more of a match for the example I gave regarding people misreading legal documentation

I think our difference in point of view here lies on how much trust we put in the author, with what I seen so far I feel I have enough trust in the author to think he investigated these claims properly and made sure they weren't just misunderstandings, and that many of those checks he did weren't included in the article for any technical/legal reasons. Much more so reading some of his comments:

> As is always the case with incredibly precise and rigorously fact-checked reporting like this, where every word is chosen carefully (the initial closing meeting for this one was nearly eight hours long, with full deliberation about each sentence), there is more out there on that subject than is explicitly on the page.

> You try to reach a critical mass of detailed, rounded understanding of a central question, integrating the most meaningful perspectives, interrogating the weak points and blind spots, and backing up the assertions with documentary evidence or strong sourcing. Eventually, you reach a point where enough sources and materials are reliably triangulating toward the same truths.

> The fact-checking process at the New Yorker is exhaustive, and can span weeks. Every sentence, assertion, and piece of underlying sourcing get scrubbed by multiple independent pairs of eyes. This story had four fact-checkers working on it for the better part of a two week period, pulling very long hours.

As I said I'm happy to agree to disagree on this point.

> So many of the claims seem to fall into the pattern that requires the person reporting the claim to judge the sole meaning of what was said

I guess that's the nature of communications between humans. Even examples of written discussions seem contentious. The only type of claims I can think about that could be outside this category are the ones about written contracts, but it's understandable we don't have access to the actual contracts, and even if we did we couldn't really prove what was verbally agreed to be put in the contracts.

> To say that's where he sits is to buy into the premise that whoever is the head of OpenAI controls our future. OpenAI is but one of many enterprises working on this

This might start a whole new discussion, but I think being the CEO of one of the companies that produce state of the are models is enough to have a high concern. My worry is that he(or any other company) won't say "stop" if a new AI is found to be more powerful but have considerable negative impacts on society. As an example it doesn't matter who has the "strongest" atomic bomb, any country that has one is a potential treat to humanity and should have rigid controls in place.

I commented specifically on Altman because the article seems to suggest he's more power-greedy, persuasive, possibly deceptive, and with strong-leverages/contacts than the average person, or even the average CEO.

(edits: formatting)