Assertion Supported
Sitaraman: OpenAI's board fired Sam Altman for lacking candor, not safety
“The board in their statement as sort of lacking in detail as it was said that they didn't fire Sam for anything but lack of candor, right? Like they didn't fire him because he wasn't AI safety enough.”
Assertion Supported
Sitaraman: Reports of a Q* letter prompting the OpenAI coup are false
“One thing that I've now heard from a few different people is that the reports about the, a letter being sent to the board about Q star that, that that's not true. And then it didn't precipitate the board's decision.”
Assertion Not checkable as stated
Sitaraman: Anthropic screens job candidates for Effective Altruism culture fit
“There's a culture fit test. Like when you get hired where you're asked that, that as far as I understand, I don't have like the questions in front of me, but a lot of the questions are sound, they're basically trying to test whether or not you'd be EA enough. …”
Assertion Supported
Sitaraman: Marc Andreessen's manifesto equated slowing AI development to murder
“When he did his tech optimist manifesto, when was that? A couple months ago. Didn't he say that any type of deceleration of AI would cost lives? And so if you're slowing down AI development, you're basically a murderer.”
Assertion Supported
Sitaraman: OpenAI's superalignment team tried to create an automated AI scientist
“And one of the projects that, you know, we reported on was they were trying to create an AI scientist. So this is like not a science, like a human scientist, like an AI scientist that would study alignment.”
Assertion Supported
Sitaraman: OpenAI only recently hired one person for 2024 election policy
“And so open AI a couple months ago, just hired a, somebody on their policy team focused on the 20, 24 election.”
Assertion Not checkable as stated
Sitaraman: Ilya Sutskever identifies as an AI safety advocate, not EA
“If you talk to people who know Ilya, he's like, I'm not EA, I'm AI safety.”
Prediction Not checkable as stated
Sitaraman: Effective Altruism will retain strong influence in AI hiring and policy
“So you're probably going to hire a bunch of EA people and you're going to have to make them happy because they could leave at any time. Right there. There's huge competition for people who actually build these systems. So there's that, but then, you know, so t…”
Prediction Not checkable as stated
Sitaraman: Anthropic likely won't fracture over EA controversies due to shared baseline
“Will they be split up by it? Probably not just because they are all coming from the starting point that EA is good. It's just maybe the execution is a problem or we are misunderstood and it's a public issue.”
Assertion Not checkable as stated
Sitaraman: All major AI labs officially denied being Effective Altruism-aligned
“For this story, we asked all the companies, by the way, if they agreed with characterization of them as EA or not, and all of the companies told us they are not EA aligned.”
Assertion Supported
Sitaraman: Anthropic employs philosopher Amanda Askell, ex-partner of EA founder Will MacAskill
“They have a philosopher on staff who. Really? Yeah. Everybody does that. Right. But it's Amanda Askell. So she's the former wife, I think, or of Will McCaskill, who is the, right, the founder of EA.”
