hey so about your favorite boy hank green.
-
@davidgerard Eh? My take was that Roko's Basilisk was Calvinist Satan, punishing people for eternity because they were predestined for punishment for not being one of the Elect who were Saved because [incomprehensible belief-based selection process].
@cstross oh no, Roko is very clear that this is the Coherent Extrapolated Volition (CEV), the AI charged with preserving human value, the *good* AI that was Yudkowsky's goal
that's why it upset so many rationalists, it was an extrapolation of Yudkowsky's actual goal to disaster
-
@cstross oh no, Roko is very clear that this is the Coherent Extrapolated Volition (CEV), the AI charged with preserving human value, the *good* AI that was Yudkowsky's goal
that's why it upset so many rationalists, it was an extrapolation of Yudkowsky's actual goal to disaster
@davidgerard @cstross It's very silly.
-
@cstross oh no, Roko is very clear that this is the Coherent Extrapolated Volition (CEV), the AI charged with preserving human value, the *good* AI that was Yudkowsky's goal
that's why it upset so many rationalists, it was an extrapolation of Yudkowsky's actual goal to disaster
The basilisk also only punishes believers, which is kind of the opposite of what the god of Reformed Christianity does.
Hilariously, this doctrine falls squarely into one of the oldest traps of Christianity: if innocently ignorant nonbelievers aren't damned, then leaving them to their ignorance is the moral thing to do, right? By alleviating their ignorance you'll just create the possibility of their damnation.
That means that a faith which wants to spread via missionaries has to either believe that innocent ignorance is damnable, or else suppress this line of thought. Christianity chose the former, and Yudkowsky chose the latter.
-
The basilisk also only punishes believers, which is kind of the opposite of what the god of Reformed Christianity does.
Hilariously, this doctrine falls squarely into one of the oldest traps of Christianity: if innocently ignorant nonbelievers aren't damned, then leaving them to their ignorance is the moral thing to do, right? By alleviating their ignorance you'll just create the possibility of their damnation.
That means that a faith which wants to spread via missionaries has to either believe that innocent ignorance is damnable, or else suppress this line of thought. Christianity chose the former, and Yudkowsky chose the latter.
@passenger @davidgerard And then there are those of us atheists who were born into and grew up in non-Christian religions and think Christian dogma is hilarious wank.
(Except when it's coming with pitchforks and torches to murder your community.)
-
@passenger @davidgerard And then there are those of us atheists who were born into and grew up in non-Christian religions and think Christian dogma is hilarious wank.
(Except when it's coming with pitchforks and torches to murder your community.)
As an atheist who grew up in a churchgoing family, I assure you that you cannot possibly think worse of Christianity than I do.
-
As an atheist who grew up in a churchgoing family, I assure you that you cannot possibly think worse of Christianity than I do.
@passenger @davidgerard Self-identified Christians murdered most of my extended family two generations ago. (I recognize that not all Christians are murderers by inclination, but it's not the kind of thing you can un-see once you're aware of it.)
-
@passenger @davidgerard Self-identified Christians murdered most of my extended family two generations ago. (I recognize that not all Christians are murderers by inclination, but it's not the kind of thing you can un-see once you're aware of it.)
My ancestors fled from Europe because they were the survivors of the massacres that Christians were carrying on against other Christians in Europe. The survivors were very much a minority.
I don't want to downplay what your family went through. I can feel kinship for it.
Did my ancestors learn from this that massacres are bad? No, they learned that their form of Christianity needed to become equally violent and oppressive, and they taught that violence and oppression to their kids. They became exactly the sort of people who would do the massacres in turn, and they did it in the name of Christ.
Not everyone is a murderer by inclination, as you say. I think a lot of it has to be taught: you are taught to dehumanise the victims, and you are taught ways to manage your conscience. To do that sort of teaching properly takes generations, and having seen it up close I don't want it to remain extant.
-
@cstross oh no, Roko is very clear that this is the Coherent Extrapolated Volition (CEV), the AI charged with preserving human value, the *good* AI that was Yudkowsky's goal
that's why it upset so many rationalists, it was an extrapolation of Yudkowsky's actual goal to disaster
Honestly the practical understanding (since it's actually impressive how little they end up reading of the things they've claimed to read) ends up feeling a whole lot like Maxwell's demon- literally.
Letting the cool people in.
The actual 'you will be tortured otherwise' seems to upset them enough to not dwell on it, and I definitely encountered a few that were simply 'It's just you don't get to hang out with the best' tier.
Edgy pascal's wager, really.
-
hey so about your favorite boy hank green. it’s such a shame that he went overboard with the LLM stuff, that was impossible to see coming, good on him for realizing he was addicted right? well wishes for the good boy!
anyway there’s smoking gun evidence that he was paid by effective altruists to do marketing for the “AI is so dangerous but so revolutionary” flavor of critihype. this is alleged by hank green expert hank green in the washington post, as well as an EA in charge of influencer marketing. he was literally being paid by TESCREALs.
someone in the same community ran a program where TikTokers were required to post an AI critihype video a day in a cult-like environment where they were subjected to “voluntary” waterboarding and other forms of indoctrination. I’m not kidding, this literal thing happened.
here’s @davidgerard’s pivot to AI with the evidence linked: https://pivot-to-ai.com/2026/08/06/the-ai-doomsday-cult-recruits-social-media-influencers/
@zzt @davidgerard Honestly one of the biggest annoyances to me with AI people and also one of the huge "you're in a religious cult" red flags is that literally *every* time it becomes relevant that I don't like generative AI the conversation *immediately* turns to "oh, you poor, unenlightened soul, let me show you the light and convert you!"
-
@zzt @davidgerard Honestly one of the biggest annoyances to me with AI people and also one of the huge "you're in a religious cult" red flags is that literally *every* time it becomes relevant that I don't like generative AI the conversation *immediately* turns to "oh, you poor, unenlightened soul, let me show you the light and convert you!"
@zzt @davidgerard Not once have I *ever* talked to someone who likes gen AI to have them open with "it's interesting you have a different opinion than me, let's talk about it and maybe we'll both walk away with something new."
It's *always* "haven't you read our pamphlet of 16 reasons why AI is amazing, yet?!" and they're *always* the same claims which were easily refuted like 4 years ago.
I'm legit so tired of arguing against the same "AI learns like people" points from 2021...
-
No, I was misreading. The phrasing of OP implied to me that the linked post had evidence for the tiktok cult thing, not the hank green claim.
*reads linked post*
and, having read through the linked post and the WaPo article, I don't see anything in there beyond "he did a single video which was a disclosed sponsorship", while also doing other videos where he said he disagrees with it?
Certainly doesn't look good but this feels very much like "internet people being internet people and turning everything into the worst possible version of reality"
@pixx @davidgerard yep, you commented without reading. thanks for clearing that up
and you’re still not reading in any depth
> Green, the YouTube star who did a paid post with ControlAI for his science channel, said in a statement that he wants to give his viewers a better understanding of this increasingly powerful technology by exposing them to a variety of views on AI. He pointed to recent unsponsored interviews on his personal YouTube channel that included critics of AI company hype, Sen. Sanders and one of the co-authors of “If Anyone Builds It, Everybody Dies.”
green is in a partnership with ControlAI that produced one sponsored video and an ongoing series of “unsponsored” (a term of art meaning there’s no sponsored tag or company logo on the video) videos with major EA figureheads pushing an AI doom framing. these aren’t critics of AI company hype, they are the hype.
-
@pixx @davidgerard yep, you commented without reading. thanks for clearing that up
and you’re still not reading in any depth
> Green, the YouTube star who did a paid post with ControlAI for his science channel, said in a statement that he wants to give his viewers a better understanding of this increasingly powerful technology by exposing them to a variety of views on AI. He pointed to recent unsponsored interviews on his personal YouTube channel that included critics of AI company hype, Sen. Sanders and one of the co-authors of “If Anyone Builds It, Everybody Dies.”
green is in a partnership with ControlAI that produced one sponsored video and an ongoing series of “unsponsored” (a term of art meaning there’s no sponsored tag or company logo on the video) videos with major EA figureheads pushing an AI doom framing. these aren’t critics of AI company hype, they are the hype.
@pixx @davidgerard this is green confessing to taking money from ControlAI, a wealthy political organization, to falsely represent AI doom as science. if you see that and don’t see a problem, American public school failed to prepare you for grown up reading.
-
hey so about your favorite boy hank green. it’s such a shame that he went overboard with the LLM stuff, that was impossible to see coming, good on him for realizing he was addicted right? well wishes for the good boy!
anyway there’s smoking gun evidence that he was paid by effective altruists to do marketing for the “AI is so dangerous but so revolutionary” flavor of critihype. this is alleged by hank green expert hank green in the washington post, as well as an EA in charge of influencer marketing. he was literally being paid by TESCREALs.
someone in the same community ran a program where TikTokers were required to post an AI critihype video a day in a cult-like environment where they were subjected to “voluntary” waterboarding and other forms of indoctrination. I’m not kidding, this literal thing happened.
here’s @davidgerard’s pivot to AI with the evidence linked: https://pivot-to-ai.com/2026/08/06/the-ai-doomsday-cult-recruits-social-media-influencers/
@davidgerard by the way, I’d like to highlight something from Green’s statement in the Washington Post because it’s worse in context and I want to make sure everyone has that context:
> Green, the YouTube star who did a paid post with ControlAI for his science channel, said in a statement that he wants to give his viewers a better understanding of this increasingly powerful technology by exposing them to a variety of views on AI. He pointed to recent unsponsored interviews on his personal YouTube channel that included critics of AI company hype, Sen. Sanders and one of the co-authors of “If Anyone Builds It, Everybody Dies.”
ControlAI is a wealthy political entity in the effective altruist sphere, and Green took their money in order to present their AI doom marketing as both science and as legitimate criticism of AI. given that AI doom presupposes the impending creation of an AI god and is the preferred framing of the AI companies themselves, it is neither science nor critique.
-
@davidgerard by the way, I’d like to highlight something from Green’s statement in the Washington Post because it’s worse in context and I want to make sure everyone has that context:
> Green, the YouTube star who did a paid post with ControlAI for his science channel, said in a statement that he wants to give his viewers a better understanding of this increasingly powerful technology by exposing them to a variety of views on AI. He pointed to recent unsponsored interviews on his personal YouTube channel that included critics of AI company hype, Sen. Sanders and one of the co-authors of “If Anyone Builds It, Everybody Dies.”
ControlAI is a wealthy political entity in the effective altruist sphere, and Green took their money in order to present their AI doom marketing as both science and as legitimate criticism of AI. given that AI doom presupposes the impending creation of an AI god and is the preferred framing of the AI companies themselves, it is neither science nor critique.
@zzt @davidgerard Well, fuck him. He's dead to me.
-
@davidgerard by the way, I’d like to highlight something from Green’s statement in the Washington Post because it’s worse in context and I want to make sure everyone has that context:
> Green, the YouTube star who did a paid post with ControlAI for his science channel, said in a statement that he wants to give his viewers a better understanding of this increasingly powerful technology by exposing them to a variety of views on AI. He pointed to recent unsponsored interviews on his personal YouTube channel that included critics of AI company hype, Sen. Sanders and one of the co-authors of “If Anyone Builds It, Everybody Dies.”
ControlAI is a wealthy political entity in the effective altruist sphere, and Green took their money in order to present their AI doom marketing as both science and as legitimate criticism of AI. given that AI doom presupposes the impending creation of an AI god and is the preferred framing of the AI companies themselves, it is neither science nor critique.
@davidgerard Green stated that he did this in order to present a variety of views on AI, but that’s not true. all of the interviews and AI videos on Green’s channel are in support of ControlAI’s AI doom narrative. there are no opposing viewpoints challenging the ridiculous and unscientific idea that godlike AI is a possible and impending risk to humanity.
all of the interviews Green has done have been with people who are either allied closely with ControlAI or are similarly tied up in Effective Altruism and AI doom (yes even Bernie, sorry about our boy).
the article is coy about this, but the “If Anyone Builds It, Everyone Dies” co-author Green interviewed was Nate Soares, a core TESCREAL personality who came up with the term “AI alignment”, now used in AI doom marketing by every EA organization.
the other co-author was Eliezer fucking Yudkowsky.
-
@davidgerard Green stated that he did this in order to present a variety of views on AI, but that’s not true. all of the interviews and AI videos on Green’s channel are in support of ControlAI’s AI doom narrative. there are no opposing viewpoints challenging the ridiculous and unscientific idea that godlike AI is a possible and impending risk to humanity.
all of the interviews Green has done have been with people who are either allied closely with ControlAI or are similarly tied up in Effective Altruism and AI doom (yes even Bernie, sorry about our boy).
the article is coy about this, but the “If Anyone Builds It, Everyone Dies” co-author Green interviewed was Nate Soares, a core TESCREAL personality who came up with the term “AI alignment”, now used in AI doom marketing by every EA organization.
the other co-author was Eliezer fucking Yudkowsky.
@davidgerard for anyone unfamiliar with Yud: oh boy you don’t want it
in short, Yudkowsky invented the modern, AI doom cultist, capital-R form of Rationalism. without Yud and Soares, the TESCREAL belief package wouldn’t have the shape or the political power that it does now. these are core figures. Yud did decades of work exposing billionaires to the Rationalist and Effective Altruist ideas that he and Soares came up with.
the reason why the interview wasn’t with Yudkowsky is because his name is now poison, as a lot of people who listen to the guy quickly realize he’s a crank.
Green’s paid video with ControlAI led quickly to an interview series where he platformed one of the biggest TESCREAL figures in the subculture.
and note that when Green says the interviews were “unsponsored”, that’s a YouTube term of art. it doesn’t mean he didn’t receive compensation, it just means he personally decided the video didn’t need a “Sponsored Content” warning or ControlAI’s logo on it.
-
@davidgerard for anyone unfamiliar with Yud: oh boy you don’t want it
in short, Yudkowsky invented the modern, AI doom cultist, capital-R form of Rationalism. without Yud and Soares, the TESCREAL belief package wouldn’t have the shape or the political power that it does now. these are core figures. Yud did decades of work exposing billionaires to the Rationalist and Effective Altruist ideas that he and Soares came up with.
the reason why the interview wasn’t with Yudkowsky is because his name is now poison, as a lot of people who listen to the guy quickly realize he’s a crank.
Green’s paid video with ControlAI led quickly to an interview series where he platformed one of the biggest TESCREAL figures in the subculture.
and note that when Green says the interviews were “unsponsored”, that’s a YouTube term of art. it doesn’t mean he didn’t receive compensation, it just means he personally decided the video didn’t need a “Sponsored Content” warning or ControlAI’s logo on it.
@davidgerard whether Green’s videos are marked as unsponsored because ControlAI paid him in one large lump sum with the understanding he’d do the rest of the AI videos, or it was an explicit part of a partnership contract, or Green’s a true believer fool who took a smaller sum for the sponsored video and bought into the belief package entirely, is invisible and of no concern to me
this is what corruption looks like when someone admits to it
it’s time to pay attention and reclaim the media literacy your education denied you
-
@davidgerard Green stated that he did this in order to present a variety of views on AI, but that’s not true. all of the interviews and AI videos on Green’s channel are in support of ControlAI’s AI doom narrative. there are no opposing viewpoints challenging the ridiculous and unscientific idea that godlike AI is a possible and impending risk to humanity.
all of the interviews Green has done have been with people who are either allied closely with ControlAI or are similarly tied up in Effective Altruism and AI doom (yes even Bernie, sorry about our boy).
the article is coy about this, but the “If Anyone Builds It, Everyone Dies” co-author Green interviewed was Nate Soares, a core TESCREAL personality who came up with the term “AI alignment”, now used in AI doom marketing by every EA organization.
the other co-author was Eliezer fucking Yudkowsky.
@zzt
Thank
you! Seeing everyone blow up about his use of genAI while this problematic barely being addressed. I get that one of those are simpler to wrap your head around, but the other is so much more damning.
Hank Green went from shrugging and going "Why do everyone hate AI?" to booster of AI doomerism.
Carl from Internet of Bugs addressed it, but I haven't really seen many others do so.
https://www.youtube.com/watch?v=4lKyNdZz3Vw -
@zzt
Thank you! Seeing everyone blow up about his use of genAI while this problematic barely being addressed. I get that one of those are simpler to wrap your head around, but the other is so much more damning.
Hank Green went from shrugging and going "Why do everyone hate AI?" to booster of AI doomerism.
Carl from Internet of Bugs addressed it, but I haven't really seen many others do so.
https://www.youtube.com/watch?v=4lKyNdZz3Vw@707Kat Carl’s video is excellent! it was my all-in-one source on this before Pivot covered it. I should give that video a link in the top level of this thread
-
@707Kat Carl’s video is excellent! it was my all-in-one source on this before Pivot covered it. I should give that video a link in the top level of this thread
@zzt I was gobsmacked when HG released both a Sci-show video with the ControlAI behind it (I went and personally checked up their ties to EA) and then his personal channel with the Yudkowski book interview within 1-2 weeks.
Not to speak of him running another EA point critiquing the critique of water usage of genAI Datacenters in another video.
I find it so astounding that resource abuse and current harms apparently "isn't that bad" while future theoretical hype b/s is worth platforming.
