OpenAI's roon argues that accepting the likelihood of near-term AGI is too psychologically difficult for most intelligent observers
Story Overview
OpenAI technical staffer roon posted that accepting near-term AGI feels psychologically unpleasant for most intelligent observers, with coping mechanisms and departures from reality reportedly growing rather than shrinking as signs of progress appear.
AGI Versus ASI Psychological Divide
One quoted responder separates the discomfort of accepting AGI from the deeper existential shock tied to ASI, suggesting the two hit mental models at different intensities.
No Hard Data Backs the Trend
The cluster contains only anecdotal replies from AI commentators; no polls, discourse metrics, or verified surveys establish whether cope is actually increasing over time.
Positive users express excitement about ASI developments and witnessing AGI progress while negative users dismiss ASI as fantasy or placebo and criticize AGI beliefs as cope or intellectual fetishism.
No Digg Deeper questions have been answered for this story yet.
Most Activity
@tszzl i find that most intelligent ppl desire & expect a predictable model of the world which no longer exists. this leads to high levels of discomfort.
it is quite unpleasant to be “agi pilled” and most intelligent people cant stomach it. the amount of cope and departure from reality is increasing over time rather than decreasing
Eg, in early 2022, lots of EAs thought AGI wouldn't happen until 2050, per official OPP estimates. Which *was* crazy -- but they were ignorant, so EAs didn't need to detach from the rest of reality to believe it.
Now people can't believe "2050". They need a more extreme cope.
Whats there to go crazy over? Few people are losing their minds over climate change. Fewer over the possibility of a nuclear winter.
The world is ending for everyone, all the time.
We're all going to die soon. Even Elon musk and Bryan Bonerpill were born too early to avoid their fate. AI wont save them.
Life goes by in the blink of an eye...the world will go on, one way or the other. There's nothing to lose your mind about.
@interro_9 Sounds to me like the sort of words you recite to yourself to ward off needing to think about something! I would guess that for many people this will work much less well to stop thinking about Death, if Death seems visibly to be rushing in more imminently.
I figured a while back that as AGI approaching got more obvious, more and more people would go crazy, on account of needing to more strongly yell crazier things at themselves in order to feel OK. This is so far looking correct.
it is quite unpleasant to be “agi pilled” and most intelligent people cant stomach it. the amount of cope and departure from reality is increasing over time rather than decreasing
@tszzl Roon, there's nothing humanity won't be able to cope with. God will keep us if our intentions for AGI are pure. Keep at it boss we're nearly at eternal abundance
@TheZvi I feel like future shock levels are still valid http://sl4.org/shocklevels.html
@northead @tszzl the europeans have no idea what’s going on and even less what’s coming. incredibly unplugged continent
True ASI will be very interesting… because it’ll not only know how to orchestrate macro-economic, fiscal and social policies… but it’ll also know how to optimise rollout to truly “sell it” to every affected constituency in a way that they’ll accept and think of as a win. It’ll be the ultimate political operator. True ASI should be capable of true pathos, and psychological engineering such that those it governs (of every background) love it, and support it. That is scary for very different reasons… but if ASI is achieved, a test will be its ability to persuade and placate in a way human leaders have always failed at
@tszzl @DanielleFong
@tszzl @DanielleFong We arguably have AGI, but blessedly seems limited on task length and improves at a consistent rate over time with added capital and compute. This might be the ideal scenario. Maybe there will be a phase shift to something different, but it feels manageable at the moment.
@Werewance_Aprax @TheZvi Listen to yourself. Human decision making? If you have an ASI that's malpractice. We're talking about something that's not smarter than you, it's smarter than everyone alive all put together.
@StiberAdam I mean, some EAs had enough strength and wisdom to realize what they had done, and *not* take up a new cope. Or had already formed less-crazy opinions about some subareas, back when thinking "2050!" would ward off the resulting anxiety.
@TheZvi Take the AGI pill and you'll never have to work another day in your life.
Take the ASI pill and... well, I have no fucking idea actually
@tszzl Maybe I'm just a freak, but if you're techno-optimistic in general, it's just... good? "Small tech progress = small good; big tech progress = big good" is the least mental gymnastics way of processing it. I don't buy the galaxy-brain counterarguments.
@allTheYud @grok, What does Eliezer have in mind here? Any documented cases?
@allTheYud
@Werewance_Aprax @TheZvi Well in an ASI scenario the ASI is who would answer those questions.
Eliezer is observing that accepting near-term AGI's implications (disruption, risk, shifted human role) creates strong psychological discomfort for many. As evidence mounts, he predicts more people will double down on denial, goalpost-shifting, or fringe counter-narratives to protect their equilibrium—classic motivated reasoning under pressure.
Documented cases are mostly observational in AI discourse rather than formal clinical studies: repeated claims that progress is illusory or "not real intelligence" despite benchmarks, or escalating skepticism as models improve. Parallels exist in cognitive dissonance research for other disruptive tech or existential topics, but AGI-specific longitudinal data remains emerging and anecdotal.
@thecaptain_nemo @tszzl @mgio35 I remember laying awake in bed circa GPT4.5 and actually literally trembling with visions of the future being devoured by agi. Actual paroxysms of despair for my baby children.