So seeing the reaction on lesswrong to Eliezer's book has been interesting. It turns out, even among people that already mostly agree with him, a lot of them were hoping he would make their case better than he has (either because they aren't as convinced as him, or they are, but were hoping for something more palatable to the general public).
This review (lesswrong discussion here), calls out a really obvious issue: Eliezer's AI doom story was formed before Deep Learning took off, and in fact was mostly focusing on more GOFAI than neural networks, yet somehow, the details of the story haven't changed at all. The reviewer is a rationalist that still believes in AI doom, so I wouldn't give her too much credit, but she does note this is a major discrepancy from someone that espouses a philosophy that (nominally) features a lot of updating your beliefs in response to evidence. The reviewer also notes that "it should be illegal to own more than eight of the most powerful GPUs available in 2024 without international monitoring" is kind of unworkable.
This reviewer liked the book more than they expected to, because Eliezer and Nate Soares gets some details of the AI doom lore closer to the reviewer's current favored headcanon. The reviewer does complain that maybe weird and condescending parables aren't the best outreach strategy!
This reviewer has written their own AI doom explainer which they think is better! From their limited description, I kind of agree, because it sounds like the focus on current real world scenarios and harms (and extrapolate them to doom). But again, I wouldn't give them too much credit, it sounds like they don't understand why existential doom is actually promoted (as a distraction and source of crit-hype). They also note the 8 GPUs thing is batshit.
Overall, it sounds like lesswrongers view the book as an improvement to the sprawling mess of arguments in the sequences (and scattered across other places like Arbital), but still not as well structured as they could be or stylistically quite right for a normy audience (i.e. the condescending parables and diversions into unrelated science-y topics). And some are worried that Nate and Eliezer's focus on an unworkable strategy (shut it all down, 8 GPU max!) with no intermediate steps or goals or options might not be the best.
You thought a right-and-proper Communist Five-Year Plan couldn't also be a self-insert fanfic? Hold Yud's beer
did you hear about Yudoslavia? They have this Eight GPU policy, if a household is about to have another GPU and they find it can’t run Crysis at max settings at 60fps, they leave it outside for the wolves. But this might just be Yudophobic propaganda
In Eliezer's "utopian" worldbuilding fiction concept, dath ilan, they erased their entire history just to cover up the any mention of any concept that might inspire someone to think of "superintelligence" (and as an added bonus purge other wrong-think concepts). The ~~Philosopher Kings~~ Keepers have also discouraged investment and improvement in computers (because somehow, despite now holding any direct power and the massive financial incentives and dath ilan being described as capitalist and libertarian, the Keepers can just sort of say their internal secret prediction market predicts bad vibes from improving computers too much and everyone falls in line). According to several worldbuiding posts, dath ilan has built an entire secret city that gets funded with 2% of the entire world's GDP to solve AI safety in utter secrecy.
tell me you didn't subject yourself to glowfic
glowfic: it's like a forum, but worse(tm)
Yep, from what I can tell second hand dath ilan world building definitely skews towards it doesn't count as totalitarianism if the enforced orthodoxy is in line with my Obviously Objectively Correct and Overdetermined opinions.
It is funny that this is the best Yud, with an IQ of 191, can come up with.
Thanks for the lore, and sorry that you had to ingest all that at some point.
Like a lot of sci-fi that I will comment on without having read, that sounds like a Dune ripoff (that I also have not read). Except, rather than "ok we tried the whole AI thing and it turned out bad", Yud is saying "my world is smarter because they predicted AI would be bad, so they just didn't try it, neener neener neener!"
Also, iterating on the name origins of "dath ilan," as I've said before, it's an anagram of Thailand. But I have a new hypothesis. It's formed from "Death Island" minus the letters "ESD". What could this mean?!?!?! Electrostatic discharge? "Eek, Small Dragon?" "Eliezer Sudden Death?" The abyss pondering continues...
OG Dune actually had some complex and layered stuff to say about AI before the background lore was retconned to dollar store WH40K by the current handlers of the IP.
There was no superintelligence, thinking machines were gatekept by specialists who formed entrenched elites, overreliance to them was causing widespread intellectual stagnation, and people were becoming content with letting unknowable algorithms decide on matters of life and death.
The Butlerian Jihad was first and foremost a cultural revolution.
Ironically, one of the biggest lore drops about dath ilan happens in a story I initially thought at the time was a parody of rationalists and the concept of dath ilan (Eliezer used a new penname for the story). The main dath ilan character (isekai'd into an Earth mostly similar to our own but with magic and uh.. other worldbuilding conceits I won't get into here) jumps to absurd wild conclusion throughout basically every moment of the story, and unlike HJPEV is actually wrong about basically every conclusion she jumps to. Of course, she's a woman, and it comes up towards the ending that she is below average for dath ilan intelligence (but still above the Earth average, obviously), so don't give Eliezer too much credit for allowing a rationalist character to be mostly wrong for once.
I don't know how he came up with the name... other fanfic writers in rationalist-adjacent space have complained about his amateurish attempts at conlanging, so there probably isn't a sophisticated conlang explanation about phonemes involved. You might be on the right track guessing at weird anagrams?
My invite-only internal prediction market is offering good odds that this is referring to either math pets or tentacles, and a solid parlay opportunity on it being both combined
I can smell the 'rape (play) is the best kind of sex actually' from over here.
Stop trying to make dath ilan happen, Yud.
I didn’t what this meant initially. “HJP” pointed me toward “harry james potter” but I couldn’t figure out “EV,” so ofc my gutter brain decided this was “Handjob Penis Envy Vagina”. But no, it’s just “HJP Evans Verres”, Verres being HJPEV’s adopted dad.
He has a credit score of 250 from this (imagine me holding two thumbs up at myself) ratings agency.
Yeah his naming is wack. I’m guessing that Yud went with “Verres” for HJP’s nonbiodad because it’s close to “veritas”, the latin word for truth. But searching “verres” gives the latin word for boar, lol. And also Gaius Verres, a notoriously bad magistrate according to wikipedia.