Dude why do we even need or want AGI?

To solve biology and physics and live forever amongst the stars, obvs.

Or to allow a tiny elite to treat the rest of humanity like cattle since they no longer depend on them for physical and mental labour.

Nerding out about Dune is tremendously cool and i love it, you can pick any thread to pull and it always goes somewhere.

The details of the plot don’t necessarily map neatly, but the aesthetics, the fantasy, the torture-versus-dust-specksness of it is what the rationalists are all about.

Isn’t aspiring for the aesthetics while ignoring the (admittedly heavily lore driven and not especially applicable to irl) substance exactly what the torment nexus meme is about though? Except I don’t think there is a dust-speckness aspect to the GP, you aren’t future-human-population-maxxing¹, the point seems to be to ensure there is a future where humanity’s collective free will isn’t utterly tethered to a prescient autocrat², and the prescriptive aspect is that we should be part of an open system instead of say locked in with the great man of history du jour, which is also in keeping with the ecological framing.

It’s been a while but I don’t think the Golden Path is even that front and center in the text, Dune 4: GEOD is basically a character study on the God emperor, who is one of the most unique and fascinating characters in sci-fi³.

I think this is why the GP is a TINA situation by authorial decree, it’s Frank Herbert going listen, I’m doing my best to write a suicidal rebirth god archetype as a layered and relatable-yet-utterly-othered character, you are not supposed to be worrying that much if there could be a GP-but-liberal with more individual thriving and less oppressive totalitarianism, I assure you he’s thought about it extensively and he thinks there sure can’t and that’s it.

Have you read the Dosadi Experiment? It’s a very strange book and a good way to get a high dose of Herbert’s obsessions from a new perspective.

The Bureau of Saboteurs books along with Godmakers were my favourite non-Dune Herbert books! I also found the Dragon in the Sea fascinating 20 years ago and think I should revisit, and also finally read the follow up collaborations that only seem to be available unofficially. Also The White Plague gave late-teens me nightmares.

I remember Dosadi as being more about FH going all out on the intrigue and deep lore to the point where the latter parts of the book are basically written in innuendo, you are literally expected to read between the lines to understand what the hell is going on, I wish my parents had as much faith in me as Frank Herbert had in his readers.

But the more you scratch the more unambiguously evil he is.

The worst aspects of FH I’m aware of are that he was a shitty fucking parent and hated Iron Maiden, unambiguously evil seems stretching it, unless you were his son.

I think there even is a case to be made about how Frank Herbert is the anti-L. Ron Hubbard, using sci-fi literature as an efficient outlet for his psychedelics/mysticism/ecology/psychosexuality obsession oscilliation and actually leaving a descent literary legacy instead of starting a cult or several, but this post is already running so long it’s starting to need an editor.

1. sigh
2. This has additional metaphysical implications with how prophecy works in the duniverse, in the sense that it’s not only about being forever in complete submission to Empire in some traditional ultracolonialist sense, you are essentially reduced to just a part of another man’s waking dream as they pick and choose what future track to lock the timeline in.
3. And is also the first person narrator of the book. Both the remaining Dune books also switched to full first person point-of-view narration with very little in terms of an omniscient guiding perspective, it’s fairly ambitious actually.

the 4000 years of necessary tragedy and sacrifice, aka the golden path, is also remarkably similar to the rationalists’ teleology! specifically the longtermist stuff! i.e. it’s okay to sacrifice lives today to ensure humanity’s survival.

Only in the broadest sense where humanity survives into the far future by spreading so far and wide that no matter the scale of a catastrophe a significant part will always continue to thrive.

However, where longtermism is about papercliping the entire universe into compute to fulfil some vague utilitarian notion of virtual happiness quota, the GP seems to be more about crippling the substructure that ostensibly causes humanity again and again be reduced to the whims of some supreme authority, be it the automated thinking machines from their past or their current much harder to escape succession of psychic tyrants let loose by selectively breeding for something humanity had absolutely no natural defence against.

The GP isn’t even a utopia, it’s a response to an immediate incredibly out of the box problem, the inevitability of an eventual dynasty of space wizard genghis khans.

i think herbert’s feelings about breeding programs are … complicated. he never presents it as ethical but he does seem fascinated by the concept and he’s very very into the idea of forging supermen from extreme environmental pressure. like “comes up several times per book” levels of fascinated.

I think they come second to his concerns about ecology and humanity’s relation to the environment. Post-desert fremen are basically water-fat cosplayers, and in general, other than the deliberately paradigm shattering kwisatz haderach, the end product of genetic adjustment are never presented as an apex for humanity, more like a good fit for their niche, like how post-emperor fish-speakers either peter-out or get subsumed by other factions.

their philosophy aligns with the plot of dune

Only in the Torment Nexus sense. Like, in the original books it takes 4000 years of tragedy and sacrifice at an immense scale to somewhat unfuck the direct implications of said breeding program and ensure humanity’s continued existence.

edit: also the genetic engineering enthusiast faction/race are absolutely never presented in a good light.

overwhelmed/exhausted

In the blog he posted afterwards he says he has 40 years of experience, so he is also grandpa old.

Sometimes I think it’ll eventually come out that the inner circle rationalists already had a breeding program going by this time, deliberately mix and matching possibly oblivious parents according to desired characteristics like increased ability to mentally rotate shapes or an above average polymarket win/loss ratio, and there will be at least a few kids that will grow up having to deal with that shit.

Barely skimmed it, but it says stuff like

for the people saying things like “I’m a PhD from xyz uni and I’m telling your LLMs are just stochastic tools that make everything up and the world will fall apart if you use them”, I’m here to tell you that you are out of date. The world of software engineering has changed dramatically in the last few months.

which is sold-in-dumpster-alley quality copium, and also that version 3.5 is going to “raise the bar enormously with regards to rsync security”, so I guess we’re looking at a notably vibe coded security model for rsync going forward?

I don’t notice any obvious signs of slop-machine

The brainwrinkles look if you put a brain though a make-it-look-like-a-sketch prompt, it’s pointlessly detailed (i.e. not pretty or crafty, not adding anything, just lines for the sake of lines), it has the piss filter hue and also there’s random orange shading. The background seems to be at least two separate layers of seemingly disjointed stuff.

Supposedly the plot hinges somewhat on AI shenanigans so this painting being noticeably GenAI’ed could the part of its point. or even a meta thing, but playing the game to make sure isn’t currently on my agenda.

There’s a Basilisk reference in the new 007 game, I guess the robot devil is firmly mainstream now.

edit: the in-game painting looks hella AI generated too.

if all you want is sex you can simply go have sex with the people who want sex. Or just pay for sex, or use grindr or whatever.

Rationalists used to more openly stew in incel culture, a big part of which is that beyond sex you are also owed undivided love and attention, so there’s probably still a big deal of self-worth attached to that.

Incels who are fine with just paying for it call themselves MGTOW and are kind of a separate subculture that I feel has mostly petered out by now, probably because it’s harder to separate from regular old women-strictly-as-sex-objects type misogyny.

maybe I’ve just missed that part of their culture.

I wasn’t around but supposedly rationalists were hugely PUA-curious in the early days when entitled nerds commiserating about not getting laid were a major voice in the community, along with a bunch of other more out there incel stuff like bi-maxxing, i.e. trying to get it on with other severely undersexed dudes to scratch the itch.

A lot of it was summarily scrubbed when they started getting money and attention.

jqwik maintainer’s anti gen AI activism makes clanker crankers sad

From github thread:

I can’t actually believe someone would be so childish and put this nonsense into their repo.

Actively opposing hyper-scaled GenAI and agentic coding is an ethics-related decision. Those who have not followed the long-going discussion may want to start reading up here: https://blog.johanneslink.net/2025/11/04/to-gen-or-not-to-gen/

Thus, one can argue that my ethical judgement is wrong or based on wrong assumptions. One could also argue that the measures I decided to take come with more down-side than up-side. Calling it childish, however, reveals IMO that the accuser has not seriously thought about the topic.

Seems like run-of-the-mill yada-yada boosterism for the most part.

Points for managing to write an entire blog post without lamenting how the persecution of nerds is a historical level atrocity, much unlike the ongoing state-sponsored genocide by Israel which is very fine and very justifiable, I guess.

LessWrong, an online community that started out as a place to discuss Yudkowski’s writings, who among other things is the If You Build It Everyone Dies guy.

LessWrong also directly gave us Effective Altruism and the Zizians.

While looking at ACX comments for the you should let claude vote for you thing I saw someone saying that the lumina guy (gobble designer microbes instead of brushing your teeth, boosted by siskind and aella who got free samples) has apparently pivoted to AI with a startup about producing AI generated literature around positive human-AI interactions to influence future generations of LLMs towards favorable alignment.

I think the later got mentioned here some time or other but I didn’t realize it was also the teeth bacteria successor grift.

Well now you’ve made me look at the image too long, too.

#1’s belt probably wraps twice around that waistline, I’d say photoshop/AI but there’s that 1/3 of the shirt hanging off to the right with nothing to fill it so yeah, and also Phil Davis¹ exists.

#10 is aspiring to be the jock from scooby doo and I don’t think that jacket can actually close, similarly to #1 the whole thing probably only works for very specific poses and angles.

#6 looks like he runs a leopard print handbag delivery service, #4 is horny youth pastor in his aunt’s kitchen, 7-9 look comfy but contrived, #5 really wants you to know he’s wearing a possibly expensive wife beater under that shirt and 2 & 3 are all sorts of dodgy, esp. #3 who looks like a tourist trap gift shop keeper whose face is noticeably untanned because he has to hide it a lot for undisclosed reasons.

#2 mostly has a raging case of resting netflix true crime serial killer face syndrome.

I get that the point probably was to only show the broad strokes of how it should look if you’re making an effort, but whatever.

I don’t know Guillaume Verdon / Beff Jezos

He is supposed to be the original instigator and public face of so-called e/acc or effective accelarationism, i.e. the rationalist (or maybe rationalist inspired idk) spin-off of people who feel whinging about alignment isn’t necessary and that it’s in fact awesome when new technology is killing people and ruining the environment because it means we’re getting to the singularity faster, and also openly rooting for fascism on main is based now.

His magic heatless AI processor startup feels like a grift to fleece investors and as far as I know has only ever produced an obviously staged video of sciency looking individuals fawning over a 3d printed wire mesh while touring what looks like a chip fab with an extremely lax contamination protocol.

I do not know if that can work because only a few percent of the company will be for sale

I think a 30% chunk will be available for non-institutional investors, which as far as these things go is supposed to be humongous and also an indication that they are very much counting on stupid money to prop their valuation.

It’s more like license to sue their pants off if they get caught propagating obviously proprietary code through the responses of their tool, and if they are doing it but you can’t tell that just means your enterprise code isn’t discernible from claudeslop so no harm done.

I’m assuming if suddenly an LLM code tool is able to do something like write a parser for an unambiguously closed source heavily copyrighted data format and the only possible leak is the devs using LLM tooling, it’s going to be a big legal deal.

Microsoft releases cost calculator for GitHub Copilot for the new token usage based billing. Previously you were being charged per request, kind of like hiring a cab and paying the same whether you went to the next corner or the next continent.

Turns out Zitron may have been seriously low balling the actual cost to subsidized billing ratio.