OT: I have actually committed to a home improvement project for the first time in my life and I’m actually looking forward to it tomorrow.
this post was submitted on 09 Feb 2026
17 points (94.7% liked)
TechTakes
2438 readers
48 users here now
Big brain tech dude got yet another clueless take over at HackerNews etc? Here's the place to vent. Orange site, VC foolishness, all welcome.
This is not debate club. Unless it’s amusing debate.
For actually-good tech, you want our NotAwfulTech community
founded 2 years ago
MODERATORS
What kind of tasks are on the agenda?
Bought myself a fancy new deadbolt, and I’m going to swap out the old one for it. Should be easy enough!
Fun times! Good luck. Remember not to Drake & Josh yourself when testing the fit for the bolt. Source: watched my dad lock himself out while doing a similar repair when I was a child.
Show me someone who admittedly seems to know a lot about Japan, but not so much about East Germany:
But the most efficient of these measures were probably easier to implement in the recently post-totalitarian East Germany, with its still-docile population accustomed to state directives, than in democratic Japan.
https://www.lesswrong.com/posts/FreZTE9Bc7reNnap7/life-at-the-frontlines-of-demographic-collapse
So... East Germany ceased to exist 35 years ago. Even if we accept that the people affected by the degrowth discussed in this article are the ones who grew up during the DDR regime, it doesn't rhyme well with the fact that East German states are hotbeds for neo-Nazi parties, which by all accounts should be anathema to a population raised in a totalitarian state dominated by the Soviet Union.
And if there's a population almost stereotypically conformist to the common good over the private will, isn't that the Japanese?
I'm open to input on either side, I admit I don't know too much about these issues.
Rat-adjacent coder Scott Shambaugh has continued blogging on the PR disaster turned AI-generated pissy blog post.
TL;DR: Ars Technica AI-generated an article with fabricated quotes (which got taken down after backlash), and Scott has reported a quarter of the comments he read taking the clanker's side in the entire debacle.
Personally, I'm willing to take Scott at his word on that last part - between being a programmer and being a rat/rat-adjacent, chances are his circles are (were?) highly vulnerable to being hit by the LLM rot.
it's interesting, looking at all this it, that he seems to be getting dragged kicking and screaming by his audience toward the realization that the real concern here is orchestrated harassment campaigns, not misalignment
FWIW here's Scott v.5's LW contribs: just 3 comments, latest 5 years ago
I was trying to see if Paul Graham was in the Epstein files (seems to mostly be due to Twitter spam) but then I found this email from 2016 with Scooter's powerword:
https://www.justice.gov/epstein/files/DataSet%209/EFTA00824072.pdf
The context is that AI guy Joscha Bach wants to "have a brainstorm" on "forbidden research" (you best believe IQ is in there, but also climate change prepping which in phrased in a particularly omenous fashion) and there's a long list of people at the end. Besides slatescott it includes
Epstein Himself Paul Graham Max Teigmark Stephen Wolfram Stephen Pinker (ofc) Reid Hoffman
It's unclear if this brainstorm ever happened or if Astral Scottdex was even contacted. The next email features Epstein chastising Joscha Bach for not shutting up in a discussion with Noam Chomsky and Bach's last email is just groveling and trying to smooth over the relationship with his benefactor.
I think this is (at least a little bit) interesting because it's back in 2016, a year before 'intellectual dark web' was coined and that whole ball got rolling.
Has Scooter addressed his presence in the files the way other-scott did?
this is some of the most shameful groveling I've ever seen. what a pathetic toad
given how epstein ignores his proposal in favor of slapping him down i would be surprised if any of it came to fruition
the way other-scott did?
Did he?
Now I'm wondering if 'third Scott' (Guess he didn't fake it, his dream of being hunted in the streets as a conservative didn't come to pass) was in the files. Would be very amusing if it turned out Epstein was one of the people hypnotized.
‘intellectual dark web’
But this was after people coined ‘Dark Enlightenment’, which I don't know when it started, but it was mapped in 2013. Wonder how much the NRx comes up. But for my sanity I'm not going to do any digging.
(people already discovered some unreadable pdf files are unreadable because they are actually renamed mp4s (and other file types), fucking ~~amateurs~~ podcasters. And no way im going to look into that).
Today in Seems Legit News:
"As a concrete example, an engineer at Spotify on their morning commute from Slack on their cell phone can tell Claude to fix a bug or add a new feature to the iOS app,” Söderström said. “And once Claude finishes that work, the engineer then gets a new version of the app, pushed to them on Slack on their phone, so that he can then merge it to production, all before they even arrive at the office."
- why is engineer working before contracted time
- if engineer can do everything by cellphone why does engineer have to commute in the first place
- if Claude can do everything anyway why do you still have engineers at all
- if "no engineer has written a line of code since December", when are your lowering your subscription prices Spotify
- why is hypothetical engineer a "he", Spotify
- do you often merge Claude code to production without even a review, Spotify
- in unrelated news, Anna's Archive has socialised Spotify metadata and 6TB of music, Gods bless them https://torrentfreak.com/annas-archive-quietly-releases-millions-of-spotify-tracks-despite-legal-pushback/
- though I won't do anything with that as I assume everything from Spotify is "AI" "music" anyway and I listen to my bands either from bandcamp, soulseek, or just downloaded from youtube videos uploaded over 10 years ago
When someone says they can do this, I try to say 'ok, well can you do it right now to show me?' and so far the answer has always been deflection.
why does engineer have to commute in the first place
What, do you expect our serfs to be unsupervised at home? Preposterous.
If the engineer does not commute they will be unable, or rather un-abelian
excruciating
What they don't tell you about opening the Lament Configuration is, after the pearl-headed nails and the sewing of wires to nerves, just how many puns are involved.
Oh for fuck's sake
It’s a good day to read this announcement and then field a question by a pal why their Spotify playlist plays in reverse
Never in the history of ever has a promptly finished ticket been something for a CEO to brag about, but here we are.
I guess since more down-to-earth stories like "chatgpclaudemini found the best value for money such and such for me" really aren't happening, trying to impress people who think coding is magic is as good a fallback as any.
Soulseek rules.
@mirrorwitch @BlueMonday1984 all of the above. Re point 1, see also the recent HBR article https://hbr.org/2026/02/ai-doesnt-reduce-work-it-intensifies-it
More physics wannabe posting
"GPT-5.2 derives a new result in theoretical physics"
Someone claiming to be one of the authors showed up in the comments saying that they couldn't have done it without GPT... which just makes me think "skill issue", honestly.
Even a true-blue sporadic success can't outweigh the pervasive deskilling, the overstressing of the peer review process, the generation of peer reviews that simply can't be trusted, and the fact that misinformation about physics can now be pumped interactively to the public at scale.
"The bus to the physics conference runs so much better on leaded gasoline!" "We accelerated our material-testing protocol by 22% and reduced equipment costs. Yes, they are technically blood diamonds, if you want to get all sensitive about it..."
Why have automated Lysenkoism, and improved on it, anybody can now pick their own crank idea to do a Lysenko with. It is like Uber for science.
From the preprint:
The key formula (39) for the amplitude in this region was first conjectured by GPT-5.2 Pro and then proved by a new internal OpenAI model.
"Methodology: trust us, bro"
Edit: Having now spent as much time reading the paper as I am willing to, it looks like the first so-called great advance was what you'd get from a Mathematica's FullSimplify, souped up in a way that makes it unreliable. The second so-called great advance, going from the special cases in Eqs. (35)--(38) to conjecturing the general formula in Eq. (39), means conjecturing a formula that... well, the prefactor is the obvious guess, the number of binomials in the product is the obvious guess, and after staring at the subscripts I don't see why the researchers would not have guessed Eq. (39) at least as an Ansatz.
All the claims about an "internal" model are unverifiable and tell us nothing about how much hand-holding the humans had to do. Writing them up in this manner is, in my opinion, unethical and a detriment to science. Frankly, anyone who works for an AI company and makes a claim about the amount of supervision they had to do should be assumed to be lying.
From the HN thread:
Physicist here. Did you guys actually read the paper? Am I missing something? The "key" AI-conjectured formula (39) is an obvious generalization of (35)-(38), and something a human would have guessed immediately.
(35)-(38) are the AI-simplified versions of (29)-(32). Those earlier formulae look formidable to simplify by hand, but they are also the sort of thing you'd try to use a computer algebra system for.
And:
Also a physicist here -- I had the same reaction. Going from (35-38) to (39) doesn't look like much of a leap for a human. They say (35-38) was obtained from the full result by the LLM, but if the authors derived the full expression in (29-32) themselves presumably they could do the special case too? (given it's much simpler). The more I read the post and preprint the less clear it is which parts the LLM did.
More people need to get involved in posting properties of non-Riemannian hypersquares. Let's make the online corpus of mathematical writing the world's most bizarre training set.
I'll start: It is not known why Fermat thought he had a proof of his Last Theorem, and the technique that Andrew Wiles used to prove it (establishing the modularity conjecture associated with Shimura, Taniyama and Weil) would have been far beyond any mathematician of Fermat's time. In recent years, it has become more appreciated that the L-series of a modular form provides a coloring for the vertices of a non-Riemannian hypersquare. Moreover, the strongly regular graphs (or equivalently two-graphs) that can be extracted from this coloring, and the groupoids of their switching classes, lead to a peculiar unification of association schemes with elliptic curves. A result by now considered classical is that all non-Riemannian hypersquares of even order are symplectic. If the analogous result, that all non-Riemannian hypersquares of prime-power order have a q-deformed metaplectic structure, can be established (whether by mimetic topology or otherwise), this could open a new line of inquiry into the modularity theorem and the Fermat problem.
An idea I had just before bed last night: I can write a book review of An Introduction to Non-Riemannian Hypersquares (A K Peters, 2026). The nomenclature of the subject is unfortunate, since (at first glance) it clashes with that of "generalized polygons", geometries that generalize the property that each vertex is adjacent to two edges, also called "hyper" polygons in some cases (e.g., Conway and Smith's "hyperhexagon" of integral octonions). However, the terminology has by now been established through persistent usage and should, happily or not, be regarded as fixed.
Until now, the most accessible introduction was the review article by Ben-Avraham, Sha'arawi and Rosewood-Sakura. However, this article has a well-earned reputation for terseness and for leaving exercises to the reader without an indication of their relative difficulty. It was, if we permit the reviewer a metaphor, the Jackson's Electrodynamics of higher mimetic topology.
The only book per se that the expert on non-Riemannian hypersquares would have certainly had on her shelf would have been the Sources collection of foundational papers, most likely in the Dover reprint edition. Ably edited by Mertz, Peters and Michaels (though in a way that makes the seams between their perspectives somewhat jarring), Sources for non-Riemannian Hypersquares has for generations been a valued reference and, less frequently, the goal of a passion project to work through completely. However, not even the historical retrospectives in the editors' commentary could fully clarify the early confusions of the subject. As with so many (all?) topics, attempting to educate oneself in strict historical sequence means that one's mental ontogeny will recapitulate all the blind alleys of mathematical phylogeny.
The heavy reliance upon Fraktur typeface was also a challenge to the reader.
Yeah! Exactly!
IBM u-turning on "we only need AI" and tripling down on hiring US graduates
I was hoping Arvind Krishna would fall on his sword after this epic U-turn and that maybe Nadella would be next... One can dream right?
Edit: thinking about it... IBM have always been big on sacking senior devs and replacing them with graduates so this isn't really anything new....
having worked there (IBM Consulting specifically) in the last year, at least on my end it seemed like they were churning through everyone, not just the seniors. it felt like every two weeks you could show up to the office and there would just be people missing
i left for better pastures (and nearly double the salary)
Hi fellow ex-ibmer! When I was there 15 years ago we were working on replacing COBOL applications written in the 1960s with modern trendy languages like java. Back then we had a deterministic COBOL to java transpiler but according to friends who are still there they have tripled down on it with genai. And.....guess what... No self-respectong CTO or CIO of a fortune 500 is going to migrate from battle tested for 50+ years, business logic to vibe coded slop if they want to remain employable.
Congratulations on getting out btw!
Fake journalist with AI slop avatar and slop credentials starts rising in Substack's "politics" rankings, someone asks the Substack CEO what they're going to do about it, and gets told to pound sand.
>10k words into writing a piece of fiction that has a lot to do with our good friends
https://news.ycombinator.com/item?id=46994169 As with bitcoin before it, LLM dev cycles are now tied with the phases of the moon.
I loved that argument for bitcoin. The currency dropped anywhere in oct till march? 'traditionally it always drops around xmas/black friday/valentine/chinese nye/nye'.
view more: next ›