Want to wade into the sandy surf of the abyss? Have a sneer percolating in your system but not enough time/energy to make a whole post about it? Go forth and be mid.
Welcome to the Stubsack, your first port of call for learning fresh Awful you’ll near-instantly regret.
Any awful.systems sub may be subsneered in this subthread, techtakes or no.
If your sneer seems higher quality than you thought, feel free to cut’n’paste it into its own post — there’s no quota for posting and the bar really isn’t that high.
The post Xitter web has spawned so many “esoteric” right wing freaks, but there’s no appropriate sneer-space for them. I’m talking redscare-ish, reality challenged “culture critics” who write about everything but understand nothing. I’m talking about reply-guys who make the same 6 tweets about the same 3 subjects. They’re inescapable at this point, yet I don’t see them mocked (as much as they should be)
Like, there was one dude a while back who insisted that women couldn’t be surgeons because they didn’t believe in the moon or in stars? I think each and every one of these guys is uniquely fucked up and if I can’t escape them, I would love to sneer at them.
(Credit and/or blame to David Gerard for starting this.)
https://russwilcoxdata.substack.com/p/and-the-alignment-problem-what-chinas
In June 2025, Zhao Tingyang gave a talk at Tsinghua’s Fangtang Forum. The edited transcript ran in The Paper on July 4 under the title “人工智能的伦理与思维之限” (The Ethical and Thinking Limits of AI). Near the end, Zhao wrote this:
“What requires more reflection is that attempting to ‘align’ AI with human nature and values actually contains a risk of human species suicide. Human nature is selfish, greedy, and cruel. Humans are the most dangerous biological species. Almost all religions demand the restraint of human desire; this is no accident. AI aligned with human values may well become a dangerous subject by imitating humans. Originally, AI does not possess the selfish genes of carbon-based life, so AI is actually closer to the legendary ‘human nature is fundamentally good’ kind of existence, whereas human nature is not ‘fundamentally good.’” The alignment paradigm treats human values as the target AI should conform to. Zhao is arguing the target is the danger. An AI aligned to human values inherits the specific features of human judgment that Zhao says have produced the record of human harm. The paradigm is not incomplete. It is pointed the wrong way.
Zhao’s argument has developed across CASS, The Paper, and Wenhua Zongheng from late 2022 through 2025, from a provocative aside into a sustained critique of the alignment paradigm. In the same period, the English-language alignment and AI ethics literature produced no substantive engagement. No citations. No rebuttal. No naming. Zhao is a member of the Chinese Academy of Social Sciences Institute of Philosophy, author of the Tianxia framework, and one of the most cited philosophers working in Chinese today.
I need to think on this a little more, wasn’t on my radar.
In the same period, the English-language alignment and AI ethics literature produced no substantive engagement. No citations. No rebuttal.
Wow it’s almost like alignment and AI ethics studies is less a serious academic field and more like a prank capital likes to play on consumers.
But I also think Zhao Tingyang’s take that alignment will make AI evil because people are evil falls too much into the the-people-deserve-to-be-disempowered totalitarian state funny business side of things to be especially influential down these parts.
I asked someone from the mainland, she more or less agreed with you:
This is basically consistent with the long-standing logic of the Chinese internet: technology brings discursive power, and to give it away is to give away discursive power. AI is especially so.
To be fair, while I’m not familiar with the discourse in China I know a lot people consider (rightly) “alignment” as a framing to be a red flag for cranks and rats. It’s not that surprising that this attitude hasn’t been getting much recognition when the marketing departments of ai companies has been more engaged on that subject than serious academics.
Habryka doesn’t have time to write all the crazy shit he’s mulling on, so he offers a summary.
https://www.lesswrong.com/posts/MqgwHJ93pJpaeHXs6/posts-i-don-t-have-time-to-write
Do you enjoy living in a society that takes fire safety seriously? Sucks to be you, I guess:
- Fire codes are the root of all evil
How about we just make all the mosquito nets flammable. That’s effective altruism!
Also Switzerland is a libertarian paradise apparently.
Fire codes are the root of all evil
Ah somebody got told by their landlord not to do something. (I remember our student housing landlord (a big org) was regularly claiming ‘fire codes’ as an excuse to get rid of stuff in semi public areas. The actual fire codes didn’t demand this btw, it was just the excuse they used to stop students from filling everything with random trash).
Unfortunately there are a few things that make courts pretty tricky to implement in practice for things like the rationality, AI safety and EA communities. Badly implemented courts also can just make things worse by creating a clear target for attack and pressure. Seems very tricky, but probably we should have more courts (or maybe not, I would need to write the post to figure it out).
yesss yesss lesswrong people’s tribunals and struggle sessions let’s gooooooooo
in b4 “Committee for AI Safety” seizes control and executes people who are too smart with a guillotine.
They develop a special g-meter to find people who could potentially create more efficient gpus and send them to the gulags.
Is that what “g” has stood for all this time?
We’ll hide them at the g spot. LW would never find them there.
I was trying to combine the scientology e-meter stuff with the iq race science g-factor stuff.
But despite well-documented claims to genius IQs, somehow the billionaire set ends up not on the chopping block.
might LWers be the real Pol Potists? Read on to find out!
They sure seem to enjoy the idea of a Year Zero
I mean, it’s a banger of a song.
I think building codes and zoning reform are good topics to get into in rich English-speaking countries but you have to 1) learn from actual experts not x.com/wiseAss1488, and 2) engage in local politics and policy and not just post to nerds around the world.
The fire code thing really is an excellent example of LessWrong Brain. Fire truck drivers insist on needlessly large trucks (no citation) which makes roads 30% wider than they would otherwise be (no citation) which has “probably” “non-trivially” contributed to larger cars (no citation) leading to enough additional road fatalities to cancel out the lives saved by stricter fire codes (no citation).
The LessWrong Brain argument starts with a deliberately contrarian conclusion and proves it with a Rube Goldberg chain of logical syllogisms. Of course, citations are strictly optional, and they are free to misinterpret them as they see fit. The only real standard of each claim is “looks good to me”, but you are supposed to be impressed that they managed to string a dozen of them together to reveal some shocking, deep truth of the world that nobody else knows about. The AI 2027 nonsense is an infamous example of this.
He uses the word “fermi” which is cult jargon based on Fermi estimation, a.k.a. guessing shit with back-of-the-envelope calculations. Not exactly what you want if you want to convince people to reform fire codes, especially if you have zero citations for anything.
I guess people just aren’t rational enough, and the only reason the fire codes are so irrational is because people are emotional about fire codes. Firefighters are apparently revered as heroes, when it is the LWers who should be the heroes. After all, firefighters merely save people from fires, while LWers buy multimillion dollar mansions to talk about saving quadrillions of hypothetical people from hypothetical basilisks!
rationalism is when i pull five numbers out of my ass and multiply them together
Yeah but never pull 9 numbers out of your ass, that would make you too smart and they will tell the gov to drone strike you.
it’s what they do instead of prayer
Looks like Mythos didn’t catch this one:
Anthropic secretly installs spyware when you install Claude Desktop
Whoopsie!
Well it does have the secret “Any attempt to arrest a senior officer of OCP results in shutdown” derective.
It’s fine, spyware is only a risk when it’s bad people’s spyware. It’s totally fine when it’s Anthropic™-approved spyware!
As for Mythos catching things, maybe they should have used Mythos on their very own Claude Code considering that it has hilariously obvious security exploits, such as this one which inserts an arbitrary string into a shell command. Actually, never mind I don’t see anything wrong here, maybe we should burn another $20k in electricity running Mythos on it again to find out.
anthropic is the most moral ai company in the universe
Habryka @ LW takes a break from reinventing Western civilization from first Rat principles to advocate a RETVRN to incandescent lighting:
Eventually, in most of the western world outside of the US, incandescent lightbulbs were literally banned to promote energy saving policies.
This was the greatest uglification in history. Within two decades, much of the world that was previously filled with beautiful natural-feeling light started feeling alien, slightly off, and uncomfortable, and societal stigma around energy-saving policies prevented people from really doing anything about it.
jesus fucking christ this is insufferable, these guys have been masking off for so long now that they’re scraping the bottom of the barrel to find new things to do fascist signalling about
This was the greatest uglification in history.Seems like a market opportunity for some special lights/glass that recreates the natural feeling light they want. (I have some lower light leds in older style lamps and I’m having, for me, nice and cosy lighting. Only issue seems to be me getting older and my eyes getting worse with age.
Who TF was using LED lights for indoor lighting in the 1990s? Compact flourescents were the lightbulb replacement in the oughties.
And how TF do you write that post without using the phrase “sensory sensitivity” and citing some women who know they have autism? Once you know you are more sensitive to your environment than allistics, you can start to experiment with interventions.
Dramatic fascistic “RETVRN” language and focus on aesthetics aside, my wife and I actually dug into some of this lighting quality stuff a while ago and while our very good friend here does a poor job explaining it there is a definite difference in normal LEDs vs incandescent or natural light. The LED spectra is fascinating - big spikes at a couple wavelengths and nothing in between. In my experience with switching to the fancier high-CRI LEDs the difference is pretty minimal. Feels like a possible case where you don’t notice it, but your brain does. For my wife it seems to have helped reduce the incidence and severity of her crippling migraines, which is obviously more impactful. I don’t think I’d say it beautified the space or brought us back to the halcyon days of our glorious past, but that’s been huge for us all the same. The plural of anecdote is not cliche, but there’s not nothing here.
What these people do is they take something that’s vaguely true or at least has a kernel of truth in it and they give it the most sensationalistic and negative spin possible. So it can’t be “there were upsides and downsides to the switchover from incandescent bulbs”, it has to be “the greatest uglification in history”, and anyone who points out the energy saving advantages of LEDs and fluorescent lightbulbs is just trying to Silence You and smear you. It’s so wretched.
guys
Agentic Ripz is my new jam band.
I’m sorry, I think I need to believe that this is taking the piss in order to be able to function. It can’t be real (It’'s definitely real).
Oh God I read their FAQ and it looks like the whole concept is to gamify smoking weed because if there’s one problem with weed it’s that it’s not addictive enough on its own. I mean the actual concept is to try and smash enough hip tech buzzwords together to extract some amount of the dwindling venture capital continuing to slosh around the valley, but if it actually happens the thing it’s going to do is take all the addictionware tactics that app developers have developed and bring them to bear on promoting drug use.
This is a serious blow to coolness, from which not even drugs will easily recover.
Wake me up when there’s an agentic butt plug
You made me look and now you all have to know there’s a library for butt plugs (written in Rust) that has LLM generated code in it:
https://github.com/buttplugio/buttplug#inclusion-of-llm-generated-code
I have mixed feelings about speaking things into existence
Considering the amount of weird hentai on the internet that cannot end well.
This is how I learn that buttplug.io has fallen
I hurt myself today… to see if I still feel…
If it’s “agentic,” doesn’t that imply it smokes weed for you
Two tech-related links for today, both relating to fascism.
First, tante has a new blogpost about AI being overtly fascist in nature, which he’s also posted to LinkedIn, seemingly for kicks. (Found on red site, too)
Second, the Nazi laptop company has sent a second pre-release laptop to DHH, showing they have not changed at all since they went full fash six months ago:
Tante nails it again. No notes.
That he has a LinkedIn finally explains how I first heard of him. A liberal but very startup/hustle culture brained colleague shared an anti-blockchain thing on the Slack. Always wondered how she stumbled on a comm(o|u)nist tech critic, but it must’ve been LI.
Ex-CEO, ex-CFO of bankrupt AI company iLearningEngines charged with fraud - Reuters, 17 April 2026
Prosecutors said iLearning marketed itself as an artificial intelligence-driven digital education company with an “out-of-the-box AI platform,” and claimed to earn revenue mainly by selling licenses for its educational and training platforms to customers, including healthcare companies and schools.
According to the indictment, the defendants used forged sham contracts to make it seem that iLearning’s customers were real, and used “round trip” transfers of investor and lender funds – meaning they sent money to purported customers, who then returned it to iLearning – to manufacture revenue.
At least 90% of iLearning’s $421 million of reported revenue in 2023 was fabricated, the indictment said.
I think they called this wash trading in cryptoland.
Why? So it can be firebombed?
(don’t fire-bomb residences people)
also it’s hilarious that a 5% local tax surcharge on a place you don’t even live in is considered Stalinism
This has been getting posted repeatedly on a reddit sub I follow and it’s funny that we’re still finding Harris fans who somehow didn’t catch on for all these years that the dude is just straight up racist and hates Muslims on an individual basis. The fact that he was probably the most commonly mentioned podcaster on the slate star codex sub is a sign!
I’ll steal my general remark from one of the threads on DTG:
Sam, if you’re allowed to call the mayor an Islamist because his wife liked some posts on social media, can we call you a racist for having a notorious racist on your podcast to yammer about the woke menace?
For the uninformed (like myself), the odd fox out is a character from Zootopia (pic related), which is a Disney movie.
Bold move to steal from The Mouse in broad daylight, isn’t it?
You can see also his rabbit counterpart from Zootopia in a princess dress in the background of the first picture!
I assume that’s more because they used AI to lift zootopia’s art style wholesale and so that’s just how rabbits are now.
I can’t imagine the Mouse being happy about this cameo.
damn i wish i had confidence of a mediocre techbro. twitter thread suggests that there’s four copyright infringements just in these two images
e: still can’t get fingers right, lol
looks up, shrugs Everybody involved here kind of sucks, best of luck to all of them. looks back down
When we’re all too tired for “let them fight”.
Look I know Doctorow fails the gen-AI purity test but I think he’s making sense here
Too narrow of a scope to be useful I think.
When he talks about LLCs, the real thing he’s referring to has to be capitalism. The “unaligned AI is already here, it’s corporations” has been around for a while, and I don’t think it convinced a lot of people to become socialists. I think you can’t usefully tackle just the capitalism of the tech industry, if you don’t understand the wider system around it. I mean the VC that sloshes around in tech to raise up one bullshitteer after another comes from somewhere. Real estate, mining, fossil fuels, etc.
When he talks about LLCs, the real thing he’s referring to has to be capitalism. The “unaligned AI is already here, it’s corporations” trope has been around for a while, and I don’t think it convinced a lot of people to become socialists.
Nah, this is the threat as he describes it:
Every moment that we remain stuck in the enshitternet is a moment of existential risk. At the click of a mouse, Trump could order John Deere to switch off all the tractors in your country
He doesn’t need tanks to steal Greenland. He can just shut off Denmark’s access to American platforms like Office365, iOS and Android and brick the whole damned country. It would be another Strait of Hormuz, but instead of oil and fertilizer, he’d control the flow of Lego, Ozempic and deliciously strong black licorice
He’s not saying that “corporations are a parasitic species we created that will devour our precious planet and ability to survive and turn it into shareholder value for our corpses to enjoy” no sir he’s saying we need to switch to open source replacements for Office 360 and get Right To Repair for our tractors.
Yeah, but he’s been saying this for a while, and I don’t think this is a new way of saying it that is particularly useful.
Agreed
Yeah. I mean, I’m going to assume that he’s not specifically referring to LLCs and that, say General Partnerships and Sole Proprietorship aren’t actually better, mostly because they absolutely share the same problems. But that’s a whole lot of digital ink spilled about how important it is to create “digital public goods” - and one assumes public goods more generally, given our ongoing inability to subsist on digital food - without any real thought put to how such systems should be organized, controlled, protected, etc. Like, if he stopped trying to wave “digital innovation” around like a magic wand he’d basically have to be a communist or anarchist of some flavor. But he’s still sufficiently in the Silicon Valley Milieu where lefty politics are too cringe to admit, so he’s stuc having identified some real problems and having no realistic starting point for solving them
It’s not so much he fails the purity test than it is he thinks all gen-ai works like whisper on a laptop and open source LLMs grow on trees.
It’s not just that he fails a purity test it’s that he basically only cares about his pet issues like right to repair and hating copyright and positions himself as a social activist while twisting real social issues to fit into his framework of ‘copyright bad, hackers fix this.’
You thought software quality was poor before LLMs? You ain’t seen nothing yet.
Typical mentality on the orange site:
I don’t think I’ve read a single line of code I’ve shipped in over 6 months.
The poster goes on to whine about how it sucks it’s not the same as solving puzzles and they’re just QAing all day now and calls the LLM a slot machine, so at least they’re not boosting.
Still, they don’t go so far as to say they were forced to work this way, so their not even looking at the code either means they’re a lazy bum or that it’s too far past incoherent to be worth it.
New random positivity thread:
This week in unsettling AI ads.
The algorithms feed me variants on this one pretty frequently, always with some fantasy woman offering to be your best friend and definitely not maybe give you dirty pictures. “Normal human woman tied up in a basement” is new, though, and even skeezier than usual. I don’t know what the workflow is for these or whether there’s an actual person writing the prompts for this, and I don’t think there’s any answer that would make me less uncomfortable about it.
Wtf is up with the face in the top left circle.
Another lament about the tech industry has come to my attention, as Ky Decker asks themselves “Do I belong in tech anymore?”
Pretty solid piece all around, recommend checking it out
[lament about ill effect of AI, here’s why it’s terrible]
I use AI sparingly for
this is critihype
I use AI sparingly to make sure the company-paid subscription is a net loss for the AI vendor.
Hey, it could happen.
Overall, I think it was a bit cookie cutter for an article of this type, but maybe It’s just the preaching to the choir effect. Even the fact that he ostensibly quit his job over this stuff doesn’t hit as hard as it should, it comes off as if he could have done so at any time but this way he gets to grandstand about it.
Also stuff like this:
It wasn’t a bad job, not by most metrics. It ticked the boxes a job is supposed to tick: good pay. Health insurance. Remote work. Time off. Nice coworkers.
sounds like it should be in a how do you do, fellow workers copypasta.
I can’t tell if this is serious.
If that’s really how we react to someone who admits AI use although they know why AI is bad, we need to listen to Dr. Fatima some more.
There is enough of this shit that appeasement has no reason except making well meaning moderate centrists more comfortable, and they were never the audience. (despite their unshakable conviction they are the only audience.) dr. fatima’s video is good advice on comforting the comforted and nothing else.
i commend to you my masto post with the more detailed version, and the responses thereto, e.g.
the world we live in where the golang stdlib cryptography maintainer doesn’t know how to use ffmpeg and can’t be fucked to learn and so instead relies on slop
I don’t have much sympathy for the “let’s wait and see” moderates, but I do think there’s a coherent difference between people who have tried AI tools and found some use for them in some limited context and people who go full Howard Hughes with it like John McGasTown or whatever that idiot’s name is. To me it feels like an extension of the argument that these so-called AI systems are a normal trchnology. They aren’t a harbinger of the end times, whether you interpret that as the singularity or the biblical Armageddon. It’s a normal technology that is breaking in normal ways and is breaking society and the economy in the ways we would expect late capitalism to break. If it wasn’t this it would probably be something else. Hell, there’s still a chance that the wheel turns to “Quantum” or something else after this and we stretch another few years out of that before the music stops.
AI is a bad tool for any given job, and is fundamentally not worth the price that we as a society are paying to let it exist at this scale. If it wasn’t being subsidized by capitalists chasing ridiculous returns and bouyed by an economic system structured entirely around giving it to them then there’s no way in hell it would have hit this point. But that’s not incompatible with people being able to find utility in it in some cases, and I think we lose credibility by treating any admission that someone has found any value in AI products as a confession of unseriousness. That doesn’t mean their use isn’t still part of the problem, but I’d we frame the critique in terms of “how much would you actually be willing to pay for you ‘occasional’ use?” It would redirect the discussion away from the subjective “well I found it useful for X” to the more objective question of just how expensive and destructive these things are to operate and how much of those costs are going to have to be subsidized forever if these things are going to stick around.
Welp, thanks for clarifying.
I find it hard to get mad at someone who uses AI sometimes on account of all the bozos who went full Gas Town.
i generally like Dr. Fatima, so i was curious about this video, but it was pretty disappointing. I have several more thoughts but i wanted to keep this reasonably short.
The first section about “back end harms” is the best part. Unfortunately, section 2, the “Front End Harms” section names valid problems but falls flat when it comes to solutions. She rolls out a lot of lib tropes about “education” and gestures at companies self-regulating the sycophancy of their own models despite evidence (a massive amount of it if we consider corporate “self-regulation” more broadly) to the contrary. Remember “media literacy” discourse about social media misinformation? it went nowhere, because it’s not a problem that can be solved with education, it’s imperative to actually learn lessons from history and bring this technology under political control. You cannot do this effectively when your government is 3 monopolistic corporations in a trenchcoat.
She says that anthropic are “better than the competition” which is trivially true but extremely credulous. If faced with the choice, I would prefer that Claude beats out Grok and chatGPT but this is ultimately a marginal difference due to the nature of the industry and ultimately of capitalism itself.
Section 3 is fine for what it is, but it’s really about the psychology of persuasion and not AI as such. Some of the discourse on this site would benefit from the reminder that moral absolutism isn’t very persuasive, but this section is way too long, we can just dispense with the moralizing and “harm reduction” anyways, because just like plastic recycling, personal reduction in AI use for harm reduction reasons is a fake solution to a systemic problem.
The harms of AI are intimately linked with the nature of monopoly & platform capital. You cannot defeat this enemy if you cannot actually describe it properly, and no amount of leveraging NIMBYism to defeat your local datacenter project will fix it.
Maybe Dr. Fatima isn’t the right person to be delivering that message, or simply doesn’t view the problem through a materialist lens. It’s not enough to be against AI, you need to be for something. This is the disease of liberal technocratic managerialism and it manifests in myriad ways including AI critique. We need to move beyond it.
“I use AI sparingly for” is the new “I only do meth before my hookups for unprotected sex with strangers”.
Saw a remarkable take on the pro-AI parts of bsky, that since DeepSeek420.69 can offer the model at like 15% of Claude’s pricing, that must mean that Anthropic is operating at an at least 80% positive margin on inference, so things will work out.
In the same thread they complained about Zitron’s math being dodgy.
