"Brand extension" ---> "Brand amputation," maybe? #neologism
Author, journalist, editor and critic combining his time in Turin, Austin, Ibiza, Belgrade etc #BruceSterling
5,151 followers 1,835 following 4,052 posts
view profile on Bluesky
"Brand extension" ---> "Brand amputation," maybe? #neologism
"Reality is what is ordained by the stark Laws of Nature instead of mere human illusion!" But maybe there is no such "law," and reality is just generated on-the-fly by some loose, daffy, LLM-style statistical process *Are we supposed to hang ourselves? We'd get used to the idea
*As a career science fiction writer, I've spent so much time with this kind of Philip K. Dick "ontological riffing" that I can't find it threatening in this melodramatic way *A "Copernican dethronement," it's only traumatic if you're fully-vested in clinging to that particular throne over yonder
*Rather a lot of ugly truths in the tragically short life-story of your reputable source there. #JohnKeats flic.kr/p/7Dd1i1
*Not to leap to the Bingster's defense here, but since you're dismissing his effort, how about you do a better job of comparing a cucumber to human cognition
*One has to wonder what forms of untrustworthy, dangerous, irrational hallucination are most commonly and implicitly trusted by everybody -- and religion, nationalism and market economics promptly raise their hands *What goofball ever "trusts poetry," not even great poets, least of all them
"Flip the script, what if humans had plant-like cognition," said the gigantic language model as it eagerly munched coal in its new Big Tech data-center
*Meanwhile, on the thinking cucumber front:
*God is not "really" a man, humans just hallucinate Him as a man, and not even consistently, and also, so what
*LLMs hallucinate because the Internet is not in touch with reality. Worse yet, the human condition is not in touch with reality. What if reality itself is hallucinative? 'God plays dice with the universe," and the dice are sometimes generated with some sloppy extra dots
*Prof. Peter Denning. I've always been impressed by this guy's sense of civilized decency. Life must be a trial for him nowadays denninginstitute.com/pjd/PUBS/CAC...
*He says that the AI-skeptics are just handwavey nitpickers who don't understand the actual software business fly.io/blog/youre-a...
"We have achieved success with the Torment Nexus: social media for raccoons and octopi"
"Hey big slop machine. Do all my prompt work for me" medium.com/@jordan_gibb...
*Why not just buy Twitter and neutralize the review system *In fact, you should push it way more. Real men die in fast car wrecks. It's a sign of revived national greatness
*It's like a moralizing trolley-problem: what if my disengaged Tesla auto-driving cybertruck has to roll over an immigrant baby on the critical average of once every 1,266.3 kilometers *It's not even an issue. It'll never be allowed to come up. "Zero Policy Dimension."
*The ideology of AI-driven-cars claims that they would LOWER the automotive death rate, and maybe yes maybe no, but that's really just the chocolate frosting *The cars are an integral part of the installed-base of blood-for-oil; it's been well-established worldwide that those roads have to roll
*Well, that's not a new decision, that's the long-term history of most national highway systems ever since the birth of the automobile. *A human fatality every 26 seconds
*What does a Cory-Doctorow-enshittified autonomous car-driving-system look and act like? Pretty easy to forecast. It "critically disengages" somewhere in the snowstorm and then demands ten bucks to turn back on
*I also quite like the euphemism "critical" because it implies that there are thousands of tiny acts of "disengagement" that are not "critical," but just sorta loose and rattly. And cracked, and flaking-off, but that'll be fine, it's just sloppy, that's how it is. #averagecriticaldisengagement
*I like the "disengagement" euphemism because the system didn't fail, or break, or bloodlly endanger the public. It just sorta... departed. The Artificial SuperIntelligence is off for some virtual coffee-break. It'll be right back, for sure #slop
*The autopilot passenger-drone smashes in flames into an urban high-rise. "Interesting critical disengagement there," remarks the prompt engineer as he stands among the robot ambulances
"It'll get safer and more secure," well, it might, but commonly, when systems do get more secure, the guys involved in running them just push the safer-systems harder in order to get more value out of the buck *When you wear a seatbelt you drive faster; Soccer Mom safe in her SUV drives like a tank
"Average critical disengagement," that's in interesting AI-industry euphemism. How long -- on average -- before the slop snaps or breaks and some human somewhere has to install more hot-glue or a zip-tie. *Good luck with that if you're moving at highway speeds when it happens.
*Also that's some pretty darn good LLM-slop generated FX on the screen there
*I admire this kind of Bluesky-speak. What does this feel like? It's kind of like listening to English-speaking bartenders speaking insider booze-jargon at some overcrowded casino-stripclub in Monaco.
*Mass-produced AI-slop pseudo-history. I kinda figured that this was gonna become a problem. #404Media *It's like a revenue-weaponized, human-labor-free modern mutation of steampunk.
*Burning Man. A great place to liquidate some Russian emigre dude who got in over his head. #BowieKnife meduza.io/en/feature/2...
*I used to warn people that weblog platforms would become dead media some day, like 8-track tapes or mimeo machines, and they were like: "How could that be? Everybody knows that the Internet is Forever" everything.typepad.com/blog/2025/08...
*"Bored Ape Yacht Club" means that I'm a crypto bro, I'm a bored ape, me and my club of pals have got yachts, and you, who disapprove of our nouveau-riche hacker aesthetic, can have-fun-staying-poor
"While the job appears glamorous, models deal with myriad challenges, including late payments, financial debt, eating disorders and sexual harassment" -- and being upstaged by their LLM lookalike clones www.latimes.com/business/sto...
*Speaking of hugely-expanded attack-surfaces, the Norks have discovered AI "pivot-from-crypto" 🇰🇵🕵️♀️
*Well, it's hard to get more brazenly crypto-indifferent to visual propriety than "Bored Ape Yacht Club," where there's whole years of art hijinks to study
*I blame fantasy that now literally lies in our hands
*A genuinely "new genre" would involve an LLM that absorbs every detail of a woman's life-history and then writes a romance-tinged self-help book where she is the protagonist and every paragraph flatters her, and it's 60,000 words and next year she gets the sequel Pop fiction with new capacities
*It's not a 'new genre," because it's actually the collapse of an aging futuristic SF-subgenre into a form of contemporary social-realism, which is kinda sad, frankly *You get used to it, tho #CyberpunkSocialRealism
*In other words, not only have the Language Models become "fine writers," they've got their own literary critics now futurism.com/gpt-5-litera...
*It also increases the "attack surface" by a crazy among and the "attack surface" is what ruined the Internet-of-Things *When your psychoanalyst AI-Girlfriend turns out to be a pig-killing Chinese Mafia contraption under the hood, because she got pwned with a prompt-injection, she won't look pretty
*I like monobloc chairs as a slop metaphor because humans have thousands of years of intense physical, emotional, cultural, aesthetic and intellectual engagement with chairs, and yet plastic monobloc chairs are one of the commonest, most ubiquitous things on this planet, from pole to pole
"Clean up the noise in that image," yeah, sure, okay, fine *YouTube has taken to doing that to people's amateurish videos, and they don't ask permission, either *So the video product is neat and looks more pro, but the userbase catches on that it's slop anyhow, and they don't like it
*I don't think "slop" is "noise." The stulf they call "AI hallucination" might be noise. Slop is just slop. Cheap, fast, gets some sort-of job sort of done, like hot-glue or injection-molded plastic. Often shiny and with lots of detail, but very human-low-effort as an artifact
*Well, the canonical non-slop example would be "AlphaGo Move 37" *It looked like a mistake, but it was outside standard human cognition in that arena and it wasn't "slop" johnmenick.com/writing/move...
*I've never liked the "magic," "witchcraft," "daemons" etc etc, it's obscurantist and doesn't get us anywhere *"Cauldron of data soup" is pretty good, but an "AI Agent" can dip around in a couple dozen different "cauldrons," while the eventual result is still "slop" on the screen
*Good try, but there's probably a lot of different kinds of it and it needs some kind of taxonomy. #jargon *If something that resembles an "artificial super intelligence" spits up a diamond-hard lump of advanced super-slop, what does that product even look like
*If AI's kill off humanity and take over the world, what are they themselves supposed to do for a "retirement system" and some stable, fulfilling old age *They're worse off than humans, they don't even have next-of-kin futurism.com/ai-experts-n...
*I'm inclined to think that all AI output is best described as "slop" when it's accurate and adequate. But that leaves a verbal hole for the proper word for AI-output that's truly superb, inhumanly brilliant, revelatory and paradigm-changing, etc *That stuff can't be described as "slop"
*Not buying the AI-pocalypse here, but it's easy to imagine a plausible future where "investing for retirement" is useless, because dotcom-bubble / AIbubble futurism.com/ai-experts-n...
*Categorizing various forms of Artificial Insanity. So bonkers that it reads like an AI hallucinated it www.psychopathia.ai
"Reasoning Token Shock," what a bizarre term-of-art #jargon
*Where did you expect it to be
*Olga Scheinpflugova, Mrs Karel Capek. Olga was a pretty actress from Prague. Olga was hot. Her husband's posh family wouldn't give Olga the time of day *She had the grave right next to his, one couldn't help but notice that www.pinterest.com/bruce2385/ol...
*Did I ever tell you about the time I visited Prague and I left a copyright-free cyberpunk floppy-disk on Karel Capek's grave "Pics or it didn't happen" *Okay, that was back in a remote 1990s era of analog cameras www.wired.com/1995/01/prag...
*Hey yeah, Canadians. Go get Margaret, that Canadian chick 📕👽🦑
*The "traditional option" is Karel Capek inventing the dystopian concept of "robots" as a nonhuman slave-labor force designed and built to crush human labor-power in the interests of oligarch financial power en.wikipedia.org/wiki/R.U.R.
*Oh, and the humanoid robot in question is also 105% Chinese, by the way youtu.be/or0iYHjoptE?...
*It's a human-shaped, networked peripheral for some Big Tech Platform's Large Movement Model, like Alexa or Siri, but with arms, legs, eyes and ears, and a database. *Is there some way to make that cute? Does it flatter you a lot, is it your psychoanalyst?
*The contemporary "humanoid robot" paradigm. There's a mechanical entity inside your house and/or workplace, but it's not an old-school individual "mechanical butler."
*Whatever happens to musicians will happen to video marketers and ad-guys. I dunno why, but somehow I didn't expect those core creatures-of-capitalism to be facing AI-extinction youtu.be/3Zvk4AMCrG8?...
*Look, ChatGPT exists, just push the button for it. #moviescript
*The movie's introduction scene where she's in an Ursula Andress 007 bikini, staring at the beaches of occupied Crimea, with powerful binoculars
*If there's ever a thriller movie made about blowing up the Nordstream pipeline, I bet that the one woman in the saboteur diving team features prominently. #Ukraine #fossilfuels #war www.pravda.com.ua/eng/news/202...
*Should I ask about lucrative product tie-ins that are also "slopjects" #BruceSterling #gaming #collectibles
*Should I ask Bing/ChatGPT to go ahead and vibe-code some of those suggested games and also set me up a royalty-structure
*Boy, that's pretty quick and handy, eh? I wonder if human game-design people even have jobs now
*Can't you just ask ChatGPT to summarize all the stories for you
*When surreptitious AI-use gnaws its way into the stuffy proprieties of some traditional company, a lot of practical hazards appear that "AI Hazard" people don't talk much about *All the human workers forgetting how to do important stuff, for instance read.misalignedmag.com/the-new-chal...
*Using a specifically trained LLM for banking services in the creole "Manglish" dialect. #language #LLM
*Try posting this text, unsourced, to your followers and ask them "What famous American novelist wrote this"
*It "sounds like Pynchon" in the way that some skeet on Bluesky sounds like some verbal quip that some human being actually said aloud
*AntiScientific Trumpist America-Firster
*Chunks of open-source code that are functional and useful but have nobody human who is willing to work on maintaining them. neilb.org/adoption/
*That would be a good bet. Drop by for "Share Festival" in Turin some time. #artdirector medium.com/@bruces/the-...
*Which of these two women is most likely to hallucinate, lie to you, manipulate your feelings, empty your bank account and then ghost you after your psychotic breakdown *Trick question, they're the very same "girl"
*There's a ton of hot-take blather about why ChatGPT5 is a flaccid bomb while ChatGPT4 seemed hot and dazzling, and it's really a senseless folk-panic there. Nobody including the userbase financiers owners and manages has a ghost of a clue about this revulsion *They all SAY why, nobody knows why
*I completely agree, and here's a bunch of other AI papers that also seem pretty good, and I'll get around to reading every one of those just any year now www.programmablemutter.com/p/the-politi...
*I also agree that AI Slop is oddly colorless and anodyne, but that's because its designers were extremely keen on "harmless helpful and honest," while no one has ever weaponized AI for "extremely charming and persuasive" or "dazzlingly idiosyncratic"
*I agree about the persnickety author cult-thing, but that's related to the singer-songwriter cult thing or the film-director cult thing, and not intrinsically necessary for published writing, distributed music or highly-viewed cinema/video
*I hear you, but I'm not buying either of these premises *Guys with deadly weapons are always "defending the peace," while educational textbooks, which actually convey coherent thoughts, are notoriously painful to read and never on the bestseller lists
*Intellectuals think that stupidity is boring and lacks charisma, but "complete stupidity" is like LSD or laughing-gas. No one is immune
*It's like thinking that QAnon fanatics and J6 patriots won't sacrifice their lives and sanity for crazy, sourceless, viral rumors that no qualified political-commentariate auteur would ever place on a screen *But they do. They exult in it, it's like their champagne
*This is a John Ruskin arts-and-crafts argument, and anyone interested in modernity and industry should carefully read John Ruskin, because he was catastrophically wrong about every single thing
*This seems like a positive aspect of AI-Slop, since it makes it clear that the combatants are phoning-it-in *Somebody at Propaganda Central told them to 10X the culture-war production and flood the zone with cheaper-faster LLM bullshit *In theory this should cause madness but it's soporific
*AI-Slop has invaded the ongoing meme-war between Russia and Ukraine *It's hard to say why, but it really sucks the Slavic bitterness, folk-hate and eye-catching sarcasm out of that poisonous and polarizing imagery *It becomes anodyne, toothless
*This is the school-of-thought that believes that AI Slop will become the artificially-intelligent equivalent of hot-glue, duct-tape, zip-ties and cheap plastic *All over the planet, you can't really trust it, who cares if the price is right
*An AI could have written this if you asked it to write a textbook example of the Gartner Hype Cycle "Trough of Disillusionment" www.theregister.com/2025/08/25/o...
*Being an author, I hear this auteur-centric argument rather a lot, but it's like wondering why anyone would ever buy an automatic weapon that was deprived of the artistic loving-care of Mr Colt or Mr Kalashnikov
*Are these slopjects, no, this is just normal, everyday digital 3D-modeled solid metal-casting. #industrialdesign #finedistinctions #weird www.digitalmetal.io
*Read all this stuff and you'll be able to describe modern AI like a real scholar, and just in time to have your university destroyed by cranks and oligarchs, most likely www.programmablemutter.com/p/the-politi...
*Whatever happens to musicians will happen to novelists youtu.be/Y4B4QuDxCos?...
*Sanctions, blood-for-oil, pipelines, Ukraine, Russia. Orban, drones-drones-drones, etc kyivindependent.com/hungary-bans...
*"Prompt injection" hacks, where your LLM agent is illicitly prompted to rob you and of course promptly does
"Here’s the part that should terrify you: there is no policy solution to this," he Zero-Policy-Dimensioned. #economics #classwar #oligarchy medium.com/the-investor...
"They're good kids, but they still believe in a "policy dimension," while I just rule this world by QAnon-soaked Presidential Decrees on TruthSocial"
*I like "Zero Policy Dimension." It's starting to sound like an actual dimension, like some place that Bizarro Donald Trump might inhabit in some DC comic book