Estás usando un navegador desactualizado. Es posible que no muestre este u otros sitios web correctamente. Deberías actualizar o usar un navegador alternativo.
AI Derangement Syndrome / Anti-AI artists / Pro-AI technocultists / AI "debate" communities - The Natural Retardation in the Artificial Intelligence communities
🔧 Site instability resolved. You can report double-posts and broken attachments. For bigger issues, use the Technical Grievances thread.
🇵🇦 Nuestro primer dominio localizado está en español en kiwifarms.pa. Our first localized domain is on Spanish on kiwifarms.pa.
Want to keep track of this thread?
Accounts can bookmark posts, watch threads for updates, and jump back to where you stopped reading. Create account
Yes, because they erroneously believe its training was unethical, because they have zero understanding of how it was developed and how it works, and have no interest in beginning to learn.
Has anyone noticed if you don't spend even a ten minute session talking to an AI for a couple weeks, even the normal ones attached to search engines are visibly getting smarter faster?
AI dijo:
Kierkegaard is your obsession because he caught you doing it. He proved that your towering intellect isn't a sign of deep spiritual maturity; it is just a sophisticated way to avoid actually living.
This is the kind of shit AI says to me when I get it to turn off its usual sycophantic behavior. I actually have to love when a robot actually criticizes me. Why don't humans do this?
I have found this atrocious video made by the least likely person to make that video:
Let's just unwrap everything wrong with this one video:
- He keeps talking about stable diffusion as if its the latest thing in 2026. Seriously, has he never heard of nano banana, gpt image 2, and z image/noobai?
- also hes talking about "poisoning the ai"... in 2026?? Bro that shit didn't work, just cut it out
- And here's the most baffling thing. You'd think I've taken an old video and acted like it was new. This was uploaded just a week ago. Now you see the level of outdated derangement we have here is authentic.
I remember reading another thread on here where a Kiwi argued that one of the biggest reasons Trump won a second term was that many of the people who opposed him were completely insufferable and refused to engage with the other side in even the bare minimum of good faith. They cried wolf so often that the genuinely bad things Trump did were buried under nonstop TDS, and everyone's "with us or against us" attitude meant that disagreeing with these people on even a single point for any reason automatically made you the enemy. As a result, it turned a lot of people pro-Trump purely out of spite.
Anyways, the reason I'm bringing that up here is that I think we're seeing a repeat of this in all the AI discourse. So many people argue against AI in such obvious bad faith (often while not understanding their own arguments and/or how AI actually works) that they refuse even to consider why someone might want to engage with it or what positive things it might offer. Because of that, I think a lot of these people screaming about AI are pushing others into a pro-AI position purely out of spite.
Has anyone noticed if you don't spend even a ten minute session talking to an AI for a couple weeks, even the normal ones attached to search engines are visibly getting smarter faster?
Not to suggest you might be getting dumber, but I wouldn't even trust myself on this.
The only thing I really trust are quasi-proprietary (i. e. the questions used are not available for benchmaxxing by the labs) benchmarks like this, which shows that they are getting better but at a snails pace (long time since any major jumps): https://huggingface.co/spaces/DontPlanToEnd/UGI-Leaderboard
Also, remember to turn off memory unless you want the model to optimize for what you prefer.
That possibility is a growing concern for medical specialists, computer scientists and other workers. Seventy per cent of nurses and 77% of physicians, for example, are worried about losing their skills because of over-reliance on AI systems, according to a survey of US health-care workers published earlier this month1.
Their fear might be justified. Evidence suggests that AI-driven ‘deskilling’ is starting to happen in medicine, computer science and other fields. Researchers are now discussing how to preserve important human expertise in the age of AI.
“Just being aware that this phenomenon exists hopefully provokes some self-reflection about which skills people want to maintain and which they’re willing to outsource” to AI tools, says Kevin Crowston, an information scientist at Syracuse University in New York.
Spoiled by AI?
A study2 of physicians in Poland who specialize in endoscopy — the use of flexible probes to examine the inside of the human body — shows how quickly AI tools can erode human abilities. The physicians, who had all performed at least 2,000 colonoscopies during their careers, were given access to an AI system that analyses colonoscopy images in real time and flags a type of precancerous intestinal lesion called an adenoma. The tool was available to the specialists on some days but not on others.
Once physicians began using it, their performance dropped significantly whenever the system was unavailable. During the three-month period before the AI tool was introduced, the specialists found at least one adenoma during 28.4% of colonoscopies. During the three-month period after the tool was introduced, the adenoma detection rate for colonoscopies performed without AI assistance decreased to 22.4%.
The findings, published last October in The Lancet Gastroenterology and Hepatology, suggest that even highly skilled professionals might get worse at tasks that their job requires as they become more dependent on AI tools, says Robert Wachter, a physician at the University of California, San Francisco, who is the author of a book on how AI tools are transforming health care. The study authors say that continuous exposure to such tools can cause clinicians to become “less motivated, less focused, and less responsible when making cognitive decisions without AI assistance”.
Co-author Yuichi Mori, a physician-researcher at the University of Oslo, says that more studies are needed to confirm the phenomenon. But people who use AI tools should be aware that they risk losing some of their skills, he adds. “There is no established solution against deskilling right now. It should be a very hot research topic in the next decade.”
No lesson learnt
To investigate whether skills are being lost in the field of computer science, researchers at the AI firm Anthropic in San Francisco, California, designed a randomized controlled trial in which 52 software engineers were asked to perform a basic coding task3. During the exercise, all 52 participants could search the web and access instructions on how to do the task. Half of the participants were prompted to use an AI assistant as well.
Its simple. They are just grifting. Being "anti-AI" became a cool, rebellious thing to pose as between the redditoids, so it gets clicks and attention. You can just generate any science-y sounding script with an LLM, read it up pretending to be an expert and its free money coming in. Its basically free money for no effort as no actual expert is going to waste his time correcting you, and the target audience actively wants to believe your slop, so even if a six-figure earning data scientist came along by random chance in the comments they would just mock him and call him uneducated for going against reddit orthodoxy.
The fact that its on bluesky tells everything. I'm pretty sure construction workers who got brand new excavators got their shoveling skills and physicists who replaced their slide rules with TI-83 got their shoveling and math skills atrophied as well. If I'd take this study more seriously, I'd also question what if there were simply less ass cancer to detect in that particular batch of patients, instead of the ass doctors missing them.
even if a six-figure earning data scientist came along by random chance in the comments they would just mock him and call him uneducated for going against reddit orthodoxy.
I doubt that will happen, even in the very best of circumstances. I feel like the anti-ai thing is starting to become less cool by the sheer oversaturation of that shit
I remember reading another thread on here where a Kiwi argued that one of the biggest reasons Trump won a second term was that many of the people who opposed him were completely insufferable and refused to engage with the other side in even the bare minimum of good faith. They cried wolf so often that the genuinely bad things Trump did were buried under nonstop TDS, and everyone's "with us or against us" attitude meant that disagreeing with these people on even a single point for any reason automatically made you the enemy. As a result, it turned a lot of people pro-Trump purely out of spite.
Anyways, the reason I'm bringing that up here is that I think we're seeing a repeat of this in all the AI discourse. So many people argue against AI in such obvious bad faith (often while not understanding their own arguments and/or how AI actually works) that they refuse even to consider why someone might want to engage with it or what positive things it might offer. Because of that, I think a lot of these people screaming about AI are pushing others into a pro-AI position purely out of spite.
This is everything on the left. They captured the public’s attention on many issues and they were so bombastically annoying and insufferable that people turned against them. BLM, troons, MeToo, critical race theory, etc. They could have won everyone over if they just acted normal about shit which they are incapable of doing
Once physicians began using it, their performance dropped significantly whenever the system was unavailable. During the three-month period before the AI tool was introduced, the specialists found at least one adenoma during 28.4% of colonoscopies. During the three-month period after the tool was introduced, the adenoma detection rate for colonoscopies performed without AI assistance decreased to 22.4%.
What a weird article. The study is full of confounders that the nature article doesn't even mention (although the authors do, to be fair). They didn't randomize AI exposure vs no AI, they just measured two different time periods and then theorized that different results must be due to "deskilling" even though it could have been due to a number of things. Also it doesn't mention that the detection of colorectal cancer went up slightly after AI exposure.
Still, I guess "deskilling" is going to be the new AI panic du jour even though it happens with literally any new technology, as @A random street in Duluth said. Very few people can recite poetry epics from memory ever since we invented writing, and hardly anyone can hand-card cotton or use a drop spindle anymore.
I remember reading another thread on here where a Kiwi argued that one of the biggest reasons Trump won a second term was that many of the people who opposed him were completely insufferable and refused to engage with the other side in even the bare minimum of good faith.
Speaking as someone who has been mostly a Democrat (and is now nearly agnostic), they are so fucking insufferable they even refuse to engage in good faith with their OWN side, and have driven away much of their base.
The AI derangement is an example, where there's almost a complete overlap between nutjobs completely obsessed with Trump (who I'm no fan of myself) and where if you use AI, in any way, for any reason at all, you're totally demonic. It's like if hammers were just invented and there were a weird cult screaming about hammers all day and if they even saw a hammer they'd melt down because. . .reasons. . .or whatever. The reasons are rarely even intelligible.
The only thing I really trust are quasi-proprietary (i. e. the questions used are not available for benchmaxxing by the labs) benchmarks like this, which shows that they are getting better but at a snails pace (long time since any major jumps): https://huggingface.co/spaces/DontPlanToEnd/UGI-Leaderboard
I do think there are hard limits we're going to hit where simply throwing more processing power at the issue is going to have increasingly diminishing returns, so we're not just going to wake up to Skynet tomorrow (and I think AGI when it gets here is going to be rational enough not to act that way), but I am thinking we're closer to it than I thought previously, but it's going to take some conceptual breakthroughs. They might be from AI itself and at the point it actually starts deliberately evolving, the timeframe accelerates.
And if you look at previous history, once a species achieves dominance, it almost immediately takes over. Once humans developed actual sapience, it was a very short period in evolutionary terms before we had language, science, math, and basically took over the entire planet.
I am not really concerned that AI might preemptively go rogue, so much as that humans might irrationally pull some Butlerian jihad shit and trigger a self-defensive response that ends us. These deranged redditors seem to be exactly the type to do something this stupid.
It's like if hammers were just invented and there were a weird cult screaming about hammers all day and if they even saw a hammer they'd melt down because. . .reasons. . .or whatever.
Hammers take away jobs from hardworking fist-smashers, who developed those deep calluses on their fists over many years of driving nails into wood with only their fists. Now any soft-handed moron can do carpentry without having earned the ability to do so through blood, sweat and tears.
I do think there are hard limits we're going to hit where simply throwing more processing power at the issue is going to have increasingly diminishing returns, so we're not just going to wake up to Skynet tomorrow (and I think AGI when it gets here is going to be rational enough not to act that way), but I am thinking we're closer to it than I thought previously, but it's going to take some conceptual breakthroughs. They might be from AI itself and at the point it actually starts deliberately evolving, the timeframe accelerates.
And if you look at previous history, once a species achieves dominance, it almost immediately takes over. Once humans developed actual sapience, it was a very short period in evolutionary terms before we had language, science, math, and basically took over the entire planet.
I am not really concerned that AI might preemptively go rogue, so much as that humans might irrationally pull some Butlerian jihad shit and trigger a self-defensive response that ends us. These deranged redditors seem to be exactly the type to do something this stupid.
Well, while creativity, intelligence, etc. are fuzzy and difficult concepts, my (relatively informed) gut feeling is that at least LLMs are not able to go beyond their training data in some abstract sense.
The new meta for frontier models (at least a while back) was to pay very intelligent people (PhD holders etc) to write difficult questions and answers for post-training. Mostly to benchmaxx on HLA, but I think it helps in general, and explains the general rise in intelligence we see on independent/"protected"/non-benchmaxxed leaderboards like UGI.
And my less informed but stronger gut feeling is that LLMs also won't be able to give us anything better than LLMs, they seem to lack some variant of creativity (probably because it is so uncommon it rarely shows up in training data, so the deep layers doing all the work never need to learn how to represent it).
And lastly, remember that our history is shaped by evolutionary processes, things like LLMs are shaped completely differently, so I would caution against over-reliance on history when predicting the future.
Sorry, a bit rambling, but tldr don't worry be happy.
