{"slug": "we-never-use-ai-for-anything", "title": "We never use AI. For anything", "summary": "A tech commentator argues that AI is a 'con man' and a 'rug pull scam,' claiming that large language models are 'nothing but a fancy autocorrect' that will never be fixed, and that AI is 'killing the environment and destroying people's lives.' The author cites a personal test with Claude, where the AI generated useless code and then lied about its necessity, and warns that dependence on AI will lead to high costs once subsidies end.", "body_md": "## AI is designed to make you dependent on it\n\nWe are still in the gravy train part of the rug pull scam. AI is\nheavily subsidised by billionaires and massive corporates, who have\nbeen sold on a false promise of replacing you, the person, replacing\nyou, the worker, with a perfect machine slave that never rests, never\ncomplains, and does everything it's told.\n\nBut what will happen once the investors start demanding returns, the\nfree money runs out, the true cost of running AI models comes forward,\nand those that have been on the hype train realize the true\nlimitations of AI? Those who have become dependent on it will be\nforced to start paying big. And the already obscene fortunes of AI\nRobber Barons will grow even further.\n\n###\nAnd that is not us. We value independence and long-term planning.\n\n### Therefore, we reject the false premise of the “AI Future.”\n\n## AI is not intelligent, and hallucinations are a feature\n\nFrom self-prompting to multiple agents talking to one another until\nthey both agree, countless resources have been spent on trying to\ncorrect AI hallucinations. However, hallucinations are inherent to the\nway AI, or Large Language Models more specifically, work, and they\nwill never be fixed.\n\nLarge Language Models are nothing but a fancy autocorrect. Where they\ndiffer from your phone’s keyboard is the size of their memory, which\nthey can use to predict the possible next word, based on a huge\ndatabase of previous words. But that’s all it is, a prediction.\n\nThe AI doesn’t know or understand what it’s writing. Give the same AI\nthe same prompt, and watch it give you two completely different\nanswers. It can guess, but it will never know. Fixing this would\nrequire a complete rebuilding of the way the Large Language Model\noperates.\n\n## AI is the most successful con man in history\n\n*“Funny how when I ask AI about something I know, it’s completely\nwrong, But when I ask it about something I don’t know, it gives a\ngreat answer.”*\n\nThis is a paraphrasing of a post I saw on Reddit a couple of weeks\nago. In addition to what was written above, corporate executives love\nAI because it is extremely good at one thing - making you think it\nknows what it’s talking about, and masterfully avoiding any\naccountability.\n\nJust last week, just for fun and to keep up with the newest\ndevelopments, I tried asking one of the leading AIs, Claude\nspecifically, to add a simple feature to a search button. The\nrequirements were simple: automatically focus the field when it\nappears so the user can start searching right away.\n\nAt first, it gave me something that seemed like exactly what I wanted,\nand it even seemed to work properly… at first.\n\nWhen I inspected the code more closely, I saw a massive chunk of code\nthat did seemingly nothing - variables being reassigned to themselves,\ndelegates and coordinators being created and never used, pointless\ntwo-way bindings that referenced one another convenience initializers\nthat initialized one variable out of ten, then silently failed.\n\nJust for the fun of it, I removed all the pointless code that Claude\nconjured up, and asked: “Even without this code, the feature works as\nit should. What is the point of all the useless code?”\n\nAnd what I got back didn’t surprise me in the slightest. “You’re\nabsolutely right! The code doesn’t need to be there, but if you remove\nit, (the feature that the AI made up that was not in the original spec\nat all) will not work.”\n\nAI is great at lying. And that’s it.\n\n## AI is killing the environment and destroying people’s lives\n\nImagine a beautiful forest full of plants, animals, and various\ncritters. A functioning ecosystem.\n\nNow, burn it all down. Bulldoze the trees, pave over the grass with\nmonochromatic asphalt. And in place of an environment teeming with\nlife, build a data center that disturbs the weather, devours only the\nmost pristine water to cool itself, and uses unbelievable amounts of\nelectricity to churn out pages nobody will actually ever read.\n\nPeople living around this kafkaesque complex will be forever tormented\nby noise they can just barely hear. The water they need to live will\nbecome brown and black with dirt and rust while the data center gets\nit in their place. They won’t be able to shower or get water out of\nthe tap, as the pressure needed will go to pushing millions of liters\ninto the useless facilities.\n\nThat is the most wretched and perverted use of any technology in\nrecent years.\n\n## AI will lead to a tyrannical police state\n\nWhat will happen once the rug invariably gets pulled? Those data\ncenters will start serving another purpose.\n\nAll around the world, governments are assembling massive networks of\ncameras, “license plate readers” that are suspiciously equipped with\nmicrophones sensitive enough to listen in on your conversations, and\nother unimaginable surveillance devices.\n\nAll that computing power built up under the guise of bringing about\nthe “AI future” will start being used to spy on you, follow your every\nmovement, listen to all your calls, and more.\n\nMuch of this infrastructure has already been established.\n\nOnce those data centers are sitting idle, once the commuting power is\nnot being used to churn out unbelievable amounts of useless slop, it\nwill be used to supercharge these already existing programs.\n\n## Nobody knows what AI even is\n\nAI has become a catch-all term for whatever system tends to be\ngenerative and produces unpredictable outputs that cannot be\nreplicated.\n\n“AI pictures” this, “AI text” that. If it produces slop, it’s AI.\n\nIt’s no wonder that people think that. A lot of the AI tools bundle\ntogether many unrelated technologies, selling them as a single “AI”\nproduct.\n\nBut if we peek under the hood, we will see:\n\n-\n*Text generators?* Those are actually\n**Large Language Models**, nothing but fancy autocorrects\n-\n*Image generators?* **Diffusion Models** that take random\nnoise and try to recreate a picture by grouping surrounding pixels\ntogether\n-\n*Image classifiers?* They’re a technology that has been almost\nunchanged for the last 20 years, and has nothing to do with the\ncurrent notion of “AI.” Optical Character Recognition has been a\nthing for decades, and works as well as it did back then, as it does\nnow.\n\nAnd none of it can reason or think. And it never will.\n\nOnly people are intelligent, creative, capable of actual thought. We\ncan’t let this unique gift go to waste.", "url": "https://wpnews.pro/news/we-never-use-ai-for-anything", "canonical_source": "https://corkmac.app/our-ai-stance/", "published_at": "2026-08-24 09:32:32+00:00", "updated_at": "2026-08-24 10:12:59.736827+00:00", "lang": "en", "topics": ["artificial-intelligence", "large-language-models", "ai-ethics"], "entities": ["Claude", "Reddit"], "alternates": {"html": "https://wpnews.pro/news/we-never-use-ai-for-anything", "markdown": "https://wpnews.pro/news/we-never-use-ai-for-anything.md", "text": "https://wpnews.pro/news/we-never-use-ai-for-anything.txt", "jsonld": "https://wpnews.pro/news/we-never-use-ai-for-anything.jsonld"}}