Fuck AI

Yes, I hate AI.

None of these things are Intelligent. Artificial or otherwise.

People act like they have answer. They don't. They can only return other peoples ideas, make things up, or at best, find patterns.

They don't provide verifiable evidence, have experience or provide any deep or meaning truths.

Furthermore, they take artistic works - across the spectrum - and reproduce it like its zero value.

It's hype for no more reason than making rich people richer.

And then on top of that there's the enormous power drains as climate change slowly switches to top gear.

What is there to love?!

I hold more against AI than not.

1. Take out all the copyrighted material all the big models illegally trained off of and their AK collapses. Their scrapers are also doing everything they can to kill small websites through DDOS by scraping everything they can as many times as possible and you know this is a feature and not a bug because it gets rid of any form of competition by default.
2. It's an attack on education and critical thinking. Why would the average Joe put in the effort to research and learn something properly lr even question the results if they believe an AI would never lie or be wrong? Critical thinking is already a skill on the decline, but I firmly believe AI is expediting this for a lot of people who either don't know better or just don't care.
3. From a coding perspective, if your software relies on AI generated code, I've heard more stories of that same software being full of vulnerabilities caused by said AI code than not. I also hold a view that if you have to use AI to understand what a program does rather than using something like a textbook written by an actual dedicated expert in a programming language or consult someone who is better than you in said programming language, you are gonna learn the absolute worst practices, like leaving a default Django password that the AI generates.
4. I view AI as extremely unprofessional. If you have to rely on AI to help your work that isn't AI related, I'll take that as you don't actually know what you're doing and am just phoning it in. It also shows how lazy and unable to think for yourself you are. I'd gladly admit I'm dumb as a bag of rocks considering I have used it to fix software errors on my Linux running laptop, so I'm not exception to my own rule. Forums and help groups for software troubles exist for a reason.
5. With the amount of energy they require ( using non-renewables ), any action you do to try and stop global warming is offset within a couple nanoseconds ( hyperbole, maybe, but it's definitely been a big issue ). I'm absolutely positive they don't use renewable, at least for the big US data centers, because that would require building new infrastructure that they're not willing to shill out a penny for. Most likely the same for just about every other country with large AI data centers out there.
6. Looping back to point 1, with them scraping everything to death, they are able to essentially embolden the deranged "artists gatekeeping art!" sickos who believe that artists are somehow gatekeeping art from everyone because they have spent time practicing and getting better. I can pretty much guarantee these deranged sickos want art for free and without having to put in any effort at all because they all don't value art or artists at all. After all, why would they support the arts when they could instead generate AI Jesus for Fakebook and get a shit ton of validation for their "art" from people without having to put in any effort?
7. AI writing is just derivative of the training data it's based on. As of now, because of the way it's trained, you can tell certain models apart from how the text is generated. "Brow furled"? Immediately, if I see that, my mind immediately goes to text generated by Anthropic's Claude AI. Also, the writing, from my experience, tends to be too well written for the average Joe. It's also as generic and logical as I usually tend to write and have been trying to shake off. That's subjective, though, on how generic and logical it reads.

Yes, I hate what's commonly being termed in the current era as AI. It's mediocrity and lies as a service.

I do not hate AI, because it doesn't exist. I'm not delusional.

I do resent the bullshit generators that the tech giants are promoting as AI to individual and institutional users, and the ways they have been trained without consent on regular folks' status updates, as well as the works of authors, academics, programmers, poets, and artists.

I resent the amount of work, energy, environmental damage, and yes, promotional effort that has gone into creating an artificial desire for a product that a) nobody asked for, and b) still doesn't do what it is claimed to do.

And I resent that both institutions and individuals are blindly embracing a technology that at every step from its creation to its implementations denigrate the human work — creative, scholarly, administrative and social — that it intends to supplant.

But Artificial Intelligence? No such thing. I'll form an opinion if I ever see it.

First of all, that which is to get fucked is Generative AI in particular. Meaning, LLM text generation / diffusion model image generation, etc. AI which consciously thinks is still sci-fi and may always be. Older ML stuff also called "AI" that finds patterns in large amounts of satellite data or lets a robot figure out how to walk on bumpy ground or whatever is generally fine.

But generative AI is just bad and cannot be made good, for so many reasons. The "hallucination" is not a bug that will be fixed; it's a fundamental flaw in how it works.

It's not the worst thing, though. The worst thing is that, whether it's making images or text, it's just going to make the most expected thing for any given prompt. Not the same thing every time- but the variation is all going to be random variations of combining the same elements, and the more you make for a single prompt, the more you will see how interchangeably samey the results all are. It's not the kind of variation you see by giving a class of art students the same assignment, it's the variation you get by giving Minecraft a different world seed.

So all the samey and expected stuff in the training data (which is all of the writing and art in human history that its creators could get their hands on) gets reinforced and amplified, and all the unique and quirky and surprising stuff gets ironed out and vanishes. That's how it reinforces biases and stereotypes- not just because it is trained on the internet, but again it's because of a fundamental flaw in how it works. Even if it was perfected, using the same technology, it would still have this problem.

AI under capitalism is dangerous, much more under fascism. In a solarpunk future it would be fine, just another cool piece of tech.

If we take the forum title here, the "fuck" is directed at the people in charge of so-called "AI" companies. The technology has value. It's just being forcefed down our throats in ways that remind us of block chain and whatever happened to block chain?!

I hate all of this Generative AI trash.

AI has been a concept in one way or another for a long time. The idea of ai is fine, this current state of slop machines and chat bots being pushed can suck my ass.

I hate it because LLMs are not AI; they're statistical models that will never achieve true intelligence. It's hallucinations all the way down, regardless of accuracy.

AI could be fine, except that in a capitalist society it's going to be used by corps & govt as a weapon against labor, a surveillance technology, and a way to plagiarize the hard work of artists.

I don't hate AI as much as I hate the nonexistent ethics surrounding LLM's and generative AI tools right now (which is what a lot of people refer to as "AI" at present).

I have friends that openly admit they'd rather use AI to generate "art" and then call people who are upset by this luddites, whiny and butt-hurt that AI "does it better" and is more affordable. People use LLMs as a means to formulate opinions and use as their therapist, but when they encounter real life conversations that have ups and downs they don't know what to do because they're so used to the ultra-positive formulated responses from chatGPT. People use AI to generate work that isn't their own. I've had someone already take my own, genuine written work, copy/paste it into claude, and then tell me they're just "making it more professional for me". In front of me, on a screen share. The output didn't even make structural sense and had conflicting information from the LLM. It was a slap in the face and now I don't want to work with startups because apparently a lot of them are doing this to contractors.

All of these are examples that many people experience with me. They're all examples of the same thing: "AI" as we are calling it is causing disruptions to the human experience because there's nothing to regulate it. Companies are literally pirating your human experience to feed it into LLMs and generative tools, turning around and advertising the results as some revolutionary thing that will be your best friend, doctor, educator, personal artist and more. Going further, another person mentioned this, but it's even weaponized. That same technology is being used to manipulate you, surveil you, and separate you from others to keep you in compliance with your running government, whether it be for good or bad. Not to mention, the ecological impact this has (all so someone can ask Gemini to generate a thank you note). Give the users & the environment more protections and give actual tangible consequences to these companies, and maybe I'll be more receptive to "AI".

I didn't hate AI (or LLMs whatever) at first, but after becoming a teacher I REALLY FUCKING HATE AI.

99% of my students use AI to cheat on any work I give them. They'll literally paste my assignment into ChatGPT and paste ChatGPT's response back to me. Yes, I've had to change how I calculate grades.

The other super annoying part of AI is that I often have to un-teach the slop that comes from AI. Too often it's wrong, I have to unteach the wrong parts, and try to get students to remember the right way. OR, if it's not technically wrong, it's often wildly over-complicated and convoluted, and again I have to fight the AI to get students to remember the simple, plain way.

The other thing I've heard from peers, is that parents are also using ChatGPT to try to get things from schools. For example, some student was caught cheating, got in trouble, but the parent was trying to use some lawyer-sounding ChatGPT argument to get the kid out of trouble. (They've met the parent before and the email seems wildly out of character.) Or in another instance, a parent sent another lawyer-sounding ChatGPT email to the school asking for unreasonable accomodations, demanding software that doesn't even make sense for the university major.

I hate LLMs because their use leads to less human creativity by pushing artists out of creating art, and lowers the quality of the art available to everyone. Not to mention they were all created in a highly unethical manner.

The rest of the slop going on with it is just a sideshow in my opinion. Replacing the very things that make us human with something artificial and 'cheap' is atrocious and I struggle to understand why everyone is going along with it.

When most people talk about "hating AI" they're talking the AI that is this wave before the next winter: (de)generative AI (whether based on LLM or diffusion or whatever ever other tripe drives things like GPT, DALL-E, Jukebox, etc.).

And yes, I hate AI in that sense, in that it is a dead end that is currently burning up the planet to produce subpar everything (words, images, music) while threatening the very foundation of cultural knowledge with obliteration.

AI in a broader sense, I don't hate. Even the earlier over-hyped-before-wintered AI technologies have found niche applications where they're useful, and once the grifters leave the (de)generative AI field we may find some use cases for AI there as well. (I think LLMs have a future, for example, in the field of translation: I've been experimenting with that domain and once the techbrodude know-it-all personality is excised from the LLMs and the phrase "I don't know" is actually incorporated properly I think it could be very valuable there. You still have to look out for hallucinations, though.)

But (de)generative AI in general is overhyped shit. And it's overhyped shit that cannot be meaningfully improved (indeed latter-day models turn out to be worse than earlier ones: ChatGPT4's suite is more prone to hallucination, for example, than ChatGPT3.5). So a whole lot of people are getting pressured, a whole lot of lives are being ruined, a whole lot of misinformation and active disinformation is being spewed by them ... but hey, at least we can have shit writing, shit art, and shit music!

I know A.I might be used for replacing jobs, but that has happened many times before, and it is mostly a positive move forward like with the internet.

This is an excuse used many times but it doesn't stand to inspection. Let's go with robots making cars. When the auto industry had massive layoffs in the '80s the median age for factory workers assembling cars was about the early '30s. What proportion of people in their '30s make any kind of transition to stable, well-paid careers when they're rendered redundant? (Hint: not very many.) An entire generation of the rust belt, in effect, because of automation, were shoved into poverty THAT WE STILL SEE TO THIS DAY. And that's one sector. Automation shit-canned a whole lot of sectors and the reverberations of that have echoed throughout my entire life. (Born in the '60s.)

The only "positive move forward" seen by these traumatically devastating technologies released willy-nilly into society with no mitigation plan is that rich fuckers get richer. Because, you know, Sam Altman needs more cash and not a punch to his oh-so-punchable face.

Oh yeah i hate "AI"

It's a lie

It's not intelligent, it's not even new tech

LLMs were improved via semantic analysis instead of lexical and then they could mimick/break down human speech well enough to translate and communicate.

Then marketing teams looked at this autocomplete bot that can mimick a person's speech and said "Yep, we can trick people into thinking this is Artificial Intelligence"

And this is what our Capitalist overloards are now doubling down on instead of us having a world where we actually understand what AI is and want real technological advancements to happen that lead to intelligent digital life.

Also it's incredibly energy inefficient, each datacenter for these things is already sucking up a city's worth of power just for us to use it to steal other people's work and ideas.

I hate that a market hype bubble has been forced onto us. LLMs are being shoehorned into tasks that they fundamentally can not actually do, they're just pattern recognition engines, they have no intelligence. Despite their lack of intelligence they are sucking up energy, municipal water, a massive amounts of compute so they can be forced onto us.

The only useful "AI" ever created is the one used in health applications to recognize certain patterns like cancer or organs. And it's only used to help doctors, it's not making decisions because it's not reliable. It's a quick sorting machine. But it's not a LLM at all.

And as others said: students were lazy but they had to work to learn something, now anyone can cheat at home and it will create an idiocracy in the next decades (and it already does with a few people around me, it's not a fantasy).

Every other AI is bullshit. But it still makes me feel good because if OpenAI can pirate all the books, I can keep on torrenting for free without feeling any guilt. Thanks ChatGPT!

I hate the endless grift that "AI" exists.

As far as technologies that are called "AI", it's just like any other technology. The usage determines the value. Under capitalism all technology is used first and foremost to violently enforce capitalism. This has already become apparent.

For example, while I enjoy generated art (apart from the environmental destruction), it's not the same thing as murdering palestinian children. The conflation of the two is grifter pseudo-science in service of evil.

Thank you for clearing things up for me. I have read every comment and I am amazed by how many aspects I missed

Well I hate LLMs

If you're talking about reading machine generated text then I'm to fucking old to eat a corporate propaganda. What's difference between AI and TV ? You can't turn off AI without turning off TV these days.
That's sad.

I hate the way people talk about AI. They’re going crazy at work pushing people to use it for everything while you have people doing less and less more poorly claiming they save 50%. Everyone is more efficient and has greater efficacy while nothing gets done.

AI is a great tool for programming - next step up from autocomplete, search, and templates. It can be useful to speed up some coding tasks but it’s rarely a final result. So I’m the Luddite where we have some non-technical manager trying to guide a bunch of us on zoom through a coding task with ai. Not only will she not listen but the only result so far has been clarifying that she needs to define requirements

And I almost missed lunch today spending almost two hours on a code review from someone who clearly had ai generate it and never tried to build, never wrote tests much less ran them, never actually put any thought into cleaning up the generated stuff to actually make it work. He had a bit of what I swear is Perl in his Java “code”.

I had to feed a write up through ai to translate it to something that “saved me time” by looking like ai wrote it

No. Copyright should be consistently enforced, pollution should be taxed, privacy should be protected, sites shouldn't be DoSed, but other than that I think it's kinda nifty on its own.

I like using chat gpt for npc dialogue in my DnD game. It can really fill out a character. Otherwise I don’t use it.

I don't hate AI, I hate the system that's using AI for purely profit-driven, capitalism-founded purposes. I hate the marketers, the CEOs, the bought lawmakers, the people with only a shallow understanding of the implications of this whole system and its interactions who become a part of this system and defend it. You see the pattern here? We can take out AI from the equation and the problematic system remains. AI should've been either the beginning of the end for humanity in a terminator sort of way, or the beginning of a new era of enlightenment and technological advancements for humanity. Instead we got a fast-tracked late stage capitalism doubling down on dooming us all for text that we dont have to think about writing while burning entire ecosystems to achieve it.

I use AI on a near daily basis and find usefulness in it, it's helped me solve a lot of issues and it's a splendid rubber ducky for bouncing ideas, and I know people will disagree with me here but there are clear steps towards AGI here which cannot be ignored, we absolutely have systems in our brains which operate in a very similar fashion to LLMs, we just have more systems doing other shit too. Does anyone here actually think about every single word that comes out of their mouths? Has nobody ever experienced a moment where you clearly said something that you immediately have to backtrack on because you were lying for some inexplicable reason, or maybe you skipped too many words, slurred your speech or simply didn't arrive anywhere with the words you were saying? Dismissing LLMs as advanced autocomplete absolutely ignores the fact that we're doing exactly the same shit ourselves, with some more systems in place to guide our yapping.