far too many people try to defend the technology as if the issues with “gen” AI aren’t inherent to LLMs as a whole. it’s tiring.
You must log in or # to comment.
skeletor invented serving cunt
I would appreciate an explanation why it is fascistic in particular. Is it about today’s AI companies? Is there something inherently fascistic in large scale algorithms that try to imitate humans? Valuing aesthetics of knowing the truth over the process of learning? I see the correlation, not as much the connection
I would say yes, yes, and yes. Mainly I’m trying to say that even if the environmental and societal factors were solved, LLMs would still be pretty fundamentally fucked, as I also said in another comment in thus thread.
I thought about this a while ago and while its more subjective and personal, I hope it helps illustrate what I’m getting at. If someone subconsciously, or even intentionally, put a piece of a song or words or whatever that I wrote into their own work, I’d feel downright honoured. I’d appreciate some credit if possible (I already think lack of attribution is an issue among creatives sometimes), but even if not, I’m happy knowing my art had enough of an effect on someone to weasel it’s way into their own! That’s amazing! On the other hand, if a piece of something I made similarly wound up in someone’s own “creation” because of an LLM, I’d feel empty, as it means nothing. Less than nothing, in fact. I’d be really upset by it.
I hope that helps?
the OC asked why its fascistic and you answered essentially with “when it outputs something trained on my data, it feels bad, thus fascistic”.
Maybe I’m missing something, but not liking something doesn’t make is fascist.
EDIT: I guess imma add this: fascism is bad. but bad doesn’t necessarily mean it’s fascist.
I can see where this guy is coming from, after all, the collection and conglomeration of everybody’s information is very fascist-esque, but reading their second comment I’m not even sure if that’s what they mean
They’re honestly kind of just rambling about it not feeling good to be plagiarised from, which comes across as a little odd
I was trying not to repeat what I already said in my other big response in this thread, so I just added another, more subjective point. Sorry, OC.
However, while its mostly my own perspective, I think its still based in the same argument: taking inspiration from someone through appreciation of their work is (usually) honouring it. If nothing else, it shows that person’s art impacted another’s. An LLM, though, does nothing more then shamelessly regurgitate bits and pieces of that and other works in an aimless fashion. There is no rhyme or reason or thread of inspiration to retrace.
Theres a reason fascists hate creativity, colourfulness, and differences between people: it goes against everything fascism stands for. LLMs were made with the intent to homogenize and steal from creatives, all while passing off output as at least somewhat original. I wouldn’t call Canada fascist, even though we do many fascistic things, like any country under capitalism. Not being strictly fascist doesn’t make those actions or laws any less so, however, and an LLM made by a non-fascist doesn’t make the concept of am LLM any less fascistic.
I don’t think you know what the word fascist means.
When I first saw the post I wondered if that was a carefully defined stance or a linguistic shortcut, and its fine both ways, really, but I was a philosophy student, this is my thing
“inherent” is a strong word, and perhaps fascistic tendencies (not convinced as to whether the whole thing is fascistic yet) are an inherent aspect to the kind of llm we are aware of. There’s a but: would it be impossible to create an opt-in LLM database? It wouldn’t happen in the world we currently know, but if we talk about inherent characteristics, that’s the exact thing we need to talk about (you don’t have to care about this at all, there’s no law about what words mean what, but it makes it easier to share ideas without confusion. And judging by the comments, there has been confusion). Do I think we could benefit from an LLM? While I hate chatbots with a passion, they dont have to be chatbots. Deep.ly is a translation only LLM, I used that back in 2018-2019 because that was the very best automated translator available for free and I had pages of technical stuff to translate (maybe they changed since then, idk). Translation can be both utilitarian and artistic, I certainly wouldn’t want a robot to translate poems, but even if, would that be fascistic?
AI art is bad, therefore fascism?
On the other hand, if a piece of something I made similarly wound up in someone’s own “creation” because of an LLM, I’d feel empty, as it means nothing. Less than nothing, in fact. I’d be really upset by it.
But why though? Personally, I would just feel nothing, maybe amusement that my work got mentioned at all. LLMs are exactly that — large language models, i.e. massive-scale mathematical algorithms whose inputs and outputs are natural language. They’re tools to solve problems.
In particular, I use LLMs extensively as a LaTeX assistant. (LaTeX is a programming language for typesetting technical mathematical documents.) I do the underlying theory myself, but I use an LLM to refactor my code and debug obscure errors. Learning the ins and outs of LaTeX is neither a very interesting nor enlightening nor useful exercise. I am happy to let an LLM take care of that crap while I go do something fun or useful with my time.
I absolutely recognize that many, if not most, people will not have any use for these tools, and I think it is obscene that the capitalists have hijacked society to put these tools at the center of our existence. That does not make the tools useless.
Additionally, I oppose the “artificial intelligence” framing when talking about LLMs, the broader machine learning field, and frankly the entire lineup of applied statistics starting from the 1950s that have lead us to this point. I do not need nor want truly intelligent machines, and capitulating to the capitalist framing that these machines are even converging towards intelligence is doing free advertising for the capitalists. In my view, these tools ought to be called “statistical learning” algorithms, where “learning” is defined as a group of related mathematical problems wholly and openly distinct from the processes that psychology, philosophy, and biology gives the same title.
And I really don’t see the moral problem with building machines in general to “learn” important tasks, especially ones that are not interesting or enlightening to do as a human. As the probabilistic framework of statistical learning and its offshoots make transparent, you don’t pick a statistical learning solution when your problem has a well-defined algorithmic solution process. You also don’t pick a statistical learning solution when you absolutely require an exact solution in every single run. Typically, you pick a statistical learning solution when a near answer is better than none at all. E.g., if I want to add a custom environment to a LaTeX document, I don’t really care about the exact trajectory that the LLM takes in its state space, as long as it produces a useful approximation of what I asked for in finite time.
Lastly, I want to remark that large language models are, nowadays, part of more complex systems. In particular, LLMs can chain together “classical” bots, classical deterministic algorithms, pre-existing code generation technologies, and even a Linux container to perform certain tasks. So when the technology first came out, if you, say, told the machine to work on a piece of LaTeX code, it would actually just use its large language model to auto complete the most statistically likely response, and it would do this based on the fact that it has LaTeX language references in its training. So it was a roll of the dice whether or not the code would compile, and the machine had no way to check it. Nowadays, I can actually give an LLM the ability to compile LaTeX code. Claude actually has a minimal LaTeX compiler in its Linux container, but I can actually give it access to my actual local LaTeX compiler (I do this inside my own virtual machine, I wouldn’t do this on my actual hardware). So the LLM can literally check its “work”. Even locally hosted LLMs can be given access to external tools. In particular, if you give it SymPy access (SymPy is a Python library for symbolic mathematics), LLMs can “like magic” be decent at math (although they still make mistakes!!!).
My best guess is that the argument is “the scale of data needed to train an LLM can never be obtained in a manner that isn’t oppressive or stealing by it’s scale” and thats not really true. there are a lot of open works, works which allow using them in datasets, etc. there are ways to get permission from people, give attribution, and just be honest about everything. the reason its not happening is because we live in a hell of a world with oppressive systems that benefit from hiding and manipulation
there’s no reason it cant be better in a better world
if the argument is generating things like drawings, renders, writing, music, videos, are inherently bad than that’s not really saying all LLMs are bad
accessibility LLMs should be a part of a better world at least. what changed my perspective is seeing someone irl who didnt speak english and just voice messaged chatgpt in their language and it translated that to me in english. was it perfect? i doubt that. do i support chatgpt? fuck no. do i believe something like this can be done properly and for public good? yes
Translators existed before LLMs.
Maybe the joke went over my head, but could someone explain how they are fascist?
I get that they are kind of a disaster for the environment and people’s brains right now, but fascism seems unrelated
All of the LLMs are built by fascists. Elon is the most obvious, where he retrains it every time Grok contradicts him or his fascist beliefs
I agree, but doesn’t that imply there could be an LLM built by someone who’s not a fascist? So the LLMs are not inherently fascistic?
I mean, there are open source models out there…
I would argue that a machine attempting to feign human intelligence, devaluing actual human intelligence in the process, is fascist. What used to be referred to as AI, like the patterns of enemies in a video game, are still authentic creations. Taking those creations and feeding them to a glorified auto complete guessing machine (with or without the creator’s consent) is an attempt to homogenize humanity itself. God, that sounds overly dramatic, but I can’t think of another way to word it. Apologies.
Basically, even putting all the real world consequences aside, the difference comes down to automating mostly menial tasks vs. handing over your agency to a machine that only supersedes a magic eight ball in it’s ability to perform slight of hand. Even if all the societal and environmental factors could be put to rest at this very moment, it’d still be a fundamentally evil concept, I feel.
I thought fascism was just extreme conservative authoritarianism.
The idea of AI is actually kinda progressive, but it’s just being driven and controlled by crazy capitalist right now
I think the idea of a machine that feigns intelligence is the opposite of progressive.
As individuals, the ideas and experiences of others should inform, but not solely dictate, our own views, right? Therefore, someone who spouts nothing but the ideas of others, even if citing each one, is unwilling to think for themself. If that person doesn’t cite their sources, it means they’re unoriginal and a plagiarist.
If we want progress towards a better world, LLMs are the last things we need. A black box that dulls our minds by pretending to be “intelligent” when it could never possibly be so while causing copious damage in every sense of the word doesn’t sound like progress to me. At best, it’s what mainstream social media algorithms do, but on steroids, and while social media itself isn’t evil, those absolutely are.Maybe calling all LLMs fascistic is an exaggeration, just like equating fascism with capitalism. Two different words exist for a reason, they are different ideologies, but the latter always benefits/leads towards the former, and LLMs share that relationship with fascism.
I can imagine fun, useful, and interesting, applications of LLMs.
It’s just that companies are going kinda crazy over investing and trying to shove them in everyone’s face for profit.
Does that include non-US LLMs like Mistral? Not that I have a high opinion of even those kinds of LLMs, but I’m not sure whether “fascistic” really captures the inherent issues of LLMs, vs. the inherent issues of US tech giants.
Models themselves, when not told to be any different, are usually left leaning and quite empathetic.
came here to mention exactly that. and how elon has had a really hard time getting grok to not be left and against him.
I honestly just dont see it. It does have the tendency to concentrate power to the wealthy top (just like factories do) which is fashy but that isnt an inherent property of LLMs. I read your other comment and I just dont get it.
How long until we start calling programming languages fash as well? LLMs are a tool and is no more fascist than the MP-40 used by Nazi Germany. LLMs don’t hold ideals much like a sub-machine gun doesn’t. Now are LLMs used by fascist? Yes obviously, but saying that fascism is inherit to a tool is just wrong. It can be used for fascist purposes, but it depends on the person using it.
Slippery slope fallacy. Some tools are inherently wrong to use. Since we’re talking about weapons of war, you could never make this argument about napalm, for instance. The flaws of LLMs are inherent to them and can’t be painted over, no matter how perfect the paint or how well intentioned the painter.
Definitely under a capitalistic economy (and ALL of the world is capitalistic right now, including China). Though if needs were met for all, and benefits shared by all, I’d say even LLMs have a place in UI/UX development, and certain parts of art, even.
It’s not bad to have certain interfaces and interactive elements have a bit more human in there. Though it is VERY bad for these corporations to crash the tech economy by skyrocketing component prices and firing people that can do a FAR better job than the slop machines…
The biggest stumbling block to most people understanding this is really accepting the idea that “the essence of technology is by no means anything technological”.
(Or if leveraging the ideas of a dead Nazi to analyze fascist artifacts creeps you out, lean on “do artifacts have politics?”)
But you’re also working against centuries of propaganda about Luddites, the industrial revolution, automation, capitalism, the commons…
Edit: Maybe most of all, the pervasive logic of neoliberalism that says modes of production can be cleanly separated from politics.
Yeah. I’m shocked to see so much… well, not neoliberalism, but neoliberal arguments under this post. I really thought this would be uncontroversial in a place like this.
I think Lemmy is generally on the correct side as far as policy when it comes to LLMs, but I think the user base generally see themselves as individual rational actors in a neutral objective shared reality. It’s a pretty common perspective in tech circles, especially ones that are enthusiastic about individualistic liberty.
That frame makes it very difficult to conceive that a thing — particularly a stateless bundle of numeric weights — could exert any kind of political influence just by existing. Let alone conceiving that an entire class of things could inherently have a particular political direction.
Because it seems as though “the difference between a tool and a weapon is the hand that holds it”. But we should know at this point, if we’re willing to be honest about the limitations of the firmware that runs inside our skulls, that weapons can wield people too. And when you have a tool in your hand, the tool also has you in its hand.
They already broke the Hippocratic oath, so, “AI” is indefensible.
The viruses are specific kinds known as bacteriophages, which only infect bacteria and are used around the world to treat patients with persistent infections. In lab tests, a cocktail of the AI-designed viruses killed E coli bugs that were resistant to natural bacteriophages.
I.e. these viruses are intended to help patients. Not a Hippocratic Oath violation (yet).
But also, if someone did use an “AI” to design harmful viruses, that wouldn’t be the “AI” breaking the Hippocratic Oath. We don’t make heart monitors and MRI machines take the oath; we make humans take the oath. (And for the record, I oppose the “artificial intelligence” framing; in my view, these are derivatives of the statistical learning framework, a mathematical theory whose definitions, while technically daunting, are otherwise transparent about the limitations of these technologies. Like them or not, calling these systems “intelligent” capitulates to the capitalist framing that these systems, which they control, deserve or eventually will deserve the same status as humans, which would thus amplify their power.)
Like people use ordinary power tools to kill each other all the time. Does that make power tools indefensible? Of course not.
