all 3 comments

sorted by: hot top controversial new old
[–] 1 point 10 months ago

Nice post.

If we are stuck with this devil, we should bend it to improve the world. Organize a union, build a better world.

  • source
  • [–] 1 point 10 months ago* (1 child)

    I disagree. This is like blaming atomic power for the cold war or the internal combustion engine for climate change or democracy for Trump (or the printing press for the moral decay of 15th century Europe, or the telescope for the Copernican heresy). Underneath virtually all modern problems are human beings making (often profit driven) human decisions. People have blamed the tools for centuries. But we need to look at the humans hiding behind the curtains and hold them accountable if we want to actually solve problems and change things.

    People love to hate on generative AI, and there's no doubt that generative AI is causing a host of absolute garbage outcomes in our modern world, but the problem is not that people use it to write code faster or make some pictures they wouldn't be able to make without it. The problem, as usual, is big corporations making big corporation decisions while the human beings that benefit from those decisions are mostly invisible and unnoticed and even when they're not are totally unaccountable.

  • source
  • hideshow 1 child comment
  • [–] 3 points 10 months ago

    You can use a hammer to build a cupboard, or to bash someone's skull in, but you can't use it to make cupcakes with (well, not very effectively, anyway, or hygienically). My point is that each tool has a limited set of purposes it can really be used for and there's no law of nature that states that all things considered "tools" always have more good purposes than bad, or that the benefits of the good purposes always outweigh the problems caused by the bad.

    So it's not good enough to flatten "AI" to the broader category of "tools" and say because something is considered to be generally true about the category as a whole, that means it's also true for this specific case; you actually have to look critically at the specific case: who does it empower, in what way, to what extent? And frankly, the ability of the current paradigms of generative AI to empower good people to do good has been minimal to non-existent, whereas bad people have been greatly empowered to do bad. People who do not value truth and to whom the end justifies the means now have an infinite propaganda machine, those who do value truth do not. So the intentions of the people who made the AI isn't even the biggest problem (though it does make things worse), it's the intentions of the users. A community-made hammer is just as effective at bashing skulls in as one made by a greedy corporation.

  • source
  • parent
  • 19 comments in c/privacy@lemmy.ml view that post
    [–] 11 points 10 months ago (9 children)

    This doesn't make sense when you look at it from the perspective of open source models. They exist and they're fantastic. They also get better just as quickly as the big AI company services.

    IMHO, the open source models will ultimately what pops the big AI bubble.

  • source
  • hideshow 9 child comments
  • [–] 3 points 10 months ago (7 children)

    Right, Betamax much? It doesn't really matter if one technology is objectively "better" on all aspects than another if the strategy to make it popular outpaces the other.

    To be clear I wish you were right (even though I don't find open source models to be free of problems) but I think the conclusion is a wish, not a logical one.

  • source
  • parent
  • hideshow 7 child comments
  • [–] 7 points 10 months ago (6 children)

    How is that wishful thinking? Open models are advancing just as fast as proprietary ones and they're now getting much wider usage as well. There are also economic drivers that favor open models even within commercial enterprise. For example, here's Airbnb CEO saying they prefer using Qwen to OpenAI because it's more customizable and cheaper

    “We’re relying a lot on Alibaba’s Qwen model. It’s very good. It’s also fast and cheap,” he said. “We use OpenAI’s latest models, but we typically don’t use them that much in production because there are faster and cheaper models.”

    I expect that we'll see exact same thing happening as we see with Linux based infrastructure muscling out proprietary stuff like Windows servers and Unix. Open models will become foundational building blocks that people build stuff on top of.

  • source
  • parent
  • hideshow 6 child comments
  • [–] 4 points 10 months ago (1 child)

    Working on (some) AI stuff professionally, the open source models are the only models that allow you to change the system prompt. Basically, that means that only open source models are acceptable for a whole lot of business logic.

    Another thing to consider: There's models that are designed for processing: It's hard to explain but stuff like Qwen 3 "embedding" is made for in/out usage in automation situations:

    https://huggingface.co/Qwen/Qwen3-Embedding-8B

    You can't do that effectively with the big AI models (as much as Anthropic would argue otherwise... It's too expensive and risky to send all your data to a cloud provider in most automation situations).

  • source
  • parent
  • hideshow 1 child comment
  • [–] 2 points 10 months ago

    I'm actually building LoRAs for a project right now, and found that qwen3-8b-base is the most flexible model for that. The instruct is already biased for prompting and agreeing, but the base model is where it's at.

  • source
  • parent
  • [–] 4 points 10 months ago* (3 children)

    It always horrifies me a little bit how much open source has been exploited by large corporations for profit while much of the open source tech they rely on they do not invest in. Meaning by and large it often feels like open source has been unintentionally the largest transfer of the wealth created by labor to the corporate class in human history because labor had lofty ideals and capitalists are happy to exploit that.

    Linux has the majority share of corporate servers and has for a long time, and yet is barely cracking 3% of the desktop (consumer, laborer) market. Corporations profit wildly from open source while the general public has not.

  • source
  • parent
  • hideshow 3 child comments
  • [–] 5 points 10 months ago (2 children)

    I very much agree there, but think of how much worse it would be if we were stuck dealing with proprietary corporate tech instead.

  • source
  • parent
  • hideshow 2 child comments
  • [–] 3 points 10 months ago* (1 child)

    Oh of course, I'm not saying we should dump open source, closed source is so much worse, it just sucks how much the great ethics of open source are exploited by those with no ethics.

  • source
  • parent
  • hideshow 1 child comment
  • [–] 3 points 10 months ago

    Yup, and this is precisely why it was such a monumental mistake to move away from GPL style copyleft to permissive licenses. All that achieved was to allow corporations to freeload.

  • source
  • parent
  • [–] 3 points 10 months ago*

    Agree, eg. Apertus (FOSS) is a great choice, but even some indie ones are pretty private, eg.Andisearch. AI can be great to improve our research, work and creativity, but it's bad if we use it to substitute our research, work and creativity. But yes, avoiding the AIs from big (US) corporations, which use it to spy and log user data.

  • source
  • parent
  • [–] 13 points 10 months ago* (last edited 10 months ago) (5 children)

    I mean... it seems painfully obvious and doesn't need much of a thesis behind it.

    The wealthy want their slaves back, but they want slaves that don't push back, never ask for more, never need a day off, don't need sleep, don't need breaks, and are needlessly sycophantic to stroke the egos of the wealthy. It's no more complex than that: the promise of LLMs was that they could have deeply exploitable knowledge workers without any of the fuss or mess of humans who want a life outside of their fucking jobs.

    Like what else has this ever been? It's been transparent since day one that this is why every business pushes AI adoption so hard, for them it has to work, they're willing to bet the future on it because they think their sheer belief in it and throwing money at it will eventually "make it work."

    On the plus side, anyone who understands LLMs understands their limitations and the problems that are baked in to how they work and how those issues can't be "fixed." So this dipshit ass all-in plan that the wealthy have is doomed to crumble because it's never going to work the way they want it to. So we've got that going for us.

    Anyway I hate tools being described as "tools of the ruling class" because it often misses the point of how such tools can be useful to the proletariat as well. Class solidarity is a tool of the ruling class, but class solidarity would be golden in the hands of the proletariat, who vastly outnumber the wealthy class and ruling class. All tools are useful, what makes a tool dangerous is who wields it and what they choose to use it for. A hammer can be used to build and it can also be used to smash in someone's skull. Tools aren't the problem: specific dangerous humans are. I don't actually have huge problems with AI LLMs providing they are open source and rolled out small scale on home PCs, I just have an issue with their industrial applications at scale and the attempt to use them to consolidate power and control. They don't have to be used that way.

  • source
  • hideshow 5 child comments
  • [–] 3 points 10 months ago (1 child)

    they also want skilled labor; they are happy to have it scrape from authors and artists - primarily the same groups who the ruling class despise.

  • source
  • parent
  • hideshow 1 child comment
  • [–] 4 points 10 months ago*

    I mean really they despise anyone with skills because the reality is they have hardly any themselves as they've spent their lives paying for everyone else to do everything for them. They can't make a meal, they can't drive a car, they can't do basic appliance repair, they don't know how to actually use a computer other than social media, they can't wash their own clothes, they can't do anything for themselves. They despise every skilled person because it betrays their egotistical view that they are simply born better than everyone else and deserve to never have to know how to do anything at all. It reveals they know nothing and are useless to society at large, just a drain on the rest of us.

    Secondly, I did say "knowledge workers" and I personally think authors and artists are a type of "knowledge work" as they require knowledge coupled with skill to do the work, just as people managing servers and databases also require a combination of knowledge and skill. Poetaetoe pohtahtoh.

    Thus it's also why they are all pushing hard for humanoid robots because they want to automate the human body after they have automated the human mind.

  • source
  • parent
  • [–] 2 points 10 months ago (2 children)

    A hammer can be used for good or bad but a shrapnel bomb can only kill and maim; not all tools are multivalent like that, some are realistically only used for evil. Additionally, divorcing discussion of a tool from the social context like this is blinding to the real consequences. We don't have small scale LLMs running on personal machines (even if we could, that's not how it is now), we have them at industrial scale controlled by a few billionaires. It's purely fantastical thinking to say that just because you could imagine a world where they'd be used for good, that actually means anything.

  • source
  • parent
  • hideshow 2 child comments
  • [–] 3 points 10 months ago* (last edited 10 months ago) (1 child)

    We don’t have small scale LLMs running on personal machines (even if we could, that’s not how it is now)

    Uhhh, it's pretty trivial to set them up. I have a local Ollama instance set up on my PC with several different open source models available right now. Just because not everyone is doing it doesn't mean it's not possible. I don't even have an especially fancy computer, either. Ryzen 7 3700X, 32gb RAM, Radeon 6600XT 8gb video RAM, not exactly top of the line. I struggle with programming logic sometimes, so I use it to help me figure out if I'm doing something right or not when I can't find an answer online, an activity I wouldn't exactly classify as "evil."

    I also wouldn't consider a shrapnel bomb a tool, it's strictly a weapon just like a gun is strictly a weapon.

  • source
  • parent
  • hideshow 1 child comment
  • [–] 1 point 10 months ago

    I get all that, I just mean you can't just ignore the social reality. Like you said elsewhere in this thread, Linux is barely cracking 3% of the desktop market. Can you run an LLM locally? Of course. Do people do it? No, except a couple of hardcore enthusiasts. That is an issue which can't just be cleanly separated from the technology itself and has to be taken into account when discussing it.

    Also, weapons are tools.

  • source
  • parent
  • [–] 4 points 10 months ago

    Instead I say we let ai play with the unreal government pyramid scheme that is money while we utilize the money in the same way as we already do, essentially let the AIs do the boring stuff.

    And then much later on when there is risk of not finding a niche that needs individuals we then work on creating automated labor so that we need not worry about responsibility as individuals while the AIs play present day jobs in the same we we presently do.

    At this point income and responsibility are solved issues all while not even remotely changing how society works other than what we as people do day to day so that way we may finally consider living life as we please instead of being beholden to fixing problems of the world.

  • source
  • [–] 3 points 10 months ago

    It's always about AGENCY and power, not performances or preferences.

  • source
  • [–] 2 points 10 months ago

    Reads like a communist hiding their power level or a liberal searching for a take on the enclosures actively happening this very moment that isn’t the fascist/libertarian one (“it’s different because it’s happening to me!”).

  • source