I usually don’t even understand the lingo they use. “Open-weighted” is the most recent one, then it usually goes down to specific “models” that everybody is supposed to know about.

These are my thoughts (I will stick to the vague “it” for now, but of course therein lies another question: “and how does all this apply to various specialised AIs”):

  • Is it really feasible to run it 100% locally? I know there’s plenty of people with very powerful rigs indeed, but still. Or are 99% of these people really saying “it would, in theory, be possible to run that locally, therefore your concerns are invalid”?
  • If yes to the previous: the software doesn’t come from nowhere and ultimately still relies on gas-turbine-powered datacenters and stolen IP and stolen personal data, no?

If what I wrote above is true, what exactly are people arguing when they say it’s still possible to use LLMs ethically or true to FOSS philosophy, because … ???

edit

Thanks to all who answered.

I guess it’s my fault for asking several questions in one, but this thread has attracted exactly the type of people I’m writing about; several even used the term “open-weighted models” without explaining it.

Asking to get arguments explained, I got more arguments instead.

  • iForgotSpells@sopuli.xyz
    6·
    13 days ago

    You can run them locally, yes. There are models that can even run on phones, but usecase is limited. But it can only be considered ethical, if the training data used is listed or ethically sourced IMO.

    AI bros on Lemmy will disagree with me, but most open weight models are still trained unethically i.e, theft. Most proponents of LLMs (who I talked to on bsky), who say local models are ethical, don’t fucking use it. They’re larping on socials about how awesome it is, but none of the ones I talked to are using it in their projects. They mess around, realise it is not as good as the “unethical” options, go right back to Claude

    Open weight models Qwen, deepseek, mistral, and the Ollama stuff etc are unethical in normal people’s eyes, but “ethical” enough for AI bros.

    From what I searched, there are very few that can be considered ethical - Olmo, Apertus, Starcoder(?). But idk anyone who uses these. My friend at IBM said they used Apertus, but it was nowhere near good as ChatGPT, so they no longer use Apertus now. And these models require minimum 6-8 GB VRAM for their lowest parameter model iirc.

    Even the open-weight model bros are lobbying to redefine what ‘open-source AI’ means. That should give you a fair idea about people behind open-weight as well

    • Dran@lemmy.world
      2·
      13 days ago

      I would unironically argue that a model primarily trained through distillation of closed frontier models, and then released open-weight with an open-source architecture, becomes “ethical” again.

      Something something Robin Hood