• Sagan_Wept@sh.itjust.works
    link
    fedilink
    English
    arrow-up
    5
    arrow-down
    1
    ·
    2 days ago

    Sorry to be pedantic; LLMs aren’t the ones making images though it’s still diffusion models. By definition LLMs are language focused. Again sorry, I feel like I’m being an asshole by explaining this, but I fell its important to understand this tech.

    • Opisek@piefed.blahaj.zoneOP
      link
      fedilink
      English
      arrow-up
      3
      ·
      2 days ago

      All good. I do know the difference very well, but I generally refuse to use the term “AI” and abbreviations more specific to image generation are not as well-known, not to mention there are lots of different approaches to it, too. I guess you could group them under NN or DL, but I digress. I opted for LLM because nowadays people don’t really interface with the image generation models directly anymore, but go through an LLM chat app, which then uses a different model in the background. Perhaps I should’ve used the word slop instead. Oh well.