@longlongfern@YourSirJoffrey@tumble_wood@thinkingshivers When you send an image into Chat GPT, for example, it first converts the image into text, I think using CLIP, and then uses that text description as the input to Chat GPT. It doesn't see the image at all. There's separate models going on, at least as it stands now.
@longlongfern@YourSirJoffrey@tumble_wood@thinkingshivers No and this is exactly the point OP is making. OP is explicitly talking about how LLM based image generators work. A person wouldn't be able to do it and neither would an AI. Humans have the benefit of truly existing in and experiencing the world which gives us a leg up over AI.
@the_boring_dad@skdh Not really, trying to turn GR into a QFT leaves you with a theory that's just as good as GR but also diverges at the planck scale even with renormalization. So it's basically just GR but more complicated and offers very little insight over the classical theory
@AkaPorco@0bNARA@thot_criime Understood. I will say it definitely feels anti-reading, and I believe most people arguing against you seem to see it that way.
@AkaPorco@0bNARA@thot_criime I just sat through a 20 minute video people were saying was a good explainer on what was wrong with the scishow video and it literally had no info, maybe two minutes total of anything actually descriptive. But no, I have to watch 3 more videos or I'm ill. Just let me read!!!
@AkaPorco@0bNARA@thot_criime Bruh what if I just dont want to sit through 30 minutes of reactions, jokes, opinions, background, arguments and just a small fraction of actual info when I just want to know literally what he said and why it's wrong 😭
@papitto_p49446@HormuzLetter Pretty sure they can see the group of armed American guys in American military gear with American flags on themselves and their American military vehicles that come with them. I dont think cloaking technology is that advanced yet.