If you had hidden cameras, you could gather data on how the language is being used in various scenarios. That might be enough for an ai to use to “learn” the language
Does this mean we can, indeed, train AI on ChatGPT/other AI outputs?
More like 1.5 total. He did "parts of 3 days " so he didn’t even parallel Jonah properly (who was explicitly 3 days, 3 nights)
Can we also get a new.lemmy.world? Having New Reddit’s mobile UI would make using lemmy easier on mobile
This, especially if you have NVIDIA.