Do you get the feeling that AI generated content is lacking something that can be incrementally improved on?

Seems to me that it's already quite good in any dimension that it knows how to improve on (e.g. photorealism) and completely devoid of the other things we'd want from it (e.g. meaning).

▲

tomrod 4 days ago | parent | next [-]

It's missing random flaws. Often the noise has patternd as a result of the diffusion or generation process.

	▲	arcane23 3 days ago \| parent [-]
		Yeah, I was thinking about this. Humans vary depending on a lot of factors. Today they're happy, tomorrow they're a bit down. This makes for some variation which can be useful. LLMs are made to be reliable/repeatable, as general experience. You know what you get. Humans are a bit more ... -ish, depending on ... things.

▲

keiferski 4 days ago | parent | prev [-]

Yeah if you look at many of the top content creators, their appeal often has very little to do with production value, and is deliberately low tech and informal.

I guess AI tools can eventually become more human-like in terms of demeanor, mood, facial expressions, personality, etc. but this is a long long way from a photorealistic video.