I am afraid to meet this man, because I know that I'm going to be strongly compelled to bite off his eyelids.
Bite my eyelids off then:
if we actually achieve parity between AI and human output (and not just some breathless Hypebeast tech CEO
claiming that they've achieved parity) then no, I don't really care whether there's a "soul" behind an output: if a machine had made the film
The Lighthouse instead of Robert Eggers I fail to see what the film would lack as a result.
Or, to put it another way: people might bemoan the death of quality craftsmanship, but if a factory could churn out goods
indistinguishable from artisan furniture in terms of durability and aesthetic value and uniqueness, I wouldn't particularly care that it came out of a soulless machine.
All that being said, I don't think we'll reach that point barring some insane revolution in our understanding of intelligence: there will always be tells, always something off, something too rote, the slop will out itself as slop not through some immaterial aura of "unintentionality" but because it
looks and sounds and is constructed like the kitsch home decor in a Dollar General.
This quibble aside, I'm not eager for AI to reach the point of general human-level performance, assuming that is even possible, since that'd mean every non-dextrous task on the planet would get outsourced to a server farm, and also that propaganda would be terrifying: AI might be "good enough" right now for pushing a message, but imagine if some spook in Langley (or ex-KGB agent in Russia, or some Chinese thought-policeman, etc, etc) could spin up 20,000 virtual disinfo agents with actual human-level performance, and not just current LLMs' shallow simulacrum thereof. It makes me shudder.
And in the meantime, I wish current AI slop could be scrubbed from the 'net because it is ugly and uninformative, and I loathe the people who use it to try and make a quick buck off of e-pollution.