this post was submitted on 20 Sep 2023
476 points (93.0% liked)

Technology

58143 readers
5618 users here now

This is a most excellent place for technology news and articles.


Our Rules


  1. Follow the lemmy.world rules.
  2. Only tech related content.
  3. Be excellent to each another!
  4. Mod approved content bots can post up to 10 articles per day.
  5. Threads asking for personal tech support may be deleted.
  6. Politics threads may be removed.
  7. No memes allowed as posts, OK to post as comments.
  8. Only approved bots from the list below, to ask if your bot can be added please contact us.
  9. Check for duplicates before posting, duplicates may be removed

Approved Bots


founded 1 year ago
MODERATORS
 

The majority of U.S. adults don't believe the benefits of artificial intelligence outweigh the risks, according to a new Mitre-Harris Poll released Tuesday.

you are viewing a single comment's thread
view the rest of the comments
[–] snooggums@kbin.social 3 points 1 year ago (2 children)

AI won't be creating anything new anytime soon, because it recycles existing art just like hack writers do now. The "best" art tends to require a supporting story, which AI won't have. Comedy changes constantly, and AI won't be any better than people trying random stuff.

You don't question your existence because other people are smarter or better at doing things, right? Is most of humanity not of any value because they aren't the best at everything?

[–] lloram239@feddit.de 2 points 1 year ago (1 children)

AI won’t be creating anything new anytime soon

It already has.

The “best” art tends to require a supporting story

ChatGPT can write that. Multi-modal models that combine text generation with audio and video are months away.

AI won’t be any better

Those claims have the tendency to not age well.

You don’t question your existence because other people are smarter or better at doing things, right?

Humans aren't that much better than me and not doing the things I want to do. AI on the other side will be much better than me, as well as do exactly what I want it to do and will be a click away.

And yeah, I had numerous experience were I would question my existence when playing around with ChatGPT or StableDiffusion. Neither of them is quite good enough yet, but they are very much on a trajectory where you can see that you have zero chance of competing with them in the future, or even getting remotely close.

The fact that we got them in the first place, not from humans doing centuries of research on art and language, but by simply by throwing huge amount of training data at AI algorithm, should be enough to question your existence.

[–] snooggums@kbin.social 1 points 1 year ago* (last edited 1 year ago)

AI writing a fictional background story about how it came up with some piece of art is not the same thing as multiple researchers telling the story of an artist. Neither of your examples are something someone couldn't do, because whoever prompted it could have done the same thing and just had not yet.

You are completely missing the point that great art is generally supported by the context of how it was made and not the end result in a vaccuum.

[–] flossdaily@lemmy.world 1 points 1 year ago (1 children)

I understand why you think that, but what you have to remember is that every great piece of art you've ever seen has been derivative of something before it.

For example, I think of the Beatles as musical geniuses. But they are the first to admit that they stole other people's ideas left and right.

Beethoven's 9th symphony is this piece of transcendental music, that was widely considered at the time to be the greatest symphony ever written.

But if you listen to Beethoven's works over time, you see that the seeds of that symphony were planted much much earlier in inferior works.

Genius and creation aren't what we think they are. They are all just incremental steps.

[–] snooggums@kbin.social 2 points 1 year ago

That is overly reductive and conflates copying (like a cover band) and creating something new (being influenced). Heck, even when some bands play new versions of existing songs they are adding their own personal touch and have the possibility of making it mean something new. Like how Hurt by NIN and Johnny Cash are the same song, but how they are performed ends up being about completely different experiences.

Even when bands like Led Zeppelin outright covered existing songs they added something to it that AI can't, and won't be able to do. AI can't have sexually charged energy that a human can have. They can pretend to, like how cover bands can pretend to be like the band they are covering, but AI won't be able to replicate the personal touch that memorable art has.

Even popular stuff with widespread appeal frequently drops off over time because it isn't the type of art that holds up over time. Hell, the Beatles mostly hold up more for when they were popular and how they have managed their legacy than any kind of technical prowess in musicianship. Without their performances, their personas, and the backstory to most of their music it is just well done music that has been superseded musically since that time. None of that will apply to AI, and without the backstory it will just end up being high quality music that won't stand the test of time because we don't have any context for it.

Hell, there were a ton of other composers during Beethoven's time that were putting out great music too, but you know who he is because of details other than his musical prowess.