431
We Asked A.I. to Create the Joker. It Generated a Copyrighted Image.
(www.nytimes.com)
This is a most excellent place for technology news and articles.
I think AI in this case is doing exactly what it's best at: Automating unbelievably boring chores on the basis of past "experiences". In this case the boring chore was "Draw me [insert character name] just how I know him/her".
Too many people mistakenly assume generative AI is originative or imaginative. It's not. It certainly can seem that way because it can transform human ideas and words into a picture that has ideally never before existed and that notion is very powerful. But we have to accept that, until now, human creativity is unique to us, the humans. As far as I can tell, the authors were not trying to prove generative AI is unimaginative, they were showing just how blatant copyright infringement in the context of generative AI is happening. No more, no less.
Creativity can be estimated by AI with randomness, but what they don't have is taste to determine which of their random ideas are any good.
I dunno man ... assume a model trained on the complete corpus of arts leading up to the Renaissance. What kind of randomness lands you at Hieronymus Bosch? Would AI be able to come up with Gonzo Journalism or modal music?
A brief glance at the history of human ingenuity in the arts really puts generative AI in perspective.
I see what you're saying, but Bosch may not be the best example because frankly his paintings often look like the early AI fever dreams lol (I mean, not really, but you can see the resemblance). But seriously, with enough randomness you certainly could get that kind of output - there's really no reason why not - but it would take god knows how many iterations and the computer doesn't have everything other than the art to determine what is good.
It's monkeys and typewriters, yknow. You'll get there eventually even just producing random pixels (I mean, admittedly one limit will always be resolution unless you actually teach the AI to operate an arm which paints).
Yea, it really boggles my mind that we now have a way to automate boring jobs like data entry of drafting some mundane documents but what humanity decides to use it for is artistic expression, the one thing it can't really do properly. It's like NFTs all over again...
What's surprising, people want to create what they imagine, they don't have the skills and/or time to draw/render it.
This is such a strange comment. The vast majority of AI use cases are LLM use cases. Generative Art is just a novelty. Most of the money and research right now is going towards the useful automation tasks, not the novelty. That people are abandoning one for the other is not a reasonable conclusion.
And NFTs were stupid for a completely different reason. Nobody is trying to sell me AI shit like it's going to make me rich and special. And at least some NFTs had real artists behind them.
Hollywood execs have been salivating at the idea of just generating media with AI, there was a whole strike about it. Same with video games, I believe game voice actors got screwed with that AI deal.
Also NFTs had some tech that could have been useful but instead people chose to use it for creating a new speculative market riddled with scams. That's my comparison to AI, interesting tech used for a very wrong purpose.