[ad404b8a372f2b9]

This is not remotely the same, scale and barrier to entry matter. With stable diffusion I can pick any artist right now and create over 1000 derivative works by tomorrow morning in his style to the same degree of expertise with no training involved and no work required.

[jstanley]

That's good!

Acting like it's a bad thing is just ludditery.

[heavyset_go]

The Luddites weren't some cult of ignorant technophobes, they were highly-skilled middle class craftsmen and small business owners who went from being able to provide for their families to dying in utter destitution. The remainder of them were tried for machine breaking and were either executed by the state or exiled to penal colonies. They risked everything because everything was at stake, I have a hard time saying that their situation and outcomes were "good", and I have a hard time saying the same about similar situations that are playing out today.

[knome]

Certainly, but automation is what allows for improvement to the whole.

Today, clothing is cheap and plentiful, along with bedding, curtains, towels and other cloth materials. Clothing would be outrageously expensive if everything were still hand spun, hand loomed, hand cut, hand sewn, and hand screened.

If the human computers[1] that predated the rise of the machine computer had done the same and won, it would have certainly been a boon for them at the time as well, but to the loss of all information technology developed hence.

The lack of a social safety net lent desperation to the luddites. Had they not faced imminent ruin and starvation as the machines eclipsed their occupation, they may not have had need to rebel against the newly emerging textiles mechanization.

AI may eventually replace traditional artists in many situations. Surely with simple images today, we can expect video examples in the future, and interactive AI generated simulations some time after that.

Do we smash the data centers now to save the artists' livelihoods and thereby avoid a future where anyone that wants can talk a computer through creating entire interactive fictional worlds via the synthesis of AI generation with feedback from their imagination?

[1] referring here to https://en.wikipedia.org/wiki/Computer_(occupation)

[trention]

>Do we smash the data centers now

Yes, the sooner, the better.

[ad404b8a372f2b9]

I wouldn't be so confident one way or another, this is too new. I think it's going to make a lot of things way more accessible and enable people to express their creative voice who couldn't before. On the other hand you're looking at the destruction of a lot of professions, and possibly overnight with the speed things are moving at. I think if we told every software engineer their skills were entirely obsolete and they had change career tomorrow the reception would be much colder.

I remember when I started working on generative models in 2015, you could barely generate a picture of a blurry 40x40 pixels face. Two years later 1024x1024 almost indistinguishable from reality. Now every week we have a new revolutionary application coming out.

[9wzYQbTYsAIc]

I think that the argument, overall, is that there are questions as to the legality of certain applications of the technology.

Society needs to change the laws regarding the preservation of value of intellectual labor, as has long been suggested.

Acting like the law doesn’t matter is a bad thing, if we are making value judgements.

[natch]

So if a particular convolution appears often, how would you assign value to that? Any ideas?

[9wzYQbTYsAIc]

Indeed, see https://news.ycombinator.com/item?id=33194623 for some initial thoughts on how this problem (and others) could be rectified

[natch]

So, an AI Karen?

>the generative AI system censors any output containing a category of error

I can see the ruleset now…

"This picture of a woman is revealing hair, so it must be censored because it is objectionable to some people and we must respect all people whose beliefs are guided in a sanctioned way."

"This picture shows unadulterated fun, which must be censored because…"

etc.

https://www.youtube.com/watch?v=IFe9wiDfb0E

[9wzYQbTYsAIc]

Or more to the point, “this picture contains copyrighted material, which must be censored because…”

etc.

I tried to be as general as possible.

The training data for a self-censorship neural network could be as robust as any given society would like.

An algorithm based on self-censorship of generated output wouldn’t require censorship within the training data for the generative neural network.

I can imagine some other advantages to that approach.