Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Copyright infringement in AI art

technollama.co.uk

Post reply on HN
Print view View original

51–60 of 155 posts

Re: Copyright infringement in AI art

#51

Post by trention »

matheusmoreira wrote: post #17
olladecarne wrote: post #11

ML revolutionized translation a long time ago and the demand and pay for translators went down. It also used the works of thousands of humans who translated text and they never saw a dime. Copyright applies to translations so we've already gone through something similar. The same will happen with art and every other medium that ML touches. I have a love/hate relationship with ML because of this. It seems that there w…

> I hope someone can engineer a new economic system that works for what's coming. What's coming is post-scarcity for all kinds of intellectual work. There will be nothing to economize: supply of both labor and product will be infinite. I don't want corporations engineering a made up economy where there is none.

>post-scarcity for all kinds of intellectual work

Unfortunately, all intellectuals also need to eat, bathe and clothe - areas where no "post-scarcity" is on the horizon (with climate change, both food and water may become less abundant). So an economic system will need to be invented.

Re: Copyright infringement in AI art

#53

Post by practice9 »

vintermann wrote: post #30
fxtentacle wrote: post #14

Earlier quoted context omitted.

Oh but you can. I've scanned the LAION training data set (which was used for mini DALLE, Imagen, and Disco Diffusion) and I recognized several images that long to business partners of mine. How does an AI know what Obama looks like? It memorized thousands of images of him, most of which were by professional photographers and, hence, copyrighted.

Well, it probably didn't memorize any single of them. One thing that has changed dramatically from back when I studied AI (a whole 5ish years ago, sigh...), to my shock, is that the number of training epochs has dropped dramatically. Sometimes down to just one. If it's only ever looked at your image and updated its weights once , could that really be enough to copy it? Only if virtually everything in your image is al…

This is a very interesting thought. If the image is very generic, IMO it should not be copyrightable, it should be public domain.

Re: Copyright infringement in AI art

#54

Post by angusturner »

trention wrote: post #49
angusturner wrote: post #46

The question that fascinates me about this whole debate is on what basis (if any) one draws the line between a machine learning from copyrighted data vs a human learning from copyrighted data. And I do think “learning” is the correct metaphor here. There are good reasons to believe that the brain is doing something conceptually similar to generative statistical models. i.e compressing inputs (/percepts) into some kin…

On the basis that the "machine" is not a human and there is nothing that necessitates allowing neural networks to do all the things humans are permitted to do.

I don’t think it makes sense to talk about these systems as though they have any agency (and I think I was careful not to).

Of course there is nothing that necessitates letting these systems do what humans do.

What I am saying is that I am allowed as a human to ingest copyrighted material, learn from it, form mental abstractions and use those learnings to generate new art. Why then, would I be forbidden from externalising some of that process into an artificial system that operates on similar principles.

The trained model obviously does not have agency. It is a creative and computational tool, much like any other digital tool or product.

Re: Copyright infringement in AI art

#55

Post by trention »

olladecarne wrote: post #11

ML revolutionized translation a long time ago and the demand and pay for translators went down. It also used the works of thousands of humans who translated text and they never saw a dime. Copyright applies to translations so we've already gone through something similar. The same will happen with art and every other medium that ML touches. I have a love/hate relationship with ML because of this. It seems that there w…

>ML revolutionized translation a long time ago and the demand and pay for translators went down.

This is incorrect.

Translators employed:

2005 - 29240

2021 - 52170

Yearly salary:

2005 - 38300

2021 - 52170

Source - BLS (https://www.bls.gov/oes/current/oes_nat.htm)

Employment growth was ~79% on a US population growth of ~13.5%.

38300 2005 $ equal 52854 2021$.

So pretty much demand for translators exploded while real wages stayed flat.

Re: Copyright infringement in AI art

#56

Post by 6gvONxR4sf7o »

I think it’s clear that what’s happening is (mostly) legal, but the better question is whether it should continue to be. It will be a bad outcome if a few people profit off of a huge group of people’s labor without their consent and without compensating them, in order to put that large group out of work.

Technological advances change the labor market all the time, but it’s absolutely different to exploit your own labor in this kind of way to put you out of a job.

I think if someone wants to build a research ML model, it should be fine to use whatever data you can get your hands on. But if you want to make a commercial model, you should have to have the content creators’ permission to use their data to train your product.

Re: Copyright infringement in AI art

#57

Post by trention »

angusturner wrote: post #54
trention wrote: post #49

Earlier quoted context omitted.

On the basis that the "machine" is not a human and there is nothing that necessitates allowing neural networks to do all the things humans are permitted to do.

I don’t think it makes sense to talk about these systems as though they have any agency (and I think I was careful not to). Of course there is nothing that necessitates letting these systems do what humans do. What I am saying is that I am allowed as a human to ingest copyrighted material, learn from it, form mental abstractions and use those learnings to generate new art. Why then, would I be forbidden from external…

Not having agency is not the issue here.

Usage of the model can (and will) be automated. Also, unlike you, the model scales. This is a qualitative difference in the output produced. You are left only with the superficial similarity of the input process.

Re: Copyright infringement in AI art

#58

Post by angusturner »

pornel wrote: post #50
angusturner wrote: post #46

The question that fascinates me about this whole debate is on what basis (if any) one draws the line between a machine learning from copyrighted data vs a human learning from copyrighted data. And I do think “learning” is the correct metaphor here. There are good reasons to believe that the brain is doing something conceptually similar to generative statistical models. i.e compressing inputs (/percepts) into some kin…

I think it makes sense to have a distinction between human and machine learning. • Human memory has natural limits in capacity, accuracy, speed, etc. Even if you wanted to, you just can't launder copyright all of humanity's works in a blink of an eye. Things done by humans are limited to human scale, but automation can elevate things to an industrial scale (it's legal to fart, it's not legal to have an industrial-sca…

These are fair points. I suppose I am more optimistic about all the potential ways this tech could augment or enhance human creativity.

Like, isn’t it kind of wild to think that the average person could have access to models that have learned from all art ever made…

I’m also encouraged by the fact that some of the best versions of these systems are open source and run on a commercial GPU.

Re: Copyright infringement in AI art

#59

Post by zyx321 »

OmarAssadi wrote: post #40
tkgally wrote: post #36

Earlier quoted context omitted.

Former professional translator here (Japanese-to-English, 1986 to 2005). I have very mixed feelings about MT. On the one hand, as you and others point out, rates for human translation have been hurt, and the quality of a lot of MT-enabled translation visible to the public is mediocre to poor. On the other hand, MT is enabling communication among people that was not possible or practical before. A small start-up is ab…

Good points! As an example, as a hobby, I am interested in things like anime video compression and such. Two of the most essential sites for me in terms of resources and information have been almost entirely Chinese. And while they appreciate and respect the international users, at the end of the day, they can’t translate each and every single forum post into English or any other language that someone may need. And a…

Speaking of MTL, I have never found a satisfactory replacement for ChiiTrans. An app that shows multiple translations side-by-side, including literal dictionary translation and phonetic translation. It would even show you alternative meanings on mouse-over. Is there any more modern app like that? The western fans of Japanese visual novels were decades ahead of their time.

And I agree on the censorship angle too. Imagine if politicians had to stoop to Youtube Video Essay levels of self-censorship, talking about the parallels between badguy Germany and present day badguyism. Or about the psychological effects of naughty trafficking and childhood naughty abuse. I do not want my political discourse limited to PEGI-6 language.

Post reply on HN