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

141–150 of 155 posts

Re: Copyright infringement in AI art

#141

Post by vidarh »

verisimi wrote: post #118
bsedlm wrote: post #111

Earlier quoted context omitted.

I'm wating for the algorithms to automate all my decision making too. so me in the future will be an algorithm suggesting what should I buy/eat/etc.. and another which will accept/reject what the former suggests. this way, I don't have to bother with any of this annoying "life" stuff. /joking

Yes, if you were ruler of the world, you're going to have an awfully huge amount of people doing f.a. There are only so many hairdressers, masseurs and plumbers required. What you need is a kill switch, something to slow down population growth. Anyway, are you up to date with your mandatory medical treatments?

Population growth is slowing all of itself. India has reached replacement, China is close to starting population decline. Few places are left outside sub-Saharan Africa with above replacement birth rates, and they are dropping there too.

Re: Copyright infringement in AI art

#142

Post by v7n »

thriftwy wrote: post #13

What I'm dreaming of is an AI which can make any song covered by any performer or band in history of music. Totally doable but would likely put you on the most wanted list above all the drug lords.

In case you, or others here, have not encountered it yet: https://openai.com/blog/jukebox/

Re: Copyright infringement in AI art

#143

Post by th1s1sit »

egypturnash wrote: post #64
OmarAssadi wrote: post #40

Earlier quoted context omitted.

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…

Similarly, I don’t see why I shouldn’t be able to transform or straight-up copy and distribute someone else’s art if it provides people with happiness. Would you stop doing it if the artists said “this is making me unhappy”?

No post body was provided.

Re: Copyright infringement in AI art

#144

Post by eru »

vintermann wrote: post #140
owow123 wrote: post #114

Earlier quoted context omitted.

> political press photographers aren't hired for originality! Yes they almost definitely are, what else would they be hired for? You wouldn't watch the same news everyday (even if it feels that way) > the number of training epochs has dropped dramatically So? Lets imagine a documented nurodiverse person (lets use a photographic memory) "snapsohts" a famous "works" and then sits in a room, after 5 years training, to p…

> Yes they almost definitely are, what else would they be hired for? Craftmansship. They have subtle, but probably very good, intuitions about lighting, facial expressions etc. and also a lot of concrete knowledge about these things. All of which an algorithm can pick up on by examining thousands of images once each. Once or many matters because if you just examine each data point one, there can be no overfitting in…

> Once or many matters because if you just examine each data point one, there can be no overfitting in the traditional sense.

What makes you so sure? If my algorithm was literally just storing stuff in a hashtable to look up later, you'd get overfitting from a single exposure.

Re: Copyright infringement in AI art

#145

Post by vintermann »

eru wrote: post #144
vintermann wrote: post #140

Earlier quoted context omitted.

> Yes they almost definitely are, what else would they be hired for? Craftmansship. They have subtle, but probably very good, intuitions about lighting, facial expressions etc. and also a lot of concrete knowledge about these things. All of which an algorithm can pick up on by examining thousands of images once each. Once or many matters because if you just examine each data point one, there can be no overfitting in…

> Once or many matters because if you just examine each data point one, there can be no overfitting in the traditional sense. What makes you so sure? If my algorithm was literally just storing stuff in a hashtable to look up later, you'd get overfitting from a single exposure.

Well gradient descent doesn't do that. And the models, while big in terms of parameter data, are not nearly big enough to actually store all the training data.

Think of it in terms of updating beliefs about the target distribution. With backpropagation, you predict based on the input, and update your beliefs according to how wrong you were. So in a sense it's unsound to re-use data - your beliefs already incorporate them! And traditional overfitting is all that - it's when you use up all the information in your training data. This was many people's objection to neural nets (and I thought it was a good objection at the time, and thought myself that the future lay with more "sound" methods, which performed better on most metrics anyway at the time, rather than with dodgy biomimicry which wasn't really even similar to biological brains at all).

But yes, there are other types of overfitting if you want to get philosophical about it. It's just that the one I and everyone used to worry about, from training too much on your data, just isn't important anymore. And most of those clever principled and less-principled regularization methods just don't matter anymore!

Re: Copyright infringement in AI art

#146

Post by xdennis »

jeroenhd wrote: post #121
Imnimo wrote: post #3

I'm not copyright expert, but I would have imagined the answer is relatively straightforward: if an image would be infringing if drawn by a human, it's also infringing if drawn by a model, and similarly for non-infringing. Does the manner in which the image is generated play into whether it infringes a copyright?

> Does the manner in which the image is generated play into whether it infringes a copyright? Is taking a picture of a painting enough to violate copyright? Is a digital copy of a VHS tape a copyright violation? Is copy/pasting an image a copyright violation? A computer is not creative because it cannot think. It can only copy what others do; it doesn't understand what it's doing, only how to do something. Equating a…

Copyright is a privilege granted to man. It would be ridiculous to treat tools as people.

Re: Copyright infringement in AI art

#147

Post by archontes »

Pulcinella wrote: post #89
archontes wrote: post #81

Earlier quoted context omitted.

It's not entirely unanswered. https://www.smithsonianmag.com/smart-news/us-copyright-offic...

This just says that the AI can’t own the copyright, just like how Photoshop, the software program, can’t be granted a copyright but a person using Photoshop can.

This is not what this says. This says that the US Copyright Office currently holds that human authorship is a requirement for any work to be copyrighted. They hold that the AI is a form of non-human creator, and thus all of its works are not eligible for copyright.

See also: https://en.wikipedia.org/wiki/Monkey_selfie_copyright_disput...

Re: Copyright infringement in AI art

#148

Post by sinity »

tkgally wrote: post #36
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…

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…

> On the other hand, MT is enabling communication among people that was not possible or practical before. A small start-up is able to find customers and suppliers in countries with different languages. People with common interests but diverse languages are able to chat and share ideas on social media. A person receives an e-mail from a long-lost relative in a language the recipient doesn’t know; MT enables them to correspond and eventually meet up. Little of that communication would be taking place if a human translator had to be found and paid for each interaction.

Yep. I translated tens of KWords between English and Polish by now - thanks to Deepl, which speeds this up by at least an order of magnitude (and the end result is probably better too).

I even planned to translate a book only available in Polish ([this one](https://en.wikipedia.org/wiki/Perfect_Imperfection)) - I translated about 3% already, but unfortunately the author didn't consent to sharing the translation anywhere - which is annoying, because original (Polish) text is freely available on libgen. And English version doesn't exist. And the book is 18 years old at this point. So the potentially lost profits are negligible or nil. Meh.

Re: Copyright infringement in AI art

#149

Post by andybak »

Does anyone have any thoughts on how this might relate to the Marvin Gaye estate vs Robin Thicke and Pharrell Williams "Blurred Lines" case? I know music has very different copyright rules to visual images but that case springs to mind as it was a situation where infringement was judged to have occurred on much looser grounds than previously required.

Could a similar expansive judgement happen in the case of AI art and shift the goalposts so to speak?

Re: Copyright infringement in AI art

#150

Post by angusturner »

cgrealy wrote: post #136
angusturner wrote: post #47

Earlier quoted context omitted.

Feel like people would have said the same with images 6 months ago tbh. (I mean sure, you could get a fake photograph of a person, but the text control that clip has unlocked is just wild…)

I’m not sure I get your argument. I’m saying that it’s impossible to distinguish between machine written music and human written music. You could maybe make the argument for distinguishing between a human and a machine playing a piece of music. But if I write a song that just goes A, D, C, G on a loop (plenty of songs like that), how would you determine that a machine song that goes A, D, G, C is not human written?

My bad, I misunderstood you! I agree with you.
Post reply on HN