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

Megaface

exposing.ai

Post reply on HN
Print view View original

21–30 of 114 posts

Re: Megaface

#21

Post by traceroute66 »

KRAKRISMOTT wrote: post #9
samwillis wrote: post #8

One of the difficulties with these training datasets is the currently understood rules around web scraping. The current legal precedent [0] is that web scraping is perfectly legal, despite what is in the websites terms of service, "licence" or robots.txt. If a human can navigate to it freely, you can scrape it using automated means. What you can't do with scraped data is republish it verbatim. Doing a data analysis o…

You can do the scraping in a jurisdiction where it is legal.

> You can do the scraping in a jurisdiction where it is legal.

No such thing with GDPR.

Why do you think so many US websites take the lazy-ass approach and block EU visitors to their websites ?

Simple, its because either you comply with GDPR or you don't process the information of citizens of GDPR covered countries. End of story.

Re: Megaface

#22

Post by zarzavat »

ilikehurdles wrote: post #6
colesantiago wrote: post #2

"All photos included a Creative Commons licenses, but most were not licensed for commercial use." I wonder what the implications are for Stable Diffusion, DALLE and Midjourney since that art images on the internet are copyrighted by default. Even with a fair use argument, there are examples in cases where AI was generating art that included the signatures of artists. https://nwn.blogs.com/nwn/2022/12/lensa-ai-art-ima…

> Stability AI is happy to follow copyright laws for their music model, because they know that music labels will hold them accountable. So this seems like a good time to point out to larger companies like @WaltDisneyCo that their copyrighted material is being stolen and used too I mean this is a pretty good point. If they're so sure this is legal, then train on copyrighted audio+video media as they already do with co…

Avoiding doing something because you don't want to get sued and subjected to a lengthy court battle is completely rational and it doesn't mean that doing that thing is illegal.

For example, for decades many TV shows came up with their own lyrics for Happy Birthday song, even though it was well known that the song wasn't copyrighted, because nobody wanted to get sued and fight that battle. Easier to just change a few words in the script.

Re: Megaface

#23

Post by dragonwriter »

542458 wrote: post #10
georgeglue1 wrote: post #4

Are there any licenses that are generally permissive, but prohibit certain programmatic, law enforcement, government, etc. usecases? It'd be interesting legal territory if someone has tried this already.

IANAL. I don’t think you can prevent scraping or use in ML corpuses in this way. Copyright prevents the creation of non-transformative copies of a work other than some protected use cases (parody, education, etc). All OSS licenses do is provide a right to copy a work provided certain conditions (attribution, copy left) are met. But the general legal consensus as far as I know is that most ML models meet the threshold…

> Copyright prevents the creation of non-transformative copies of a work

It also prevents transformative derivatives.

Both nontransformative copies and transformative derivative works may meet (in the US) the exception for fair use, which is the usual argument for nonlicensed use in ML training.

Re: Megaface

#24

Post by satvikpendem »

samwillis wrote: post #8

One of the difficulties with these training datasets is the currently understood rules around web scraping. The current legal precedent [0] is that web scraping is perfectly legal, despite what is in the websites terms of service, "licence" or robots.txt. If a human can navigate to it freely, you can scrape it using automated means. What you can't do with scraped data is republish it verbatim. Doing a data analysis o…

There was an updated ruling in November 2022 showing that HiQ was ruled against and that they reached a settlement with LinkedIn, so I'm not sure that web scraping is entirely legal.

https://www.natlawreview.com/article/hiq-and-linkedin-reach-...

Re: Megaface

#25

Post by Scaevolus »

schemescape wrote: post #19
gwern wrote: post #11

Earlier quoted context omitted.

This is apples and oranges. SD et al are defended on the grounds of being transformative use ( https://en.wikipedia.org/wiki/Transformative_use ): they do not distribute (ie copy ) the original training images, and they are not a derivative work due to transformativeness, so the license of the original images is completely irrelevant. (Details like 'signatures' are also irrelevant: if I write a style parody of Willia…

Does anyone know if attempts have been made to trick these ML models into reproducing original copyrighted inputs verbatim (edit: or close enough)? Edit: Asking about verbatim copies wasn't really a great question. I should have asked about producing things that are "close enough to cause legal trouble" (whether that be due to copyright, trademark, or something else).

That's a known failure mode called "overfitting" or "memorization", where a specific input text is very accurately reproduced.

I'm not aware of it occurring for any copyrighted inputs, but it occurs for many famous artworks-- it's nearly impossible to convince Stable Diffusion to style "Mona Lisa" at all.

Re: Megaface

#26

Post by sdenton4 »

schemescape wrote: post #19
gwern wrote: post #11

Earlier quoted context omitted.

This is apples and oranges. SD et al are defended on the grounds of being transformative use ( https://en.wikipedia.org/wiki/Transformative_use ): they do not distribute (ie copy ) the original training images, and they are not a derivative work due to transformativeness, so the license of the original images is completely irrelevant. (Details like 'signatures' are also irrelevant: if I write a style parody of Willia…

Does anyone know if attempts have been made to trick these ML models into reproducing original copyrighted inputs verbatim (edit: or close enough)? Edit: Asking about verbatim copies wasn't really a great question. I should have asked about producing things that are "close enough to cause legal trouble" (whether that be due to copyright, trademark, or something else).

That's not really how memorization in neutral networks works. For classifiers, memorization is more like learning a hash function and a lookup table; no need to store the full image at all. Even for very large models, the weights are a tiny fraction of the size of the original data.

It's probably helpful to think of embeddings for generative models in a similar way; it's a very specific embedding function, like a locality sensitive hash, which doesn't require actually storing the data.

Re: Megaface

#27

Post by polygamous_bat »

schemescape wrote: post #19
gwern wrote: post #11

Earlier quoted context omitted.

This is apples and oranges. SD et al are defended on the grounds of being transformative use ( https://en.wikipedia.org/wiki/Transformative_use ): they do not distribute (ie copy ) the original training images, and they are not a derivative work due to transformativeness, so the license of the original images is completely irrelevant. (Details like 'signatures' are also irrelevant: if I write a style parody of Willia…

Does anyone know if attempts have been made to trick these ML models into reproducing original copyrighted inputs verbatim (edit: or close enough)? Edit: Asking about verbatim copies wasn't really a great question. I should have asked about producing things that are "close enough to cause legal trouble" (whether that be due to copyright, trademark, or something else).

One, the diffusion model's possible output space contains every RGB image ever. But two, it cannot ever possibly contain the original inputs verbatim, because (the size of the model)/(the size of the training set) comes out to be something like 0.2 KB per image. Unless it's an incredible compression algorithm, diffusion necessarily have learned something from the input rather than copy-pasting things, as claimed upthread.

Re: Megaface

#28

Post by pbhjpbhj »

gwern wrote: post #11
colesantiago wrote: post #2

"All photos included a Creative Commons licenses, but most were not licensed for commercial use." I wonder what the implications are for Stable Diffusion, DALLE and Midjourney since that art images on the internet are copyrighted by default. Even with a fair use argument, there are examples in cases where AI was generating art that included the signatures of artists. https://nwn.blogs.com/nwn/2022/12/lensa-ai-art-ima…

This is apples and oranges. SD et al are defended on the grounds of being transformative use ( https://en.wikipedia.org/wiki/Transformative_use ): they do not distribute (ie copy ) the original training images, and they are not a derivative work due to transformativeness, so the license of the original images is completely irrelevant. (Details like 'signatures' are also irrelevant: if I write a style parody of Willia…

>if I write a style parody of William Shakespeare and add a '--Willy Shakespeare' at the end to round it off, have I revealed that I have secretly copied his work? //

I doubt you're suggesting SD, Dall-E, etc., are producing parodies so bringing in parody considerations muddies the water a lot. Also, Shakespeare's works are out of copyright.

If you sell a painting signed with a [facsimile] signature of Dali then it's pretty hard to say you didn't copy the signature, as a minimum. Thats likely to be a trademark violation too. Now, suppose you include aspects in the image specifically associated with the artist, and a signature, ... there's no way to genuinely deny that is a derivative.

Re: Megaface

#29

Post by fragmede »

pbhjpbhj wrote: post #17
samwillis wrote: post #8

One of the difficulties with these training datasets is the currently understood rules around web scraping. The current legal precedent [0] is that web scraping is perfectly legal, despite what is in the websites terms of service, "licence" or robots.txt. If a human can navigate to it freely, you can scrape it using automated means. What you can't do with scraped data is republish it verbatim. Doing a data analysis o…

>Or is it republishing of the original data? If it's publishing _data_ then you're fine under regular copyright as it only protects artistic works and not things like data. You might fall shy of other IP legislation but not copyright. YMMV, this is not legal advice and represents my personal opinion unrelated to my employment.

The CFAA would be the thing to look out for.

Re: Megaface

#30

Post by polygamous_bat »

colesantiago wrote: post #2

"All photos included a Creative Commons licenses, but most were not licensed for commercial use." I wonder what the implications are for Stable Diffusion, DALLE and Midjourney since that art images on the internet are copyrighted by default. Even with a fair use argument, there are examples in cases where AI was generating art that included the signatures of artists. https://nwn.blogs.com/nwn/2022/12/lensa-ai-art-ima…

> Even with a fair use argument, there are examples in cases where AI was generating art that included the signatures of artists.

I went through the post, and I am not sure whether I agree with the analysis of the examples. Diffusion models are conceptual parrots, and it is possible that "25% images contain a scribble in the bottom right corner, so the model will make a scribble in the corner" is what is being construed as a signature in this post.

I think a large part of outrage from the artists about diffusion model "stealing" art comes from a place of disbelief that machines can be this good without "stealing", and it's perfectly natural. In fact, it's unnatural to me how good machines have gotten in image generation, and it is a field I've been following for five years now. However, because I understand the model and can implement it myself, I can convince myself it doesn't need to steal, just needs to be able to model correlations at some ungodly level.

Post reply on HN