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

Stable Attribution

stableattribution.com

Post reply on HN
Print view View original

141–150 of 365 posts

Re: Stable Attribution

#141

Post by mrtksn »

bioemerl wrote: post #135
mrtksn wrote: post #129

Earlier quoted context omitted.

I don't think we can look into this from copyright standpoint. It was already not working very well in the digital age, now it's completely useless. I don't object that people should learn new skills and find new jobs, my concern is that people who actually put huge effort to produce that "training data" are now left dry. They should be compensated and move on, I'm completely against hampering the abilities of AI in…

The problem is, the net contribution of most artists to these systems is next to zero. It's greater than zero, otherwise the AI wouldn't exist, but these AI are trained off of 5 billion plus examples. Examples that required work to collect and sort through. Examples that required millions of dollars worth of compute time to make into something useful. What is there to compensate them for? These machines are not actua…

I agree, the situation is not ideal. These machines should't been trained on their work in first place.

Currently compensating the artists would be like the Pirate Bay compensating the studios for their production costs through the gambling ads they run. No new move would have been ever made if that was the case.

Re: Stable Attribution

#142

Post by luxuryballs »

mrtksn wrote: post #46
goldemerald wrote: post #8

This is a great website, but not in the way the authors intended. Based on some of the examples they explicitly provided, it is clear to me Stable Diffusion creates novel art. Here's a random example https://www.stableattribution.com/?image=a2666aee-0a1a-411b-... I will admit this is a nice tool for verifying the creations of SD aren't pure copies, so I think it will be useful for a time. But as AI-generated images s…

The discussion over novelty is useless, all these legal structures like copyrights patents royalties and licenses are about creating business structures to allow artist or inventors be compensated. They were already imperfect in it and now suddenly a new technology drags their work into the wilderness with no stuctures for compensation. The images created by the AI are indeed novel but they feed on the work of people…

Sounds like fair use to me, unless I’m mistaken… what could be wrong with replicating a style that someone else introduced? It’s entirely possible someone else came up with the same style hundreds of years before and it just never became well known. People copy various anime styles all the time, why can’t they use a computer program to help them do it?

Clearly the argument can’t just be “because they should hire someone to do it”, heck what if that person is totally booked? It’s not taking away from them in the slightest to produce additional artwork using a similar style if they weren’t available to take the job, so that seems irrelevant.

As long as you’re not taking their work and saying it’s yours or selling it I don’t see why you can’t mimic the style.

Re: Stable Attribution

#144

Post by kmeisthax »

minimaxir wrote: post #120
kmeisthax wrote: post #68

Earlier quoted context omitted.

The AI does NOT build or use reference boards for specific prompts. The only reference it has is the prompt itself, which gets distilled down into a list of 512 numbers, each one of which the AI associates with a particular image feature (or set of features). The only reference material it has is the training set. Most images are not actually retained in the model; but certain statistically significant or repeated im…

> The AI does NOT build or use reference boards for specific prompts. The only reference it has is the prompt itself, which gets distilled down into a list of 512 numbers, each one of which the AI associates with a particular image feature (or set of features). Minor technical clarification: For SD 1.X, CLIPText encodes a prompt and passes a (77, 768) [edit: up to 77] matrix to the core UNet. For SD 2.X, OpenCLIP pas…

Huh. I thought those got projected into the (512) shared CLIP space before getting passed to the conditional blocks.

Re: Stable Attribution

#145

Post by waffletower »

wokwokwok wrote: post #98
social_quotient wrote: post #86

Earlier quoted context omitted.

So the parent post to yours said he’s taking piano lessons. Let’s suppose he becomes great. And let’s suppose he writes his own music. How is Taylor Swift supposed to get paid by him from his early lessons which have undoubtedly played a key role in not only his love for music but the song he writes today?

You are fundamentally not addressing the issue. Previously, someone was being paid to produce something. Now, they are no longer being paid and the business can get the same thing “for free”. You can argue the same issue about out sourcing programmers to south east Asia. Is it good or bad? “…but how do we compensate the programmers who lost their jobs?” It’s just economics right? …but saying that doesn’t address the…

Use of AI models is far from "free". Even given their prior existence, ignoring their research and training cost, AI models like Stable Diffusion require energy and computing resources to execute, and require iteration and selection labor for raw image content to be created. Once raw content has been created, the content still requires processing (i.e. color correction) and integration (i.e. editing, scaling etc.) labor. If analog reproductions are desired, they also require capital to produce.

Re: Stable Attribution

#146

Post by wellthisisgreat »

the copyright people are insufferable. In my experience those who complain the most about copyright and "AI art theft" and whatnot, either

1) have a vested business interest in not automating this kind of stuff (e.g. they churn out low-effort quickturnaround copy / design), so the automation is coming for yet another mind-numbing half-automated (sorry non-stop CMD+C / CMD+V isn't really "creating" anything novel or of value). Yeah those SEO spam gigs will be written by robots now, how quaint.

2) are copyright justice warriors who are raging for the sake of it, meaning they aren't even designers, artists, writers or anything like that.

3) do create some kind of art / text etc., but of a quality that is not putting them at risk of getting the badge of honor that is "in style of ".

At the core of all this rage is just envy at someone being smarter and more successful than you are - in computer science, in statistics, in business, in art, in writing. Yeah someone did something so good it made it into Stable Diffision as a "named prompt". Yeah someone was actually capable of creating Stable Diffusion.

Protectionists are disgusting.

Re: Stable Attribution

#147

Post by mrtksn »

luxuryballs wrote: post #142
mrtksn wrote: post #46

Earlier quoted context omitted.

The discussion over novelty is useless, all these legal structures like copyrights patents royalties and licenses are about creating business structures to allow artist or inventors be compensated. They were already imperfect in it and now suddenly a new technology drags their work into the wilderness with no stuctures for compensation. The images created by the AI are indeed novel but they feed on the work of people…

Sounds like fair use to me, unless I’m mistaken… what could be wrong with replicating a style that someone else introduced? It’s entirely possible someone else came up with the same style hundreds of years before and it just never became well known. People copy various anime styles all the time, why can’t they use a computer program to help them do it? Clearly the argument can’t just be “because they should hire some…

IMHO the current laws and concepts don't work anymore, at all. Copyright laws were already very problematic and caused all kinds of troubles but what's happening now is well beyond all that.

I also think that the use of AI is a fair game, the problem is that the thing itself is made possible by the hard work of thousands of people. By hard work, I mean years or even decades of development. They didn't learn to draw perfect photorealistic paintings, they developed styles and methods that capture something about humans and the machine took it from them.

Re: Stable Attribution

#149

Post by waffletower »

This site is broken from my vantage. I uploaded a Stable Diffusion render of a banal trash can sitting on a lawn. It returned a picture of fiery ruins floating in the sky. While the attributions it returned for this image I did not upload looked somewhat similar in style, they were definitely different enough to indicate that Stable Diffusion created something new and different -- assuming that Stable Attribution is definitive in the referencing relevant training sources. I don't believe that it is.

Re: Stable Attribution

#150

Post by tgsovlerkhgsel »

mrtksn wrote: post #75
josephg wrote: post #65

Earlier quoted context omitted.

The core of your concern (& argument) seems to be the problem of existing business models becoming disrupted by this technology. But the point of good law isn’t to protect an established business model. If that were the case, we would have outlawed the loom because it displaced weavers, and the camera (sorry portrait painters) and the iPhone. (How many telegraph operators are left? None!) When an artist learns to dra…

Sure, it's disruptive and that's fine. The problem is that there's no mechanism to compensate the people who enable your new disruptive machine. As I said, humans learn but the phase humans do it still allows for people be compensated. I'm not anti-AI at all, I think its great and I think the artist who can leverage it can hugely benefit from it but it is not fair to just take something from people away.

The "problem" is only a problem if you assume that everything creates a right to be compensated.

The busker doesn't get compensated if I listen to him as I walk by (unless I choose to). He doesn't get compensated even if he has a really cool, kinda-unique idea, I get inspired by it, and start doing something very similar (without outright copying him in the sense of copyright). He doesn't get compensated even if I get my friends to join me and put him out of business.

Copyright is already an artificial piece of compensation that was added, and that is debatable (copyright can be seen as "theft from the commons"). It intentionally covers certain things and doesn't cover others, and this seems to fall under "others", as long as the training data was obtained legally.

I expect that to be the real problem, especially where content was reposted without permission.

Post reply on HN