- cross-posted to:
- [email protected]
- cross-posted to:
- [email protected]
We Asked A.I. to Create the Joker. It Generated a Copyrighted Image.::Artists and researchers are exposing copyrighted material hidden within A.I. tools, raising fresh legal questions.
You must log in or register to comment.
Copyright is a scam anyway so who cares?
I can’t believe all the simping for copyright that’s come out of AI. What the fuck happened to the internet? On a place like lemmy no less.
People are smart enough to understand the difference between someone copying for personal use and a billion dollar corporation copying to generate millions while laying off all the creative people. The latter is what these non-open-source AI companies are enabling - for profit too.
deleted by creator
Nah, that’s like saying capitalism is a scam.
Copyright and capitalism in general is fine. It’s when billion dollar corporations use political donations to control regulations
Like, imagine a year after Hangover came out. 20 production companies all released Hangover 2.
Imagine it was a movie by a small Indie studio so a big studio paid off the original actors to be in their knockoff.
Or an animated movie that used the same digital assets.
We need some copyright protection, just not a never ending system
Capitalism is a scam.
It’s an unsustainable system predicted on infinite growth that necessitates unconscionable inequality.
You dropped the “unregulated”.
Socialism is still capitalism. It’s just regulated and we use taxes to fund social programs.
And the second sentence is more caused by not taxing stock trades. If we had a tax that decreases the longer a stock is held, it would prioritize long term investment and companies would care about more than the next months earnings.
All shit that can be solved with common sense regulations.
Socialism is still capitalism. It’s just regulated and we use taxes to fund social programs.
Tell me you don’t know the definition of socialism without telling me you don’t know the definition of socialism.
While Socialism may not be against Free Markets, Capitalism /= Socialism.
Socialism is when the workers collectively own the means of production.
So like in a factory, everyone employed there also has ownership stake in the company, and they can vote on leadership internally. Instead of relying on government regulations to be able to have things like paid lunches and guaranteed sick days, instead they can come to collective agreement on those things, with the vote of every worker/owner. They can still sell products on a free market, but the “capital” part of the equation has been removed.
In capitalism, in a factory, the factory has been purchased by a Capitalist who, by definition, is someone with a lot of Capital (money/wealth) and they bought the factory whole with the capital. Now, they are going to hire workers with the capital as well, and the workers have to follow any and all their rules, like a little fiefdom of a dictatorship and nobody gets an opportunity to vote on leadership. They have no control over pay, working conditions, or much else, and they rely on the Government to enforce it otherwise.
Socialism really doesn’t have anything to do with government regulation, taxation, or social safety nets.
socialism has nothing to do with taxes
Why do all the accounts with “commie” in the name have no idea about any economic systems?
I don’t think any of them actually support communism either, it’s just weird I block so many and they always keep showing up
there is a cure for political illiteracy
I’m not aware of any common sense regulations that take precedence over corporate profits, but keep up the good fight I guess.
…
If we had a tax that decreases the longer a stock is held, it would prioritize long term investment and companies would care about more than the next months earnings.
Did you miss that?
I know starting a new paragraph for each sentence helps more people understand because they tend to skip paragraphs, but my comment was only 6 sentences
I figured two sentences a paragraph wasn’t too much.
I didn’t miss it.
You misunderstood my comment. These regulations you’re talking about either don’t exist or they are ineffective, because we are suffering the greatest economic inequality in history while corporate profits are at an all time high.
But good luck, I hope you become super rich doing business so you don’t have to suffer with the rest of humanity. Have fun buying clean water tokens in a couple decades!
If they were already in effect, we wouldn’t need them because they already existed…
When we have them corporations pay lots of money to politicians to convince them we don’t need them because they’re working.
So when someone says we need regulations, it’s a pretty safe bet the regulations theyre talking about don’t exist yet, and they’re saying they should…
You can tell that’s what they mean, because it’s literally what they’re saying.
At least most of the time most people can.
I’m not sure why you keep not getting this and acting like you’re over replying. Then replying again.
Blocking is easier, it’s what I do when people don’t understand basic stuff while having an attitude about it.
Let me show you an example.
I know it is popular to say that infinite growth is a requirement of capitalistic systems. First of all a proof is never provided. Secondly, Japan might be a counter example, where the GDP has stagnanted for two decades.
Those are fundamental features of the current system. If you want to suggest a copyright system that does protect smaller creators from bad actors but doesn’t allow the mega-corps to bully and control everyone, then feel free. But until such a system is implemented it see no reason to defend the current one which is actively harmful to the vast majority of creators.
That’s the system that was implemented…
It worked fine until corporations realized both parties like money. Which didn’t take long.
And I just said the current situation isn’t good…
deleted by creator
deleted by creator
Copyright issues aside, can we talk about how this implies accurate recall of an image from a never before achievable data compression ratio? If these models can actually recall the images they have been fed this could be a quantum leap in compression technology.
Results vary wildly. Some images are near pixel perfect. Others, it clearly knows what image it is intended to be replicating. Like it gets all the conceptual pieces in the right places but fails to render an exact copy.
Not a very good compression ratio if the image you get back isn’t the one you wanted, but merely an image that is conceptually similar.
Compression is actually a mathematical field that’s fairly well explored, and this isn’t compression. There are theoretical limits on how much you can compress data, so the data is always somewhere, either in the dictionary or the input. Trained models like these are gigantic, so even if it was perfect recall the ratio still wouldn’t be good. Lossy “compression” is another issue entirely, more of an engineering problem of determining how much data you can throw out while making acceptable compromises.
Chat GPT it’s over 500 gigs of training data plus over 300 gigs of RAM, and Sam Altman has been quite adamant about how another order of magnitude worth of storage capacity is needed in order to advance the tech.
I’m not convinced that these are compressed much at all. I would bet this image in its entirety is actually stored in there someplace albeit in an exploded format.
I purchased a 128 GB flash drive for around 12-15$ (I forgot the exact price) last year, and on Amazon, there are 10 TB hard drives for $100. So, the actual storage doesn’t seem to be an issue.
RAM is expensive 128 GB of RAM on Amazon is $500.
But then again, I am talking about the consumer grade stuff. It might be different for the people who are making AI’s as they might be using the industrial/whatever it’s called grade stuff.
It depends on what kind of RAM you’re getting.
You could get Dell R720 with two processors and 128 gigs of RAM for $500 right now on eBay, but it’s going to be several generations old.
I’m not saying that the model is taking up astronomical amounts of space, but it doesn’t have to store movies or even high resolution images. It is also not being expected to know every reference, just the most popular ones.
I have 120tb storage server in the basement. So the footprint of this learning model is not particularly massive by comparison, but It does contain this specific whole joker image. It’s not something that could have been generated without the original to draw from.
In order to build a bigger model they would need not necessarily just more storage but actually a new way of having more and faster RAM connected to lower latency storage. LLMs are the kinds of software that become hard to subdivide to be distributed across purpose-built arrays of hardware.
If you ignore the fact that the generated images are not accurate, maybe.
They are very similar so they are infringing but nobody would use this method for compression over an image codec
I made a novel type of language model, and from my calculations after about 30gb it would cross over an event horizon of compression, where it would hold infinitely more pieces of text without getting bigger. With lower vocabulary it would do this at a lower size. For images it’s still pretty lossy but it’s pretty cool. Honestly I can’t mental image much better without drawing it out.
Hmm this sounds like a similar technology to the time cube
Holy shit I didn’t even think about that.
Essentially the model is compressing the image into a prompt.
Instead of the bitmap being 8MB being condensed down into whatever the jpeg equivalent is, it’s still more than a text file with that exact prompt that gave.
But it’s not deterministic.
It’s just a little bit lossy
I mean, that randomness is just faked. Keep a consistent seed and you’ll get consistent results.
I like that thought too, surely better than calling it AI.
deleted by creator
I mean, only if you have the entire model downloaded and your computer does a ton of work to figure it out. And then if any new images are created the model will have to be retrained. Maybe if there were a bunch of presets of colors to choose from that everyone had downloaded and then you only send data describing changes to the image
I was thinking about this back when they first started talking about news articles coming back word for word.
There’s no way for us to tell how much of the original data even in a lossy fashion can be directly recovered. If this was as common as these articles would leave you to believe you just be able to pull anything you wanted out on demand.
But here we have every news agency vying to make headlines about copyright infringement and we’re seeing an article here and there with a close or relatively close result
There are millions and millions of people using this technology and most of us aren’t running across blatant full screen reproductions of stuff.
You can tell from some of the artifacts that they’ve trained from some watermark images because the watermarks kind of show up but for the most part you wouldn’t know who made the watermarking if all the watermarking companies didn’t use rather unique patterns.
The image that we’re seeing on this news site of the joker is quite exceptional, even from a lossy standpoint, but honestly it’s just feeding the confirmation bias.
“how much of the data is the original data”?
Even if you could reverse the process perfectly, what you would prove is that something fed into the AI was identical to a copyrighted image. But the image’s license isn’t part of that data. The question is: did the license cover use as training data?
In the case of watermarked images, the answer is clearly no, so then the AI companies have to argue that only tiny parts of any given image come from any given source image, so it still doesn’t violate the license. That’s pretty questionable when waternarks are visible.
In these examples, it’s clear that all parts of the image come directly or indirectly (perhaps some source images were memes based on the original) from the original, so there goes the second line of defence.
The fact that the quality is poor is neither here nor there. You can’t run an image through a filter that adds noise and then say it’s no longer copyrighted.
The trained model is a work derived from masses of copywrite material. Distribution of that model is infringement, same as distributing copies of movies. Public access to that model is infringement, just as a public screening of a movie is.
People keep thinking it’s “the picture the AI drew” that’s the issue. They’re wrong. It’s the “AI” itself.
Optimal tip-to-tip efficiency has been achieved.
It’s not as accurate as you’d like it to be. Some issues are:
- It’s quite lossy.
- It’ll do better on images containing common objects vs rare or even novel objects.
- You won’t know how much the result deviates from the original if all you’re given is the prompt/conditioning vector and what model to use it on.
- You cannot easily “compress” new images, instead you would have to either finetune the model (at which point you’d also mess with everyone else’s decompression) or do an adversarial attack onto the model with another model to find the prompt/conditioning vector most likely to create something as close as possible to the original image you have.
- It’s rather slow.
Also it’s not all that novel. People have been doing this with (variational) autoencoders (another class of generative model). This also doesn’t have the flaw that you have no easy way to compress new images since an autoencoder is a trained encoder/decoder pair. It’s also quite a bit faster than diffusion models when it comes to decoding, but often with a greater decrease in quality.
Most widespread diffusion models even use an autoencoder adjacent architecture to “compress” the input. The actual diffusion model then works in that “compressed data space” called latent space. The generated images are then decompressed before shown to users. Last time I checked, iirc, that compression rate was at around 1/4 to 1/8, but it’s been a while, so don’t quote me on this number.
edit: fixed some ambiguous wordings.
You can hardly consider it compression when you need a compute expensive model with hundreds of gigabytes (if not bigger) to accurately rehydrate it
You can hardly consider it compression when you need a compute expensive model with hundreds of gigabytes (if not bigger) to accurately rehydrate it
You can run Stable Diffusion with custom models, variational auto encoders, LoRAs, etc, on an iPhone from 2018. I don’t know what the NYTimes used, but AI image generation is surprisingly cheap once the hard work of creating the models is done. Most SD1.5 model checkpoints are around 2GB in size.
Edit: But yes, the idea of using this as image compression is absurd.
The AI one is better :trollface:
it’s a shame that @NYTOnIt isn’t being updated anymore
It’s not infringing, that’s like saying advertising is infringed by being copied.
If you show your images in public and thet get picked up by crawling spiders, you don’t have a case to curtail its spread.
For fun I asked an AI to create a Joker “in the style of Batman movies and comics”.
The Heath Ledger Joker is so prominent that a variation on that movie’s version is what I got back. It’s so close that without comparing a side-by-side to a real image it’s hard to know what the differences are.
deleted by creator
I can take any image you give me and make a stable diffusion model that makes only that image.
You are confusing bad conduct with bad technology.
Just like mowing down children is not the correct way to use a bus.
Sensationalism and the subsequent tech bro takes is actually unbearable if you just know how the technology works.
Stop pretending to know gen art if you just used one once and know IT! Please stop spreading misinformation just because you feel like you can guesstimate how it works!
This is a classic problem for machine learning systems, sometimes called over fitting or memorization. By analogy, it’s the difference between knowing how to do multiplication vs just memorizing the times tables. With enough training data and large enough storage AI can feign higher “intelligence”, and that is demonstrably what’s going on here. It’s a spectrum as well. In theory, nearly identical recall is undesirable, and there are known ways of shifting away from that end of the spectrum. Literal AI 101 content.
Edit: I don’t mean to say that machine learning as a technique has problems, I mean that implementations of machine learning can run into these problems. And no, I wouldn’t describe these as being intelligent any more than a chess algorithm is intelligent. They just have a much more broad problem space and the natural language processing leads us to anthropomorphize it.
Wah fuckin Wah.
deleted by creator
Get rid of copyright law. It only benefits the biggest content owners and deprives the rest of us of our own culture.
It says so much that the person who created an image can be bared from making it.
I took a gun, pointed it at another person, pulled the trigger and it killed that person.
