Copyright and why it's broken. - Tom Scott

100%.

It gets tricky, though. For example, I'm using a website called "Sudowrite" to help me write a novel. I've been kicking this idea around since 2007. I have a general idea for what it should look like, but I always struggle with Act 2.

Literally over a decade's worth of notes. And not a good Act 2.

But I was able to use ChatGPT and Sudowrite (especially its "Story Engine" tool) to finally understand what Act 2 was missing. And now I'm able to rewrite what I've already done, making it better. AI is a tool just like a word processor is a tool.

Lest anyone think I'm writing an ad here - I'm not. Per their FAQ, Sudowrite says that I own the copyright on anything that I generate with their stuff.

Who owns the copyright to what I write?

You do. Anything you write in Sudowrite and anything Sudowrite suggests for you belongs to you.

But if I don't modify it, that's clearly not true (as you mention). Furthermore, I can actually have it suggest things that might run counter to that idea.

I've had it suggest lines from Kafka - good lines, too. I've read Kafka, so I recognized them... but what if I didn't? I don't own the copyright on those lines, as Tom Scott points out in OP's video. Kafka's original German is public domain... most translations are not.

You can highlight some text in the tool and say "Write this in the style of Douglas Adams." It knows who Douglas Adams is. It knows what his work sounds like. And the only way it knows is because its model was trained on his work. When I did this, one of the suggestions included Zaphod Beeblebrox, which was certainly not mentioned in my text. It also suggests spaceships and aliens and futuristic gadgets, all written in the kind of prose that you'd expect from Hitchhiker's Guide to the Galaxy.

How would it know that, if it hadn't read Hitchhiker's Guide to the Galaxy?

It's why Sarah Silverman is suing OpenAI. While the model may be a bunch of statistics, it also must know what her text is like - to some degree. We can argue over how, but going back to the AI suggesting Zaphod Beeblebrox... if I didn't know HGTTG maybe I'd think that's a cool name for a character? How can Sudowrite say I own the copyright when it's clear that they don't own it, either?

Which sort of brings me back to the beginning. AI has the potential to be a wonderful tool - again, like going from a typewriter to a computer. I have had this idea for literally 16 years now, and Sudowrite was literally a game changer. I knew all of act 1, act 2 was... ehhhh, and then act 3 was never satisfying without a good act 2. I knew where I wanted to go, but not how to get there. AI really helped, because it understands story structures - and how to make good stories (with some prodding - it's not perfect). And now, whenever I'm stumped, I can type some stuff into the prompt and it'll generate ideas for me.

But that only works if we really figure out where the line is for copyright. I'm trusting what Sudowrite is telling me... but I'm taking a risk, because what if they're wrong?

The end bit, that AI created works aren’t copyrightable, is already settled. However, any work a human does to tweak or select AI generated content, if it is itself creative, is copyrightable.

AI created works are copyrightable and guidance from the U.S. Copyright Office isn't law, so it's also not settled. Guidance reflects only the office’s interpretation based on its experience, it isn't binding in the courts or other parties. Guidance from the office is not a substitute for legal advice, and it does not create any rights or obligations for anyone.

I read all that in Tom Scott's voice.

Artists and scientists can make a living of their work without copyright. CC BY SA and creative commons in general doesnt prevent you from selling your work and people can support art/science without it being behind paywalls.

Honestly, show me a large-scale work that was CC BY SA that made money. A book, movie, or game that was funded only by those donating. They are extremely rare. Especially as I know a few people who have released GPL, freeware, and CC BY SA games and have made far less than what they spent on the project. It doesn't prevent you from trying to ask for what are essentially donations but I can tell you with certainty you get far less money asking for donations than you do selling.

We need to move to a culture where we don’t try to enforce everything with the threat of jail/punishments and move to a culture where we support things we love because we want the creators to be able to continue creating without worry

I absolutely agree, just don't take away artists' ability to live first. Move towards that culture first, then get rid of the tools that currently make artists money. Copyright, right now, is almost the sole mechanic in society that allows people to make money off their art on a real scale.

Well... maybe.

Artists are able to work off of commissions, assuming that there is a demand for their art. (Getting that demand is the tricky part.) If people don't want their work on its own, then they have to work at a corporation - maybe making concept art, or drawing animation cels, or whatever. None of that art is owned by them; it's typically in the contract the artist signs when they become employed. Anything they make belongs to the corporation.

I used to work for Disney - in their theme parks, not as an artist - and even my employment contract said that any idea I had while Disney was my employer was property of Disney. Literally, if I had an idea on the job, I could not monetize it. If I thought of an idea for a video game or novel or movie, Disney owned that idea just because they were my employer.

Now. Could they enforce that? No way. But they could try, and as Tom points out then it doesn't matter if I'm in the write or not - Disney has expensive lawyers, I do not.

Scientists need grant money to do science. You have to convince a panel of experts that you have a good idea, and that your idea is worth throwing grant money at. Then you use that grant money to pay yourself and your assistants while you perform an experiment. This grant money can be from a university... or it could be from a corporation doing research and development for new concepts or ideas. If you make a discovery, the corporation might be able to patent that, since you were on their payroll at the time.

Making things Creative Commons doesn't magically make money appear. When you get paid by someone wanting to publish your work, they are specifically buying out your copyright on that work - they can do whatever they wish with it after. (Famously, this is why the first Harry Potter book is called "Sorcerer's Stone" in the US, because the publisher owned the copyright and changed the name.)

Creative Commons, therefore, is completely at odds with traditional publishing, since you can't sell your copyright to them. You can still self-publish, of course... but that's a whole can of worms. Not to mention that it's incredibly easy these days to have AI churn out 80k words of BS and sell it on Amazon for $1.99. You don't need many sales to break even.

Related oldie (but good): Everything is a Remix

No one can prove if a piece of art is AI made and no one has to admit it.

There are tools that are being used to attempt to detect if a piece of work is AI-generated. If those tools say something was, it's then on you to prove that you hand-created it. Even some artists are already having issues because things "look" AI-generated. The onus is on the creator to prove they have the copyright when dealing with copyright infringement.

So realistically, if you make some AI-generated content, I steal it, what do you do? How do you stop me from using your content?

Not sure why artists are brought up here but I guess that’s one of the highly affected groups.

Just to talk about that particular consequence, however, I don’t agree with your take. There are AI trained on works of specific artists, and the end result is that the AI is really good at producing work that’s similar to that artist’s work, effectively creating an alternative to that artist, even if it’s of slightly lesser quality and a lack of depth of the original. While this would likely not affect the artist in the short term, in the long term, new prospects who don’t yet know the artist well enough would likely be unable to tell the difference in quality, and may even go straight to the AI model since that’s distributed cheaply or even free. It may also negatively reflect on the original artist to people who don’t know the artist, as the works from the AI would likely be more abundant, and people not in the know may think that the original artist was in fact just producing their works through AI. It is highly discouraging for artists who have worked hard to hone their craft, only to have people think that their works have little difference or even a mimicry (don’t underestimate misinformation).

There has been many instances where such training was done without the knowledge of the artist. Imagine just waking up one day, and finding that there’s someone or something that can very closely reproduce your works, one’s you’ve taken many years of practice to produce, of which its quality is almost unique to yourself. There’s a blatant lack of respect for the hard work that people put into their craft, one that seemingly belittles their blood and tears, and could even be a mockery of their existence. Some artists don’t have other jobs; their art and craft is their job, and some may have even sacrificed learning the skills needed for other jobs to pursue their passion.

Saying that AI is not intended to replace artists, but to improve accessibility, is like saying ATMs weren’t meant to replace bank tellers. True, there’s much less skill required for bank tellers, and getting cash out of banks is an important process that should be swift with almost no errors, so replacing bank tellers with ATMs is a general good, except for the bank tellers, which then banks can retrain them for other jobs. Since then, the job has virtually gone extinct, and almost nobody would want to become a bank teller, and if anyone would like to, they would need to perform better than ATMs. Artists require great skills and creativity, many of which are not easily trained or obtained. Seeing an automated system produce works that are acceptable by most people would either greatly discourage new artists or perhaps even entirely remove the idea of becoming an artist for most people. It raises the barrier to becoming an artist: not only do you need to stand out, you also need to be good enough such that people can’t just train an AI model on your work to produce results that are highly indistinguishable from yours. How many more years do people need to train to be that good? For those with a job but wish to become an artist, abandoning their job to focus on their craft will likely become a much more difficult choice to make. Also, I don’t doubt this would further rise the prices of commissions due to how much work artists would have to put in, and this would only get worse at a rate that’s much faster than a scenario without AI.

So a line should be drawn somewhere. AI trained on public works or artist-approved works are definitely okay. All other options will likely need further discussion and scrutiny. We’re talking about the possibility of ruining an already perilous career path, whose works are coveted.

Seems like wishful thinking to assume that technology being used to replicate artists' styles is not going to affect their livelihood. Unlike large media companies, even a few lost clients can make a huge difference for smaller artists. Then there's also upcoming artists, if they aren't already well-known and better than AI, how are they going to get a foothold on an artistic career? That might kill their chances before they get anywhere.

It's not misinformation:

In other cases, however, a work containing AI-generated material will also contain sufficient human authorship to support a copyright claim. For example, a human may select or arrange AI-generated material in a sufficiently creative way that “the resulting work as a whole constitutes an original work of authorship.” Or an artist may modify material originally generated by AI technology to such a degree that the modifications meet the standard for copyright protection. In these cases, copyright will only protect the human-authored aspects of the work, which are “independent of” and do “not affect” the copyright status of the AI-generated material itself.

So this means work generated by AI is not copyrightable. The work generated by human touch is. https://www.federalregister.gov/documents/2023/03/16/2023-05321/copyright-registration-guidance-works-containing-material-generated-by-artificial-intelligence which is the original source of your post.

"You can not copyright works produced by AI." is a 100% true and factual statement. You can have the AI work as a part of a bigger thing that you retain the copyright on or you can modify it and retain copyright on only the modifications but the AI work in itself is not copyrightable and your copyright on the bigger thing doesn't supersede the lack of the copyright on the AI works.

So in the case in question where someone used AI work to generate images for a comic book. The comic book itself is under copyright but anyone can take those images and modify them, redistribute them without credit, etc. Just like if you used AI images for textures in a game, the game is under copyright but those textures could be lifted out of the game and used elsewhere.

I work in the games industry and can tell you AI isn't going to get to the point any time soon to fully replace game artists. Likewise, I'm an engineer and chatgpt can generate code from prompts but it's not very good. The 3d models that ai can generate are terrible. If you are a 3d artist you need not worry about ai.