How one author pushed the limits of AI copyright | US Copyright Office grants copyright for work made with AI, with caveat

Dangerous territory they're treading into. Some of those individuals (asshats would be my preferred term) quoted later on in the article are the ones eager to drag the legal system into that dangerous ground.

USCO gave her a copyright on the book as a product, not as a work of art. Meaning, no one else can just grab it, copy whole, and put back up for themselves. However, anyone can take the text of the book and do pretty much whatever they want with it, including derivative work of any kind, free and clear. They would have a much easier time getting a favorable summary judgement if the copyright holder took umbrage and initiated a suit against them, as one example of how narrow and weak this "copyright" is compared to a "normal" one.

Sure it's nice that this specific person apparently put in genuine, honest effort and used an AI generator as an assistive tool. It basically served as her much less expensive ghost writer, and one that would never get upset at her when she (again) insisted on rewriting or revising what she was receiving.

But she could have done all that with voice software; dictating to the computer. Except, you know, she'd actually have to write. She couldn't just say "I need some paragraphs about John and Marsha talking over X" and then pick through what the AI (ghost writer) comes up with to fiddle with. What she did is cute, and is a use for the technology, but what she did wasn't exactly writing.

The problem with where this leads is lawyers, and greed. Lawyers are a problem because they're professional problem makers. That's what they do. They're paid to figure out how to get away with shit. Paid to pick through rules and find loopholes, exceptions, oversights, where there's room to push and succeed in pushing. On its face the profession serves a societal purpose, but in practice lawyers often stand ready to assist with problems other lawyers created in the first place.

On greed, I specifically mean the greed people have for finding the lowest possible effort level in anything. Now when you're trying to figure out how to dig a ditch (or similar tasks), that kind of greed is actually useful. And has contributed to the rise of humanity.

But the greed that a decision like this from the USCO is going to enable is the kind of greed demonstrated by Thaler and Abbott (highlighted in the article) who simply want to enable copyright for computer generated material (e.g, AI copyright).

Abbott is a supporter of Shupe’s mission, although he’s not a member of her legal team. He isn’t happy that the copyright registration excludes the AI-generated work itself. “We all see it as a very big problem,” he says.

Abbott is a lawyer with a legal group organized to push for AI copyright. From an article of his:

Patent protection should be available for AI-generated works because it will incentivize innovation. The prospect of holding a patent will not directly motivate an AI, but it will encourage some of the people who develop, own, and use AI. Allowing patents on AI-generated works, therefore, will promote the development of inventive AI, which will ultimately result in more innovation for society.

I categorically disagree with that intention. He says he's talking about patent protection, but he wants IP created by AI to enjoy the same protection human IP does, except it won't be AI that actually owns it.

In a perfect world, it's nice to think that an AI researcher (like Thaler) could "benefit" from AI he or she develops, by securing copyright protection for works generated by the AI that computer scientist develops. But what'll actually happen in such cases is whoever controls the AI will receive the financial benefits of that copyright.

Remember, copyright is not about creativity, but legality. Copyright is the legal system's framework for safeguarding (and encouraging) creators. Occasionally you see copyright cases pop up where an author (here, any artist, not just a writer) sues over creative differences of some sort; but copyright cases almost always hit the courts because of money.

Society cannot afford to let computers generate copyrightable work. As James Cameron wrote so memorably in Terminator, "... and (it) absolutely will not stop. Ever!"

What happens if AI material qualifies for copyright is easy. Everyone and their mother, and certainly every single company (along with a whole slew of newly incorporated ones) with even the vaguest connection to a creative industry sets up computers. Banks and banks of them. All doing nothing but churning out endless Creative Work. Looking for the payday.

Sure 90, 95, even 99.9% of all those computer cycles might do nothing but crank out worthless trash, fluff never to be seen beyond the minimum wage employee who might glance over it while tying it up in a bow to send off to governmental copyright offices for certification, and those government officials who issue the copyright. That won't matter to the companies though, since if they only get one qualified homerun every year or so, that's free money.

Even the energy is becoming free, since solar and other renewable energy sources are coming online and improving at a steadily increasing pace. But we digress.

Everything and everyone will be swamped. Period. That's what corporations do; whatever it takes to profit right now. They don't give a shit about "societal harm" or "long term issues." Gotta get the money in the bank now, before this quarter closes. So The Market can see it and react favorably. So my bosses and corporate board can reward me.

Copyright offices will be buried. There's already a long delay to review and act upon applications. What happens when the volume goes up tenfold (or much, much more) within the span of a year. Which is exactly what'll happen if the USCO is maneuvered, manipulated, or tricked into allowing computer (that does not stop, ever) created material to qualify for copyright.

Because, again, it's not the computer that enjoys any income from all that work; it's the person who wants that income. Even if it's only pennies per work, they just crank out more work and get more pennies. A penny here, a penny there, soon enough you're talking real money.

Scammers and conartists already flood ecosystems like Amazon with all sorts of tricks looking for those pennies, as one real-world example happening right now. Amazon's already started to struggle with, and respond to, those scammers turning to AI. Before, scammers would use random text generators, or outright theft, to pull together some sort of text that could be posted and start pulling in pennies. Now they're using AI to do it, and those are harder to catch with minimal effort. Amazon has responded by starting to enforce AI restrictions that are headed towards a ban, because the people turning to AI aren't disabled honest actors like Elisa Shupe, but asshats panning for pennies.

Sure the US has the economic capability to scale up their copyright office to meet the demand of a world that drops a blizzard of "content" on it every single day. But why should they need to just to accommodate bad actors working in bad faith simply to generate revenue off vast volume?

What about other countries though?

And what about consumers? In books, something I've heard from people about indie publishing is they hate it because "it lets so many people just write. Who's vetting any of this?"

In other words, there are readers who prefer the out-of-sight, out-of-mind gatekeeping traditional publishers do to keep most books from ever seeing a reader's eyes. That's another digression, but I mention it here as an example. Old school radio would be another example; the DJ curated the playlist, and you relied on the DJ to do it. Some music lovers don't like how Spotify makes "everything available" and long for curation to help guide them to find things to listen to.

Now I see some validity in curation, but I feel that's a market problem. Or, rather, a market opportunity. People who can spin up to become recommenders. Except, of course, if that becomes a thing (recommenders) corporations will just do what they did to radio, and take over the recommending to push their only own shit.

But who curates when it's not hundreds of books per day, but tens of thousands? Each and every day. Probably that many songs. Probably at least ten times that number of individual pictures of some sort. And rather than dozens and dozens of TV shows, and another dozens and dozens of movies per quarter, scale that up by at least a factor of ten too.

Consumers are going to be buried just like the USCO will. You'll think "okay, I have a weekend coming up, I need a show to watch." But when you go looking, it'll be hundreds. How do you know which ones are to your tastes? That you'll like?

And remember, you won't even have some of the indicators you have today. Like, for example, I knew I was interested in Shogun because I was familiar with Hiroyuki Sanada and always love to see him in anything aimed at the English market. I was familiar with the original Shogun mini series from the 80s.

AI generated content won't have any of that. No actors you can use as a touchstone, knowing you can rely on reacting favorably to their charm or their style or anything. No lineage from the writer or director, where you can again decide you might like this new thing because you liked older things from that person.

Nope, just an endless flood of material. All enjoying copyright protection, meaning they can just churn and churn and keep churning until they come up with something viral that becomes valuable. Something that goes nova straight to the top of the charts. And because they'll own it (since their computer created it), they'll have all the same protections someone who honestly creates something will.

The legal system will collapse under that weight too. Copyright cases are bespoke because human creativity is bespoke. A human sits there arranging words or paint or whatever, and other humans have to examine it to decide how similar (or not) something is to something else. And people sue all the time, thinking "their" idea was "stolen", not understanding how creative frameworks function and mean that "okay, stories of this genre are going to have shared elements and just because both stories involve a father seeking revenge on a criminal who did his family wrong doesn't make the one a copy of the other." And so on.

What happens when that flood of AI content starts coming out? On a daily basis, some company will look over their back catalog of forgotten crap that never took off, and find examples of things that "seem real similar" to something that goes viral. They'll reach out to that viral owner and demand money or else. And the or else is cranking up the lawyers.

And Abbott wants to enable all of that shit. He's a copyright lawyer. Who gets hired to legally debate and litigate copyright cases? Could it be copyright lawyers?

He's a bottom feeder looking for job security. But because he's a lawyer, he knows how to "work the system."

I pray he fails spectacularly. Because only someone like him benefits in the world he wants to create.