A.I. artist claims people are stealing his work

So does this guy have a legitimate complaint that he should receive a copyright on his artwork? The article seems pretty dismissive of the idea and he sounds like a tool, but I don’t know a lot about the settled law.

AI Artist or BS Artist?

We seem to be living in a time when otherwise intelligent people are being constantly overwhelmingly bamboozled by obvious scams like Crypto, NFTs, and now AI. All of it is using the same MO, that it can create money out of nothing as long as you jump on the bandwagon early before it crashes. Which sounds like a Ponzi scheme to me.

To paraphrase Jack Nicholson’s Joker, the world needs an enema.

Case law is very thin as yet, and clearly applicable legislation is even thinner, but the safe bet, for now, is “no.”

Which suggests very quickly that most newspapers, magazines, and advertising won’t be copyrightable either. Since that stuff is increasingly AI-generated at least in part.

The giant two-edged sword of all of this is that AI works by ingesting vast amounts of copyrighted works scraped for free, then creating (“creating?”) new works given all that background knowledge.

The idea they can absorb all that training data without paying those authors and then be able to protect their work and charge for it seems … inconsistent.

The AI authorship crowd had better be careful what they wish for. They just might get it.

It’s not “his” work, though. It is the work of a computer program. My artistic “talent” - and excuse me for bastardizing the word - is nothing short of abysmal, but even I could push a button and have at it. If he didn’t break the rules, the rules are no longer adequate. Submissions by human hands only!

And the Great Robot Replacement begins…

Careful. “They” is doing a lot of work, there.

The “They” who absorbed training data and created a model are the giant AI companies with their billion dollar models trained on everything on the internet.

The “They” who created the art and is now asking for protection is a guy who got the tool and used it at home.

And while he isn’t painting with a brush… Neither is someone who uses Photoshop. Or a photographer, for that matter.

If a photographer, who captures what something looks like in the world, is an artist because of framing, and picking a subject, and so on - then why isn’t a guy who generates a bunch of AI photos, honing the craft of prompting an AI model, and then freezes parts of the image while transforming others (something you can do with Midjourney) until he gets an aesthetically pleasing result?

And a photograph isn’t a photographer’s work, it’s the work of nature or architects or whatever, he just captured the image that was already there.

You really think that you can generate AI art as good as his? Since you seem to think that generating AI art simply requires pushing a button, I highly doubt that.

It quite famously, absolutely, is exactly that.

The technology has been improving in leaps and bounds. I have no doubt I could create something that good today, but when he made that piece two years ago it was very difficult to squeeze that much quality and detail from generative programs.

Of course it isn’t. There are a number of skillsets involved in generating this stuff, especially if you’re trying to be consistent or have very specific outcomes in mind.

It’s no more “pushing a button” than writing this post is “clicking submit.”

Except it isn’t; at least, not any more so than a camera is a magic “push button to make art” device.

That’s true in a way, I guess today it’s less about generating a passable image and more about being able to get the result you envisioned out of the tool.

Anyone can push the button and get a “well drawn” image, just like anyone can point a camera and click to get a photorealistic image.

But not anyone can make art that evokes a desired feeling or conveys a desired idea using that tool.

Eta: I see you later make the same point about a specific outcome rather than a “good” image.

I think the AI Artist should retain copyright not because of the work required to create the art, but because it’s essentially work for hire.

If I hire a human to draw me a picture of Donald Trump riding a Patriotic Pterodactyl, that (ahem) “art” is mine, I own the copyright.

If I ask an AI to draw me the same picture, why is it different?

Under no circumstances should the AI company have copyright, but I don’t see why someone who coordinates the production of a specific work of art can’t copyright it.

From a different article on the same:

He inputted “at least 624” text prompts before Midjourney delivered the image he envisioned, according to the Copyright ruling.

These were used to adjust the scene, tone and focus of the image, he argued to the Copyright Office, including details about “how colors [should be] used” and what style and era of art the image should display. He used a “writing technique” for prompts that he had “established from extensive testing.”
[…]
After Midjourney output the base of the image, Allen then used Adobe Photoshop to “remove flaws and create new visual content” before feeding the image into another platform, Gigapixel AI, to increase its resolution and size, according to the latest copyright decision.

He’s never (AFAIK) revealed the prompt he used so it’s difficult to judge just how complex or “artistic” is was if we’re making an argument for effort. I know, from much experience with Midjourney, that often the most interesting images are pretty far removed from what you actually prompted for. He mentions that he had a vision of Victorian ladies in space helmets which I never got from the image at all. Even now, knowing what it’s supposed to be, it’s still “ehh…”

Props for manipulating what I’m guessing was MJ v3 into kicking out something worked within its limits to make an award-winning piece. Looking back on my v3 renders, it’s not quite so charitable. Usually some good moments but the overall piece always has some weird and questionable bits which makes me believe that he had to spend some time in post-production manipulating it in Adobe.

(I definitely agree that it’s not just “pushing a button” especially for a professional piece that needs additional work to make it match the vision; the rest is just me musing. I’m a bit agnostic on the copyright issue)

He also complained that “the court should recognize “his ‘creative input’ into Midjourney, which included ‘enter[ing] a series of prompts, adjust[ing] the scene, select[ing] portions to focus on, and dictat[ing] the tone of the image,'” which he claimed was “on par with that expressed by other types of artists and capable of Copyright protection.”

Which sounds like baloney. I’m a photographer, but I’m not gonna say my art takes as much skill or difficulty as, say, a glass blower. Not all art is created equal.

Agreed, and that’s why I don’t refer to a photographer as an artist. I refer to him as a photographer. The A.I. creator who produced that program is not an artist, he is a great programmer.

Then you’re in the minority, because I think most people consider photographer’s to be a type of artist.

The AI creator is not a part of this discussion, at all.

He should be because it is his creation that is pulling the wagon artistically speaking.

But if I go outside and use a $5 disposable camera to snap a photo over my shoulder of my house without even looking, it’s still covered under copyright, isn’t it? “Effort involved” doesn’t seem to be a criteria for copyright with slight exceptions that you can’t copyright a short phrase or common symbols.