Month: December 2023

Harris Brewis is back with a four-hour examination of plagiarism on YouTube. Yeah, it is a big one; I watched it in two parts because I needed to charge my headphones, because that is the world in which we now live, and it effectively monopolized my lazy Sunday.

Its subject matter is a little inside baseball — I do not watch nearly enough YouTube to know any of the creators examined — but it is a thoughtful look at what plagiarism on YouTube is like and a particularly damning exposé of one person specifically. It is extremely long but it kind of needs to be in order to accommodate examples in a fuller context.

As you may expect, there is a part of the conclusion which touches on generative media tools: how they can be used to disguise plagiarism in text, and how they themselves are farming the broader works of others. I found this interesting as I have been writing about intellectual property and generative tools a little bit. I am not sure if Brewis’ presentation of copyright is correct — that is to be determined — but it does seem like these tools do something more akin to plagiarism than strict copyright violation. Attorneys at Heer Law — no relation, as far as I know — note that plagiarism “is an ethical offence — rather than a legal offence — and it does not necessarily encompass copyright infringement”. That feels right to me as a description of what something like ChatGPT does in producing something nominally original.

Matthew Lane:

I’m not saying that there aren’t issues with AI that need to be addressed, especially worker exploitation. AI art generators can be especially infuriating for artists: they use a lot while giving back little. In fact, these generators are arguably being built to replace artists rather than to provide artists with new tools. It can be attractive to throw anything in the way to slow it down. But copyright, especially copyright maximalism, has done a terrible job of preventing artist exploitation.

As I wrote recently, this is an area of interest for me as both a technology-curious person and as an artist. There are as many ways of using generative products in ways that feel creative and encouraging as in ways that feel kind of like a cheat — for example, to duplicate a specific artist’s signature style. That is unfortunate; it is also not new.

I am not a lawyer — I have a fine arts degree — but I understand it is generally legal to replicate another artist’s work. You can repaint a painting or make a sculpture that is a duplicate of another work. (Update: To be clear: for personal use.) What is illegal is when that work is presented as though it was a creation of the original artist. That is, you can sculpt a famous statue and maybe even sell it as a copy (Update: in some places, sometimes!), but you cannot claim it is a figure by that artist — and you especially cannot do so in conjunction with a sale. It also depends where you live. This is not legal advice.

It is also perfectly legal to make things in the style of a specific artist. There are plenty of people who make music intended to sound like Joni Mitchell, take photos that look like something Edward Burtynsky would do, and make movies that imitate Wes Anderson’s style. Artists lift from each other all the time. (Update: This also depends on how a court interprets it; as I was reminded by email, a jury determined Robin Thicke’s “Blurred Lines” sounded too similar to Marvin Gaye’s “Got to Give It Up”.)

But there is an obvious power imbalance between if some random person recording a note-for-note cover of a Lana Del Rey song without permission, and if Lana Del Rey duplicated an unknown artist’s song. That is the uncomfortable tension of what some of these generative tools enable. A contributor to a stock photo website is likely no Lana Del Rey of their own craft, but they get paid to produce flexible illustrative photos for ads and brochures. If these images can be generated instead by a rich company that has money to burn, photographers do not get paid any more for this kind of work. That goes, too, for people who have been expected to churn out website filler, or even application developers. We are not going to solve this with stricter copyright laws.