We drastically need copyright reform for text, imagery, video. It was never designed for this AI era.
If you take a concept like "fair use". Let's say I embed your photo and express an opinion about it. That's what fair use was designed for. In-context relatively harmless usage of the content of others, for the sake of expression, culture and education.
That's not the same thing as "let me suck up all content ever created without permission, attribution or compensation, mangle it and sell it via the backdoor whilst making you obsolete".
You can't call that fair use, they are wildly different usages at wildly different scales with wildly different impact.
We need a new copyright category specifically for AI usage. If nothing is expressed, no training permission is given. One can opt-in and allow for training, allow for training under conditions, etc.
Honestly, I think it's completely unfair for AIs to train on this data.
I work in ML so I'm aware of the consequences but society wasn't.
My step-daughter is finally crushing it as an graphics artist and she is really pissed at tools like Midjourney.
I asked her about it and she said "yes, they steal the artwork of real artists and generate fake knockoffs" ... and I don't think her opinion is invalid.
In addition, we're all kind of forced to hop on to AI whether we're a programmer or artist just to buy ourselves a little more time, delaying the inevitable. Actually, perhaps accelerating the inevitable by contributing to it.
Even in an utopian world where we would have an economic model to support this (UBI), the outcome still sucks. It wipes out human culture. There's no point in creating/producing anything as almost anything can be produced by anyone, at incredible quality, at no cost and with little skill.
Hence, your daughter being or becoming an incredible artist would have no meaning, except perhaps for herself enjoying the process of creating art.
> Hence, your daughter being or becoming an incredible artist would have no meaning, except perhaps for herself enjoying the process of creating art.
There are lots of points and arguments to be made in this general area, but I have to ask, is this really so bad? I mean, what is the point of our lives and everything we do, other than to generally spend the rest of our time doing things we enjoy for their own sake?
If we're comparing "your daughter is an incredible artist, and here's a job for her designing product packaging for a multinational conglomerate" to "your daughter is an incredible artist, and the multinational conglomerate is using a diffusion model to design their packaging", I think it's really hard to say that the former is better than the latter. Of course, it all depends on the economic model, but the line I am quoting is within that assumption you made of the economic model being able to support this. In that case, I am for the latter wholeheartedly.
Economic incentives are great to get people "hustling", but they are rarely aligned with the human values you wish to protect, and mostly by chance if they are. Your daughter's artistry is better "spent" on art for art's (and personal enjoyment's) sake than on drawing clip art for an obscure HR form somewhere, IMO.
Nothing would stop her from continuing to create art the human way. The most intrinsically motivated will certainly do so.
But it's only half the story. Besides the process of creating art in itself being rewarding, the other rewarding part should be how other people relate to it.
One might have trained themselves for thousands of hours and this will be reflected in the output. Most people suck at art thus the skill, dedication and creativity are recognized as such. This system has merit and scarcity.
The new system has no merit as any fool can type in a few words. Nor does it have scarcity which means an overabundance of output. Both contribute to a lost sense of meaning in creating and even consuming art.
If tomorrow we will all be as fast as the fastest runner, running will become quite pointless. There is no reward or recognition for running fast. In fact, you can't even call it fast anymore, as anybody can do it.
I wonder what would happen to a world where AI runs the economy. Not everyone has some hobby or passion that brings meaning into their lives. Some people just work, come home, and spend their free hours consuming some form of entertainment. Without work, would those people just have more free hours? The elimination of human labor could be disastrous to mental health.
A good point, and I've been puzzled by how hard this split in characters is between individuals.
I know several people that without external force (work, duty) would have absolutely no idea what to do with themselves. Even their free time they organize around work-like chores or spend it on passive media.
These people seem to lack any sense of wonder, of curiosity or exploration. And it seems a permanent and fixed state. This is who they are. You can't change it.
I would not worry about this problem though because surely in the hypothetical situation of no commercial work, there's plenty of other work we can make up.
>There's no point in creating/producing anything as almost anything can be produced by anyone, at incredible quality, at no cost and with little skill.
Does this imply that some significant portion of art "value" is derived from scarcity (e.g. there is more value to creating/producing art when a smaller portion of the population can do so)?
From a strictly financial sense that makes sense, but it does seem morally at-odds with anything that makes art easier for humans to produce.
Is it "good" or "bad" to enable a larger population to produce more art?
Is it "good" or "bad" to enable a larger population to produce higher quality art?
Culturally, both seem like they'd be good. In our current economic model, they're probably both bad.
With an economic model that supports artists financially and removes the need to transmute "art" into "money", I don't think we'd see human culture wiped out. Without a financial incentive to create art, what's the point in creating/producing anything if not to contribute to human culture?
This perspective seems insane to me. I'm undecided, but if I were to put forward an argument that AI art will be bad for culture regardless of economic model, it would be something like "AI art will always be worse (in some way) than human art, but it will also be cheaper than human art, and thus will replace it in basically all commercial fields, which would be bad for culture." Maybe I'd say it's worse because it's inherently soulless, or just that as a practical matter AI is be better at doing the bare minimum than humans are, or something like that.
If I thought that AI art would allow almost anything to be produced by anything at incredible quality, at no cost and with little skill, that sounds like a Sci-Fi utopia to me, an almost unimaginable world in which all limitations on self-expression are lifted. A world in which making a movie or a TV show or a video game becomes a weekend project. It sounds wonderful.
I think if we had an economic model to support this, we would definitely be in a way better system then we are right now. So many artists and musicians don't have the resources right in our current system, and have to seek day jobs or stop making art already.
> Hence, your daughter being or becoming an incredible artist would have no meaning, except perhaps for herself enjoying the process of creating art.
This is the most meaningful reason for creating art. In fact, I'd argue human expression is the defining element of art (AI output not being art in that sense of the word), and economic motivations just pervert it.
Artists don't generate art in a vacuum. Everything is a fake knockoff of everything else.
I believe the cream will still rise to the top, and the best artists will still create something totally different, and/or use AI tools to generate something better than they could create otherwise.
No, it’s not in isolation. Doesn’t matter, because fair use applies to humans, not robots. When you go from “human that does X” to “human that operates machine that does X” you’ve changed the situation.
We’ve already been through this with cameras, which are technically just the same as using your eyes and your memory. Yet both legally and morally we all feel that operating a camera doesn’t grant you the same rights as you have by just being and looking. Strolling through the park and seeing the kids playing is very different from bringing a zoom lens and a camping chair.
That said, society could agree to a fair use that applies to ML-trained models. It could simply cover all non-commercial applications, or at the very least research.
> My step-daughter is finally crushing it as an graphics artist and she is really pissed at tools like Midjourney.
> I asked her about it and she said "yes, they steal the artwork of real artists and generate fake knockoffs" ... and I don't think her opinion is invalid.
Creativity doesn't exist in a vacuum. New creations are based on long-term absorptions of existing concepts & discoveries, & the decision to advance or rebel against any combination of said concepts & discoveries.
The nature of the work will change to focus more on the final product, wherein humans still hold an advantage over art generation models in terms of errors in the produced artwork. There's the possibility that such errors will be corrected with the use of an additional model down the pipeline that's solely focused on correcting said errors, but they're not foolproof either.
There will also be a larger emphasis in some niches over the documentation of the creation of said artworks, as it currently exists in some niche circles I'm in. Reductively, it's the knockoff Gucci handbag problem, wherein the remedies towards it will be the same here:
- (Tech) Serial imprinting / rollover keys / embedded signatures for verification
- (Social) Shaming & ostracization of individuals that buy knockoffs
I'm hesitant on using the legal system to solve such a problem, as the way the current copyright system is set up, it makes it near impossible for a new artist to NOT step on an existing artist's style in some form or another, even if unconsciously doing so.
In my opinion copyright is a law that is always at odds with the free flow of information. I'd hate for that law to start influencing how I interact with user generated text on the internet. As we see on YouTube, nuance for copyright loses to erring on the side of enforcing copyright even when the use is fair.
There's no better thinker IMO on this topic than Stephan Kinsella. (C)opyright law started in the 1500's as a form of censorship. There is no reason for it, other than censorship (or if you are in the top 1%, a great way to extract monopoly profits from the rest).
If you take a concept like "fair use". Let's say I embed your photo and express an opinion about it. That's what fair use was designed for. In-context relatively harmless usage of the content of others, for the sake of expression, culture and education.
That's not the same thing as "let me suck up all content ever created without permission, attribution or compensation, mangle it and sell it via the backdoor whilst making you obsolete".
You can't call that fair use, they are wildly different usages at wildly different scales with wildly different impact.
We need a new copyright category specifically for AI usage. If nothing is expressed, no training permission is given. One can opt-in and allow for training, allow for training under conditions, etc.