When a person does this, in general, nobody bats an eye - as far as I know, this practice is exceedingly common.
That's right.
Was the technology created with a nefarious villainous plan to push out artists?
Almost certainly why it's gotten so much investment recently. We've heard various people in business leadership get excited about firing their design and illustration teams. I've seen business owners on this forum talk about how they've shifted to using AI images. And then there's the writer's strike in Hollywood, where the use of AI is currently a major point of negotiation that the studios don't want to budge on. It is very naive to not consider the impact of this technology on creative labor, even if it's an inferior product.
a lot of the anger stems from the pushback on some of the angrier anti-AI proponents, ones that commonly discredit others' work and advocate for draconian IP laws.
The pushback is warranted. I don't know what the solution here is, but it seems to me that these AI models are generating a huge amount of value off of unacknowledged labor, at the expense of the people who made that labor. I think people are rightly fed up over yet another example of rent seeking tech platforms getting the money for labor they didn't do, open models like stable diffusion notwithstanding.
Almost certainly why it's gotten so much investment recently. We've heard various people in business leadership get excited about firing their design and illustration teams. I've seen business owners on this forum talk about how they've shifted to using AI images.
My question was whether generative AI was created with ill intent, not whether there are parties that can profit off of it. You're conflating the computer scientists, enthusiasts and hobbyists (that likely represent the vast majority of the audience here) with business owners and other people whose job is to maximize profit. Back when generative AI wasn't nearly as coherent, only the first group cared much about it.
these AI models are generating a huge amount of value off of unacknowledged labor, at the expense of the people who made that labor
The original discussion compared human artists to generative AI, so the question here is that, if human artists are allowed to use non-AI tools to do the same thing, what makes AI tools different? Say, someone puts in a lot of time analyzing the works of the most popular contemporary artists, and then becomes capable of replicating their styles, while charging a fraction of the price. Are they in the wrong here?
You're conflating the computer scientists, enthusiasts and hobbyists (that likely represent the vast majority of the audience here) with business owners and other people whose job is to maximize profit. Back when generative AI wasn't nearly as coherent, only the first group cared much about it.
The computer scientists and enthusiasts are the ones who wrote the programs that scraped the data. They are just as responsible for the situation as the business owners, and this comment goes back to my earlier statement that people working in the valley don't think about the consequences of the things they build nearly enough.
Say, someone puts in a lot of time analyzing the works of the most popular contemporary artists, and then becomes capable of replicating their styles, while charging a fraction of the price. Are they in the wrong here?
There are not. This is the exact same argument that you made before. If you can't understand the distinction between artists learning the trade and a gigantic VC funcded company producing a system that can generate thousands of images that vaguely resemble your work then I'm not sure what else I can say.
When an artist learns from another artist, they cannot exactly replicate that artist's work. Their own experiences and even muscular structure changes the result, and typically, one artist is not really a threat to the market for the orginal artist's work. New artists contribute to the field and maintain the knowledge of skills so that those can be passed on to other artists.
AI systems are practically built to digest entire corpuses of people's work then generate huge amounts of similar work by the truckload. Nobody is learning anything from the production of these images, we've automated the production of cultural artifacts which is totally perverse, and we've made it harder for artists to make a living which impacts the field negatively.
Shallow arguments like "The AI system learns just like humans" are a slap in the face.
Comments
That's right.
Almost certainly why it's gotten so much investment recently. We've heard various people in business leadership get excited about firing their design and illustration teams. I've seen business owners on this forum talk about how they've shifted to using AI images. And then there's the writer's strike in Hollywood, where the use of AI is currently a major point of negotiation that the studios don't want to budge on. It is very naive to not consider the impact of this technology on creative labor, even if it's an inferior product.
The pushback is warranted. I don't know what the solution here is, but it seems to me that these AI models are generating a huge amount of value off of unacknowledged labor, at the expense of the people who made that labor. I think people are rightly fed up over yet another example of rent seeking tech platforms getting the money for labor they didn't do, open models like stable diffusion notwithstanding.
My question was whether generative AI was created with ill intent, not whether there are parties that can profit off of it. You're conflating the computer scientists, enthusiasts and hobbyists (that likely represent the vast majority of the audience here) with business owners and other people whose job is to maximize profit. Back when generative AI wasn't nearly as coherent, only the first group cared much about it.
The original discussion compared human artists to generative AI, so the question here is that, if human artists are allowed to use non-AI tools to do the same thing, what makes AI tools different? Say, someone puts in a lot of time analyzing the works of the most popular contemporary artists, and then becomes capable of replicating their styles, while charging a fraction of the price. Are they in the wrong here?
The computer scientists and enthusiasts are the ones who wrote the programs that scraped the data. They are just as responsible for the situation as the business owners, and this comment goes back to my earlier statement that people working in the valley don't think about the consequences of the things they build nearly enough.
There are not. This is the exact same argument that you made before. If you can't understand the distinction between artists learning the trade and a gigantic VC funcded company producing a system that can generate thousands of images that vaguely resemble your work then I'm not sure what else I can say.
When an artist learns from another artist, they cannot exactly replicate that artist's work. Their own experiences and even muscular structure changes the result, and typically, one artist is not really a threat to the market for the orginal artist's work. New artists contribute to the field and maintain the knowledge of skills so that those can be passed on to other artists.
AI systems are practically built to digest entire corpuses of people's work then generate huge amounts of similar work by the truckload. Nobody is learning anything from the production of these images, we've automated the production of cultural artifacts which is totally perverse, and we've made it harder for artists to make a living which impacts the field negatively.
Shallow arguments like "The AI system learns just like humans" are a slap in the face.