Use your AI generation all you want but don’t enter a painting contest using machine generated content trained on other people’s work without their consent.
The fact that you’re comparing human artists to slop machines is really sad. There is no “silicone brain” making any of this stuff. I think you should take a few minutes and learn how this stuff works before making these comparisons.
Wow thank you for this comment. It helps detail your level of knowledge on this subject, which is very helpful to myself and others. There is nothing else to discuss here on my end.
Alrighty, so generative AI works by giving it training data and it transforms that data and then generates something based on a prompt and how that prompt is related to the training data it has.
That’s not functionally different from how commissioned human artists work. They train on publicly available works, their brain transforms and stores that data and uses it to generate a work based on a prompt. They even often directly use a reference work to generate their own without permission from the original artist.
Like I said, there are tons of valid criticisms against Gen AI, but this criticism just boils down to “AI bad because it’s not a human exploiting other’s work.”
And all of this is ignoring the fact that ethically trained Gen AI models exist.
It does not think, it is not capable of creating novel new works, and it is incapable of the emotion necessary to be expressive.
All it can do is ingest content and replicate it. This is not the same as a human seeing someone’s work and being inspired by it to create something uniquely their own in response.
I never claimed that Gen AI has consciousness, or that what they produce has emotions behind it, so I’m not sure why you’re focusing on that.
I’m specifically talking about the argument that AI is bad because trains on copyrighted material without consent from the artist, which is functionally no different than humans doing the exact same thing.
This isn’t me defending AI, this is me saying this one specific argument against it is stupid. Because even if artificial consciousness was a thing, it would still have to be trained on the same data.
And that means humans don’t learn art the same way a machine trains on data. Even if they learn from other artists, a human’s artistic output is novel and original.
How exactly is a generated image not novel? You’re not going to get the same image twice with the same prompt. Everything it generates will be original. It’s not like they’re just providing you with an existing image.
And still the argument I’m hearing is that it’s fine for humans to use artistic works without consent or credit just because it’s a human doing it.
Just because the underlying processes are different doesn’t mean the two are functionally different.
I also think it’s funny because I’m betting the Venn diagram of people who think AI using publicly available artwork to train on is bad and people who think piracy is good is almost a single circle.
Use your AI generation all you want but don’t enter a painting contest using machine generated content trained on other people’s work without their consent.
Do human artists usually get consent before training on content freely available on the Internet?
There are plenty of reasons to hate on AI, but this reason is just being pissed that a silicon brain did it instead of a carbon one.
The fact that you’re comparing human artists to slop machines is really sad. There is no “silicone brain” making any of this stuff. I think you should take a few minutes and learn how this stuff works before making these comparisons.
Right, because computers don’t use silicone.
But Gen AI is modeled after the way the brain works, so maybe you need to learn how it works before arguing against an accurate comparison.
Wow thank you for this comment. It helps detail your level of knowledge on this subject, which is very helpful to myself and others. There is nothing else to discuss here on my end.
Alrighty, so generative AI works by giving it training data and it transforms that data and then generates something based on a prompt and how that prompt is related to the training data it has.
That’s not functionally different from how commissioned human artists work. They train on publicly available works, their brain transforms and stores that data and uses it to generate a work based on a prompt. They even often directly use a reference work to generate their own without permission from the original artist.
Like I said, there are tons of valid criticisms against Gen AI, but this criticism just boils down to “AI bad because it’s not a human exploiting other’s work.”
And all of this is ignoring the fact that ethically trained Gen AI models exist.
GenAI is a glorified Markov Chain. Nothing more.
It is a stochastic parrot.
It does not think, it is not capable of creating novel new works, and it is incapable of the emotion necessary to be expressive.
All it can do is ingest content and replicate it. This is not the same as a human seeing someone’s work and being inspired by it to create something uniquely their own in response.
I never claimed that Gen AI has consciousness, or that what they produce has emotions behind it, so I’m not sure why you’re focusing on that.
I’m specifically talking about the argument that AI is bad because trains on copyrighted material without consent from the artist, which is functionally no different than humans doing the exact same thing.
This isn’t me defending AI, this is me saying this one specific argument against it is stupid. Because even if artificial consciousness was a thing, it would still have to be trained on the same data.
My entire post was a rebuttal of the “functionally no different than humans doing the same thing”.
Humans take inspiration and use it to express themselves uniquely, genAI just steals and replicates. They are in no way “doing the exact same thing”.
deleted by creator
Humans aren’t machines, dummy
And?
And that means humans don’t learn art the same way a machine trains on data. Even if they learn from other artists, a human’s artistic output is novel and original.
How exactly is a generated image not novel? You’re not going to get the same image twice with the same prompt. Everything it generates will be original. It’s not like they’re just providing you with an existing image.
And still the argument I’m hearing is that it’s fine for humans to use artistic works without consent or credit just because it’s a human doing it.
Just because the underlying processes are different doesn’t mean the two are functionally different.
I also think it’s funny because I’m betting the Venn diagram of people who think AI using publicly available artwork to train on is bad and people who think piracy is good is almost a single circle.