I'm neither an artist or a software dev, but I do work in tech - here's my perspective:
A piece of art created by an artist will likely be posted publicly, where it can be used as part of a training set for an AI art program. This has come up in news articles, I read on recently where an artist's work was used without her consent, and the AI proceeded to produce almost identical images.
A piece of code written by a software developer will likely only be posted publicly if it is open source, so the probability of an AI code program being trained on closed-source code is slim to none. Obviously there are issues surrounding attribution, and I've seen evidence of copilot spitting out some open source code verbatim, bar the license or any attribution.
I think the more open, sharing nature of coders, especially those in the open source space, is what has led to AI code being more accepted than AI art, where artists are far more concerned about protecting their intellectual property.
First, because ultimately code is about getting something that works. Whereas ultimately art is about expressing yourself and connecting with another. That's why artists get into it. It's not about getting a nice end product alone it's about expressing yourself in the form of an end product. For artist, the art lies is in this expression (and in recognizing this expression in others).
A coder might like to code, but they also understand that if the program works correctly, then it doesn't matter if they or an AI coded it. Sure, they'd be out of a job, but it would still be a valid program.
Whereas an artists sees an AI work not just as something that will take their job, but also something that doesn't fulfil the purpose of art. They see it as a chef would see a dish not just cooked by a robot, but also made up of styrofoam.
Another reason of course is even more pragmatic. Coders was high paying well cushioned jobs, and they feel they can work as coders for the next 20-30 years, even with AI rising.
Whereas artists (especially in illustration and graphic design) are already squeezed, and making a pittance, and they see the little they manage to get diminishing because of AI, not in 20-30 years, but here and now: less prestigious publications, cheapo ads, websites, etc. will already start increasingly use some AI service to generate illustrations.
Well, because it's likely to replace at least a portion of the talent needed for creating art. AI is going to do a lot of the heavy lifting, and "manual" art will be a specialty. Artists will still be needed, but the AI tools will drastically change entire industries... less artists will be needed, and the few talented ones left will be training models to sell for everyone else.
Programming has higher stakes than art, and requires more "correctness", so it's likely to take much longer for AI to properly replace programmers. Even then, programmers would just switch to AI engineering, which by then will be needed everywhere.
Because it's likely to replace at least a portion of the talent needed for creating art. AI is going to do a lot of the heavy lifting, and "manual" art will be a specialty. Artists will still be needed, but the AI tools will drastically change entire industries... less artists will be needed, and the few talented ones left will be training models to sell for everyone else.
Programming has higher stakes than art, and requires more "correctness", so it's likely to take much longer for AI to properly replace programmers. Even then, programmers would just switch to AI engineering, which by then will be needed everywhere.
A lot of very disappointing comments in TA about how artists can't handle disruption, or they think they're special, or take themselves too seriously.
The situation is a bit like what happened with machine translation. Art and translation are both necessary and badly underpaid lines of work. The AI's work may not be perfect, good, acceptable, or sometimes even coherent on the most basic level, but many companies will balk at paying a modest wage to a skilled worker for necessary work, and will either ship the garbage the AI gives them, or more often they'll use its existence as a bludgeon to reduce the wage further to unacceptably low levels. It happened to translation, it'll happen to artists.
6 comments
[ 3.3 ms ] story [ 25.2 ms ] threadA piece of art created by an artist will likely be posted publicly, where it can be used as part of a training set for an AI art program. This has come up in news articles, I read on recently where an artist's work was used without her consent, and the AI proceeded to produce almost identical images.
A piece of code written by a software developer will likely only be posted publicly if it is open source, so the probability of an AI code program being trained on closed-source code is slim to none. Obviously there are issues surrounding attribution, and I've seen evidence of copilot spitting out some open source code verbatim, bar the license or any attribution.
I think the more open, sharing nature of coders, especially those in the open source space, is what has led to AI code being more accepted than AI art, where artists are far more concerned about protecting their intellectual property.
A coder might like to code, but they also understand that if the program works correctly, then it doesn't matter if they or an AI coded it. Sure, they'd be out of a job, but it would still be a valid program.
Whereas an artists sees an AI work not just as something that will take their job, but also something that doesn't fulfil the purpose of art. They see it as a chef would see a dish not just cooked by a robot, but also made up of styrofoam.
Another reason of course is even more pragmatic. Coders was high paying well cushioned jobs, and they feel they can work as coders for the next 20-30 years, even with AI rising.
Whereas artists (especially in illustration and graphic design) are already squeezed, and making a pittance, and they see the little they manage to get diminishing because of AI, not in 20-30 years, but here and now: less prestigious publications, cheapo ads, websites, etc. will already start increasingly use some AI service to generate illustrations.
The argument falls down amongst those who consider art to be decorative. They are not wrong, they just have a different perspective.
AI communicates as much visually as it does in written form - ie not at all. But it can produce decorative things.
Programming has higher stakes than art, and requires more "correctness", so it's likely to take much longer for AI to properly replace programmers. Even then, programmers would just switch to AI engineering, which by then will be needed everywhere.
Programming has higher stakes than art, and requires more "correctness", so it's likely to take much longer for AI to properly replace programmers. Even then, programmers would just switch to AI engineering, which by then will be needed everywhere.
The situation is a bit like what happened with machine translation. Art and translation are both necessary and badly underpaid lines of work. The AI's work may not be perfect, good, acceptable, or sometimes even coherent on the most basic level, but many companies will balk at paying a modest wage to a skilled worker for necessary work, and will either ship the garbage the AI gives them, or more often they'll use its existence as a bludgeon to reduce the wage further to unacceptably low levels. It happened to translation, it'll happen to artists.