Technology

The big issue is, as you said, a capitalism problem, as people need money from their work in order to eat. But, it goes deeper than that and that doesn't change the fact that something needs to be done to protect the people creating the stuff that goes into the learning models. Ultimately, it comes down to the fact that datasets aren't ethically sourced and that people want to use AI to replace the same people whose work they used to create said AI, but it also has a root in how society devalues the work of creativity. People feel entitled to the work of artists. For decades, people have believed that artists shouldn't be fairly compensated for their work, and the recent AI issue is just another stone in the pile. If you want to see how disgusting it is, look up stuff like "paid in exposure" and the other kinds of things people tell artists they should accept as payment instead of money.

In my mind, there are two major groups when it comes to AI: Those whose work would benefit from the increased efficiency AI would bring, and those who want the reward for work without actually doing the work or paying somebody with the skills and knowledge to do the work. MidJourney is in the middle of a lawsuit right now and the developers were caught talking about how you "just need to launder it through a fine tuned Codex." With the "it" here being artists' work. Link The vast majority of the time, these are the kinds of people I see defending AI; they aren't people sharing and collaborating to make things better - they're people who feel entitled to benefit from others' work without doing anything themselves. Making art is about the process and developing yourself as a person as much as it is about the end result, but these people don't want all that. They just want to push a button and get a pretty picture or a story or whatever, and then feel smug and superior about how great an artist they are.

All that needs to be done is to require that the company that creates the AI has to pay a licensing fee for copyrighted material, and allow for copyright-free stuff and content where they have gotten express permission to use (opt-in) to be used freely. Those businesses with huge libraries of copyright-free music that you pay a subscription fee to use work like this. They pay musicians to create songs for them; they don't go around downloading songs and then cut them up to create synthesizers that they sell.