iWonder: In 5 years, would GenAI start making human-produced art obsolete?
I have a question about GenAI and artistic production that I am not seeing enough discussions on. Five years from now, perhaps sooner, humans simply won’t be able to compete with AI for incredibly beautiful and mind-boggling art — infinite streams of personalized on-demand stories, music, pictures, video clips, even full movies rendered inside immersive 3D worlds. With sufficient randomization and saliency/novelty measurement (for simplicity, imagine picking points in an embedding space that is distant from all existing points) humans won’t be able to compete. What happens at that point to creative people?
Today we have systems that can generate text based stories, and we have systems that can create videos from text. Combine these two, feed in user feedback, and the above is already possible. So we are almost there.
Those in defense of GenAI say thay in the next 1–2 years, there will be ‘AI assisted’ creation. But 3 to 5 years from now, there simply won’t be a need for that kind of supervision. There will be humans in the loop, but only on the consumption side. Implicit personalized feedback is all that the systems would need for generating infinite streams of increasingly engaging personalized content.
Some questions:
- Do you think this is inevitable like I strongly feel?
- How do you feel about AI generated content? Right now I personally feel a lot more connected to and fascinated by human generated content no matter how impressive the output is. But over time would most of us just get used to AI generated content?
- Related to the above, would the definition of what is beautiful art evolve to be something that prioritizes human generated art forms?
- If the above happens, what would become of all the artists out there?
