Sure, garbage in, garbage out and all that. The autonomously generated stuff tends toward generic as an inherent byproduct of being a closed loop system. But that doesn’t mean a real artist couldn’t look at some boring ass slop and be inspired to explore new directions.

I think one of the common themes I’m circling these days is that “human in the loop” is a common concept around ensuring outputs from AI systems are acceptable, but a better way to look at it is that generative AI should never have a direct connection to final output. As inspiration or iteration, I think there’s potential value, but ultimately, whether it’s code, art, or content, a human should create what goes out. Using AI for intermediate acceleration is a much healthier approach than the “look how many people we can replace!” angle that’s so popular in tech.

This doesn’t solve any of the many other issues with generative AI these days, but it at least feels like a more sensible approach to the creative concerns.