This piece is great. And it’s totally true that we’re surrounded by walls of slop. I stop reading things all the time when I notice the tinge of AI.
That said, I don’t think that just because this is the way things are, this is this is the way things have to be. Even with our interactions with AI. And I have hope.
Ted Chiang had a great piece in the New Yorker years ago called “Why AI isn’t going to make Art”[1] where he describes something similar. There are 10,000 choices to make in every good pieces of art. AI makes the average choice for every decision and the average choice is derivative and uninteresting. There in I think lies both the problem and the solution. Rather than autocompleting our work, we’d all be so lucky to have a captivated and motivated reader who asks all of the questions this author suggests, so we can get better and better at asking them ourselves and making great art.
If it were just a problem of making too average choices, it could be fixed by sampling at a higher temperature. But I think there's a bigger problem in that the correct choices depend on the context in a complicated way, and LLMs can't handle arbitrary deep dependency chains, so by default you get the average choice based on superficial cues instead of the average choice based on careful planning. Which is why for coding, chain-of-thought and tool calls are necessary to make the implicit explicit.
Making artistic choices explicit and subjecting them to a planning process with iterative refinement would probably already improve quality a lot (consider image generation models using a prompt rewriting to create more varied and detailed outputs) but to get the most out of it, you'd likely need a model specifically trained for that process. And considering that coding is much more lucrative, it might take a while before we get there.
That said, I don’t think that just because this is the way things are, this is this is the way things have to be. Even with our interactions with AI. And I have hope.
Ted Chiang had a great piece in the New Yorker years ago called “Why AI isn’t going to make Art”[1] where he describes something similar. There are 10,000 choices to make in every good pieces of art. AI makes the average choice for every decision and the average choice is derivative and uninteresting. There in I think lies both the problem and the solution. Rather than autocompleting our work, we’d all be so lucky to have a captivated and motivated reader who asks all of the questions this author suggests, so we can get better and better at asking them ourselves and making great art.
[1] https://dmd-program.github.io/dmd-300-master/assets/Chiang-A...