What fascinates me about high-profile failures such as Woody Allen’s “Stardust Memories” isn’t the question, “how did this great artist go so far off the reservation?” Rather, it’s unpacking the eccentricities of creation — how someone who is capable of making resonant, effective work can knowingly take the risk of failure, of experimentation, with knowledge of the risk of rejection, even hatred.
Nowadays, there’s a language for taking those risks: creators are (hopefully) aware that their work is subject to near-instantaneous criticism. It’s not unlike performance, in its many permutations: take risks, perform to your greatest ability and preparation, and learn from the failures.
This is encouraged in a lot of inspirational writing, but it’s rarely embraced. The expectation and perception remains that our creative products represent us in some unchanging way, because they’re eternally digitally preserved. Which is constricting, unless you decide “I don’t give a fuck” and hope that your mistakes will be washed away by the firehose. Which, as machine learning and data science grow ever more sophisticated, is increasingly unlikely.