In June of 2026 I wrote an article about AI usage while working on the initial part of the memoir—specifically after our first draft got written and revised. The actual work organizing, writing, modifying, and cutting was always my own (and my wife’s). No fine-tuning of a model would reproduce our story or our voice and we wouldn’t want it to. Memoir is far too personal and is a relationship with the reader based on trust.
Even after that work—our second major revision—we used beta readers for more feedback. At the end of the summer, we contracted with a professional editor to do more extensive work on the manuscript to reduce the word count and move the story more meaningfully. At the time of this writing, we’re over two months into that process with more time to go.
As time passed, I felt like the article provided less help than originally intended. The expectations around AI usage are always moving (though these are good practices)—and even change from agent to agent or publisher to publisher. There’s also been so much more substantial editing that I thought the article aged poorly. Our second revision was 10,000 words shorter than our first, and our current goal is 20,000 off of our second. What the article provided was one snapshot of a very long and (currently) unfinished process that has taken countless hours from at least a dozen people over fourteen months.
Thus, I saved a copy of the original article but have removed it from the Substack. If anyone wants to look at the original version, please let me know.