Finally read Co-Intelligence
Mollick's four rules for working with AI, checked against a year
Ethan Mollick's Co-Intelligence came out in the spring and I finally got through it. It's the first book I've read that describes AI use the way I actually experience it, as a daily working relationship rather than a technology trend.
His four rules held up well against my year. "Always invite AI to the table" is roughly how I already operate, since I bring it into PRDs, prep, and analysis by default and let the results argue for themselves. "Treat it like a person, but tell it what kind of person" is basically a description of writing custom GPT instructions, which I've been doing since late last year without having a phrase for it.
The one that stuck with me most is "assume this is the worst AI you will ever use." I spend real hours keeping my GPTs' knowledge current, and some days that feels like too much upkeep for a tool. That rule reframes the upkeep as practice for whatever comes next, and given what the tools have done in just the past year, I think he's right.
Link: https://www.penguinrandomhouse.com/books/741805/co-intelligence-by-ethan-mollick/