This website uses cookies
Read our Privacy policy and Terms of use for more information.
Two things decide how anyone answers that, and neither one is the technology.
Most bad AI output is a review problem, and reviews are a management job.
The same loop I run with product teams, taught to a room that had never built anything.
How to move towards multiplayer AI and why more AI-enabled workflows won't get you there.
The trip didn't fix anything, but for once the rest actually held.
Sorted by what it's for, with what we actually use, what we dropped, and what I turned down.
Token efficiency is everybody's problem.
Everything I showed you last week ran exactly the way it was supposed to.. and I still ended Saturday too fried to pick up my own to-do list. A note on the part of this job that never shows up in a dashboard.
One day inside the system I actually run on.. 7am to midnight, $21, 863 blocked commands, and the hour I couldn't reach my own box.
