[ AUG 14, 2026 ]
Mr. Meeseeks (Metaphor for AI)
Author Pascal Rettig | Written by a Human
I'm not a big Rick & Morty enthusiast, but a colleague pointed me to this episode as a metaphor for working with coding agents, and boy, does it fit.
The key takeaway is that the longer an agent stays alive, the more and more pain you're putting it in—don't treat it like a friend you'll be sad to see go. Give it a task, let it execute, and then let it go.
The more comfortable you are with /clear and starting anew, the easier it will be to develop repeatable habits and practices that you can put to work via a repeatable process—and the easier it will be to slowly give agents more and more autonomy.
Your important context should live in your support and steering documents, not in your session transcript, where it will be lost to history at the next compaction.
AUG 14, 2026
AI is planting the seeds of its own destruction
One of the obviously exceptional things about the current AI investment boom that we're in is how much money is being dumped into AI and LLMs, and how much...
AUG 12, 2026
VORA: Value Over Replacement Agent
I’m not the world’s biggest baseball fan, but I’ve always loved VORP - Value over replacement player. It’s an idea I’ve riffed on numerous times for various situations - basically take an average of X and evaluate individual emails, flows, etc against that average. What do those specific X bring to the table vs. the average? What makes them special and unique?