Managing agents, live today at 1 PM ET
There was a point this year, for me it was around July, August, where models got smart enough that I stopped reading everything my agents do. I don't review the specifics anymore. As long as I have a good idea of exactly what the outcome and the eval look like, I let them go. Agents are not a set-and-forget kind of thing. You create them, then you keep a watch on them for a little bit, normally it's a few days. Once they're working like you want, you let them go for a week and you come back and you check them. Agents don't want anything. It's up to us as humans to want things and to give them direction. Today at 1 PM ET I'm doing the second AI in Context live: managing agents, for people who
Keep watching
More from AI in Context
ClipYou have not yet sufficiently defined what is my business objective.
From Ep 01: Keep it in the same chat
Oct 15, 2026
ClipI do not go back to chat.
From Ep 01: Some form of database that lives atop all of your agents
Oct 14, 2026
ClipWhat is the outcome I want to achieve?
From Ep 01: It's not about the tooling
Oct 14, 2026
The inbox audit
Your team's year, read and mapped.
The inbox audit shows where your team's work actually goes, then we walk you through it.