OpenAI wants you to hire an AI that keeps working after you close the app, and I want to know how much I can trust it with my inbox before I hand over the keys. I still haven’t really given my keys to the kingdom to Meta’s Muse yet, even though Meta has access to my Facebook and Instagram.
Since I’m neck-deep in experimentation directly with Codex and Claude Code, I believe I’m controlling access a bit more tightly, but I realize I’m still slowly loosening my grip. Dots and Muse are frictionless entry points for consumers to experiment with agentic AI.
Dots is a slower implementation of OpenClaw, hopefully a bit more controlled. But I’m sure people will be jumping in headfirst.