When not to act
In a moment obsessed with agency and output, the most agentic thing I did was to stop producing publicly for a while :)
Everyone’s talking about how AI makes agency easier for us all. What if one increasingly valuable skill is knowing how to channel all the agency now available to us? Yes, including when not to act just yet.
There are a lot of incentives right now to do more - use more AI, use the latest dot/muse/instinct/grokbot, keep up, form a take on every launch, yet every breaking development has 47 takes attached before you have formed your own. AI makes producing seemingly polished thoughts easier at the same time that being chronically online makes forming your own thoughts harder.
AI boosted my agency, but it also made it easier to avoid sensemaking. These last few months, I had to go quiet so I could get back to hearing myself think clearly, and the conviction I arrived at was pretty simple: agency needs a compass. We don’t have a map to navigate these times, but a compass can help with:
- Attention: knowing what deserves my attention rather than reacting to everything
- Problem selection: knowing what problems I actually care about, and distinguishing “I can do this now” from “this is worth doing”
- Sensemaking: sitting with information long enough to form my own interpretation, and using AI to extend my agency without outsourcing my judgment
Sometimes, more agency means resisting motion for motion’s sake, instead channeling that energy into something worth working on.