🌰 focus in an agentic age

As a software engineer, LLM-use is non-negotiable at work. When I wrote 🌰 loot boxes, casinos, and work chat, Slack was the “Instagram for work”: the loot box of notifications winning against the Type-2-fun of deep, single-minded focus. LLM agents exasperbate the loot box effect: I must refresh to find out if it finished early or got stuck midway through The agent is doing the building. The building is what matters. Unblock the agent ASAP. The agent has produced 3000 words and 12 PRs. It would take sustained focus to review all this. Or…I could trust the agents reviewing the code & all the unit tests we have now, sign off on whatever this is, and find another shiny thing to tell an agent to do. People tell me that the job is now “babysitting agents” but that doesn’t feel right. When I take the time to deeply understand what the agents are doing, I often catch simplifications and corrections. But I struggle to justify making the time and, frankly, struggle to focus during that time.…