When you hand off a task to an AI agent, you tend to stop thinking about it. That's not laziness — it's a well-documented cognitive reflex. But Herbert Simon, the Nobel-winning economist who coined the term 'bounded rationality,' noticed something important: the moment we delegate a decision to a system, we also tend to delegate our judgment about whether the system is deciding well. The risk with copilot agents isn't that they'll fail spectacularly. It's that they'll succeed just well enough that you stop auditing them. The concrete habit worth building now: before accepting any agent-generated output, spend ten seconds asking 'what would I have done differently, and why?' Not to second-guess the tool — but to keep your own judgment from quietly atrophying.

In the last week, when an AI tool gave you an answer you accepted, what assumption did you let it make on your behalf without checking?

Drawing from Pragmatism / Decision Science — Herbert Simon