A minute (or less) for AI optimism — More time to focus (the math is worse than I thought)

We used to have a tidy rule for this. One project, a developer gives it 100%. Add a second, and you don’t get a clean 50/50 — you lose about 20% off the top just switching between the two. Gerald Weinberg wrote that down decades ago, and it always felt a little abstract.

The newer research makes it real (and scarier).

Gloria Mark’s work at UC Irvine found that it takes about 23 minutes to refocus after an interruption fully. And our attention span on a single screen has dropped from ~2.5 minutes twenty years ago to about 47 seconds today. Microsoft’s 2025 data landed the punchline: the average knowledge worker now gets interrupted every two minutes — roughly 275 times a day.

I did some quick calculations on one source of that noise.

Say you handle 1,000 messages a month. Reading and replying is maybe 30 seconds each — call it 8 hours. That’s the cost we actually see and budget for.

Based on 1,000 messages per month.

But if even 1 in 10 of those messages breaks your focus mid-thought, you add ~23 minutes of recovery each time. Now you’re closer to 47 hours. More than a full work week, every month, just climbing back to where you already were.

And here’s where I land on the optimistic side…again.

A lot of what fragments our attention is small stuff that doesn’t really need us — triaging messages, drafting routine replies, pulling context together before the real work can even start. That’s exactly the layer AI is getting good at. Fewer things tapping you on the shoulder all day.

Clearing the noise is only half of it. The other half is protecting the quiet that’s left, rather than filling it with more inputs and more screens.