Build at night, decide in the morning

After years of remote work, my rule is simple: build at night, decide in the morning, and never merge anything I wrote after midnight without rereading it in daylight.

Two different peaks

My best code gets written between 10pm and 1am. My best decisions get made at 9am. Learning not to confuse the two took me embarrassingly long.

Late night me writes brilliant solutions to problems that morning me realizes we do not have.

Why the split works

Deep focus and good judgment are not the same mode. At night there are no interruptions, so I can hold a whole problem in my head and move fast. In the morning I am slower, but I am better at asking whether the thing should exist at all. Using each mode for what it is good at means I keep the speed and lose most of the regret.

Takeaways

• Your best building hours and your best deciding hours may be different.
• Use quiet late hours for implementation, not for scoping decisions.
• Reread anything written after midnight in daylight before merging it.