Nobody warned me the second half of an engineering career is mostly subtraction. The first half you spend acquiring — languages, patterns, opinions. The second half you spend putting things down, one at a time, usually after they've embarrassed you.
I stopped being the best engineer in the room. Not by choice, at first — the rooms changed. For a while I competed quietly, reviewing code I had no business reviewing, keeping one corner of the system that was mine. It helped no one. The day I gave it away, the team got faster and I slept better, in that order.
I stopped rescuing things. There's a particular high in parachuting in at 11pm to save a release, and I chased it for years before noticing the cost: every rescue taught the team someone would always catch them, and taught me nothing. Now when something's heading for the wall, I say so early, once, and let the people who own it steer. Sometimes they hit the wall. Walls turn out to be good teachers.
I stopped mistaking motion for progress. Banks manufacture motion — committees, working groups, status decks about status decks. I used to measure my weeks by how full they were. Now I ask a meaner question: what decision got made that wouldn't have. Some of my emptiest-looking weeks score highest.
And I stopped having opinions about everything. A senior person's casual preference has a way of becoming a junior person's requirement. These days I try to hold a few real opinions and make them count. The rest of the time the most senior thing I can say in a design review is the thing I avoided for years: I don't know. What do you think.