When I started my career in IT, I thought the hardest problems in computer science were cache invalidation and naming things.
Years later, I understood that the hardest problems are human ones.
Along the way I hit plenty of technical walls: bugs I couldn’t explain, side effects that made no sense, complex problems solved with even more complex abstractions, projects that looked impossible at the start. Every one of them eventually gave in.
Today, with AI writing, reviewing, and even debugging code alongside us, that ceiling has moved even further. What used to take a senior engineer a week now takes an afternoon. Code, as a bottleneck, is basically solved.
Since I became a people manager, I’ve learned the hardest problems are about people and organizations. You can tail a log and debug a piece of code. People don’t work that way.
A candid feedback beats any test suite. A quiet team is a louder alarm than any alert.
None of this gets solved once and for all.
Code is solved. People are not.
That’s what makes the job worth doing.