For a chunk of my early career, I worked as an engineer in the standard IT-corporate mode and told myself the money justified the misfit. It didn't, at least not in the long run. And the misfit itself is worth naming, because I don't think it's rare.
The IT job demands moderation. Don't dig too deep into one thing, because tomorrow you'll be on a different thing. Don't scope-creep, because the sprint is only two weeks. Don't rewrite the whole subsystem just because you now understand it better. These are correct disciplines for the environment. They keep large teams shipping predictably.
They're the wrong environment for a specific temperament: the person whose way of thinking is to go deep into a domain and then make something of what they found. That temperament isn't better or worse than the moderate one. It's just structurally misaligned with corporate engineering.
The mismatch shows up in stages, usually in this order:
- Boredom. The work isn't hard, it's just shallow-by-design.
- Restlessness. The deep work starts happening in notebooks and on weekends.
- Distance. The fit gets more obviously wrong with every quarter.
None of this is anyone's fault. The environment is doing what it's supposed to do. The temperament is doing what it's supposed to do. They just don't fit.
What to do about it:
- Recognise the pattern early rather than late. The longer you rationalise the misfit as "easy money", the more expensive it becomes in the ideas you didn't get to explore.
- Move to work where depth is the deliverable. Agency work, R&D, indie, teaching, research. The specifics matter less than the shape.
- Accept that the "easy money" of the misfit job has a hidden cost. It's paid in the domain you never got to sit with long enough to say something interesting about.
This isn't an anti-corporate argument. Corporate engineering suits many people well, and those people should stay. The point is honest self-assessment about which side of the line your temperament sits on, and the earlier the better.
Drafted in January 2023
Updated for site in 2026