I had a tough week, a week ago. My manager pointed out some of the negative side effects of my aggressive adoption of agentic coding, prompting me to rethink how I’m approaching my work.
In particular, he pointed out that the fact that while I was able to take on some audacious tasks, the fact that I don’t yet have a deep understanding of the systems being manipulated puts a large burden on the reviewers to catch things that an expert in the codebase would have done. The result is frustration and damaged trust.
He was very clear that I’m not the only person in the org experiencing this. It’s a pattern he and other managers are seeing with people who joined around the time I did.
Whether it’s just a “me problem” or a bigger issue, it was a tough pill to swallow. I hate disappointing my manager and my teammates. Secondly, I felt I was doing a great job of two of the things that we are being repeatedly instructed to do by our leadership: taking initiative and aggressively adopting the new tooling.
This past week, I took it way back to the basics (i.e. how I programmed 6 months ago). I’m trying to rebuild my workflow to prioritize my team. Instead of taking so much initiative to try things, even if they might not be successful, I’m trying to lock in on just a couple things and go way deep on them. I feel like I’m on a rebound from the low I felt coming going into the weekend.
Some of my takeaways from this situation:
- I need to prioritize the people around me. Understanding what they need and how I can be most helpful to them. Then, think about what I can do for the broader company.
- My tendency is to try to get as far as I can before asking for help. In some situations, this may be good, but the problem is that by the time I reach out for input, the work is too big to digest. I need to more proactively and interactively bring people in.
- The power of the new agentic coding tools has raced far ahead of the ability of organizations of people to adopt entirely new ways of doing work. I suspect that people who join 9 months from now will arrive into an engineering org where everything is agent-first, and we have evolved to mitigate or accept the consequences. But that is not today.
- I do feel that the time spent aggressively experimenting helped me understand what’s coming, and for a time, Hopefully, the perspective I can bring from my experience can have an impact.
I’m really hard on myself. There’s an element here of me hyperprocessing basic feedback. But, I do believe the feedback is on-point and speaks to things I need to do better at. I take my work very seriously and although adapting is difficult, one of my best skills is stepping back, reconfiguring my perspective, and adjusting.
Never Down, Always Up 