Engineering video
1:00 · 19 August 2025
Debugging and reading error stack traces are scarce, highly needed skills for fixing production issues, especially since developers who rely strictly on AI will not know how to solve them.
Hear it at 0:18 →Developers must get extremely good at reading code—which is done much more often than writing it—and quickly refactoring it as AI generates more code.
Hear it at 0:46 →you'll have to fix lots of the mess generated with AI, and most people won't know how to do it, because they really didn't take the time to get better. They're paralyzed, or they just got distracted by AI and they're not really uh studying coding anymore, they don't want to get better at it. They just disregarded it. They was like, oh, it's worthless because AI does it. Uh you want to get really crazy get at debugging.
It's a very scarce skill. And it will be very much needed when things will break in production. Uh the people that vibe coded their way into the job won't know what to do. But the people that know software engineering, they will immediately find problems. with the debugger. Get crazy get at reading error stack traces, then just copy paste it and send it to GPT and ask it to fix it for you. It is a very good skill.
A good developer is a find solutions almost instantaneously. And get crazy good at reading code. It's something we do a lot more than writing code. It's always been like that. But now with AI, it got even worse. So you have to be very fast at reading code and then refactoring it really quickly and committing.
