Kendall Miller argues that as AI generates code, engineers need to shift focus to building mental models of problems and understanding what good solutions look like. The real skill is articulating problems clearly, knowing what correct answers require, and staying curious enough to analyze challenges deeply.
Chapters
So to start at the first side, it is, we have seen that more and more what AI is doing, what generative AI is doing, is emphasizing the job always of an engineer or developer was never about writing the characters of text on the screen. It was about creating a mental model to solve a problem incrementally or understanding the problem, the actual problem to solve, and ultimately knowing what right looks like. On screen. And we found that those skills you still have to have. The people that were very— it turns out people that are very successful on the never write the code yourself path and the people that are very successful as principal engineers where they wrote every line themselves. And again, it turns out you gotta be really good at knowing what right looks like. You have to be able to express the problem you're trying to solve at the right level. We need people to be curious. We need them to want to solve problems and be able to analyze problems to solve. We need them to have the engineering background to understand what right looks like.