Sadly, I've seen multiple well-known developers here on HN argue that reading code in fact isn't hard and that it's easy to review AI-generated code. I think fundamentally what AI-generated code is doing is exposing the cracks in many, many engineers across the board that either don't care about code quality or are completely unable to step back and evaluate their own process to see if what they're doing is good or not. If it works it works and there's no need to understand why or how.

I think this is equally true of writing. Once you see something written one way, it's very hard to imagine other ways of writing the same thing. The influence of anchoring bias is quite strong.

A strong editor is able to overcome this anchoring bias, imagine alternative approaches to the same problem, and evaluate them against each other. This is not easy and requires experience and practice. I am starting to think that a lot of people who "co-write" with ChatGPT are seriously overestimating their own editing skills.

Reviewing code is basically applying the Chesterton’s fence principle to everything. With AI code there’s typically so much incidental noise that trying to identify intention is a challenge.

But then again I’ve found a lot of people are not bothered by overly convoluted code that is the equivalent of using a hammer for screws either…