More road reviews less code reviews

I've been reviewing a draft essay from a friend and ex-colleague of mine. He's a super experienced senior software architect and he's written about how so much has changed since AI. It's a great read! He articulates some of the worries happening across the world of software development right now.

He touches on why the role of code reviewing has become important yet also explores why it's a waste of talent.

Let's refer to it as the fallacy of the code review gate. He's got a compelling argument that it's a waste of talent for senior engineers to be doing it yet utterly essential given he describes AI agents as overly confident junior developers. 

I tried to come up with an analogy to help his writing. 

It’s like commenting on a painting of a thousand dots no one wanted. Maybe the painter enjoyed using AI to create that painting and they could do it again but better. But maybe someone could’ve questioned the need to paint a picture of a thousand dots in the first place. So why am I reviewing it? More road reviews than code reviews. Like what road should we be following to get us to where we think we need to go.

Should our senior roles be getting closer to working out what to build and have more influence over the a product roadmap? Probably. More road reviews than code reviews. It certainly beats spending time reviewing volumes of code that AI created in the hope that those features might be the ones our customers need.

I look forward to sharing his essay once it's ready.

P.S. This is a good reminder to review other people's writing. Make lots of notes whilst doing so and it sparks more ideas to share, like the reflections on this Moment.