So my main point is to stop asking whether and how code review in its current form can be saved, but to have an open conversation about what we are trying to achieve here and the costs and benefits of alternative models.

What would these other models be? The article assumes LLM-generated code leads to the loss of current review processes, which they call ‘modern reviews’.

I feel like my team does and will continue to do reviews for the same reasons and with the same gains as before. LLM may generate code, but I expect my colleagues to take ownership of whatever they produce, to understand it, take responsibility, and describe it. (Which is an issue/gap for some, but that was an issue even before generating code.)

Don’t forget the value of actually knowing your codebase if you wrote it… or the value lost if you just have an LLM do the whole thing.

I feel like a big problem is that LLMs only produce previously learned logical patterns, they don’t do logical reasoning of their own. And code review was always bad at catching logic bugs as well.

As a result, even with us diligently doing code reviews at $DAYJOB, we have more logic bugs than I have ever seen in any codebase before.

I’ll spend three weeks discussing an existing feature with other devs and fixing up logic bugs before we realize that the entire feature is conceptually illogical and needs to be removed/replaced.
This has happened twice for me in the past three months, which should tell you a lot about how much time is wasted here.

TL;DR: I feel like we need something different than code review to catch these logic bugs. Maybe more of a presentation of the feature implementation to the other devs, and then a discussion of that.

I see all kinds of shit code come out of peoples’ coding agents. Thankfully, we are under no pressure to remove human reviews.

Nice article, thanks. I like how the writer came up with an economic metric for understanding code review in terms of dollars in proportion to the volume of code.

That is the right way to discuss this with upper management.

The other barrier is the illusion of objectivity. In the other comments, and in many programmers experience with AI, there are code short cuts and issues that are objective in that no one would disagree with their value and importance, but they are subjective in that they cannot be described so cleanly that an LLM or human can’t misunderstand them. That it can be described so perfectly as to be findable 100% with no false positives or missed positives.

And then on top of that are the purely subjective goals like ‘simplicity’. Give programmers the goal of simplicity and they wind up simply pushing the complexity onto someone else’s desk. LLMs (subagents, etc…) are even worse at that.

And it goes all the way up and down, the more you elevate or granulate to architectural decisions or algorithmic decisions, different illusions appear.

Code Review has never really been 100% successful, nor has Quality Control/assurance, Testing, or any other measure. At best it just forces people to look at code with another perspective to hopefully distill myth from reality. And that is itself valuable, if not for being able to show due diligence and effort.

Which brings us back to the article’s equation of value. Given production, liability, are calculated in business terms from the real dollars spent or projected to be spent holds up, its a good start, no?