The real value isn’t in the code - Jon Ayre

Good post from Ayre highlighting what a lot of people don't see when teams create software. I think in part there is an almost "magician" aspect to the act of writing software. All the funky syntax, specialized tools and such. To a non-technical person it really does look like magic. But to people experienced in the craft it is the realization of a lot of other inputs. Reading Ayre's essay he uses an example project that he was involved in to show how small the coding part really is. There is a lot of truth to this. People love the "build a house metaphors" with software, even though I think it is fundamentally flawed (your house doesn't move for instance), but to pull from it the coding is probably a bit like just looking at the framing and drywall going up. That isn't a house unless you do all the other stuff.

There are a lot of metaphors for software. One I use often is that software isn't even an asset, but is instead a liability. There is a solution that needs to exist, and that solution is understood through careful understanding of the problem, designing a great answer, and then at some point it becomes instantiated in code. Now we have a problem since that code will have issues (there is no perfect code) and will have to be iterated over time. Plus, this solution will likely evolve and the code will have to reflect that.

In this model I would suggest the code is actually a liability. The asset that we have is the design thinking, solutioning, domain understanding that went into it. As long as we have that asset we can modify the code, the liability, to continue to evolve. Less code? Less liability.

The backdrop to Ayre's post and my own commentary is the role that AI is increasingly playing in coding. The code part is a small but critical part of the solution. And as Ayre and others have pointed out, the limit there is time and AI enhanced development tools are going to help people in this craft to do more with the time we have.