Taste Is All That's Left

For most of my software writing career, I've come to believe that the hardest part was making the software exist in the first place. This belief is not new. I learned through building web applications, watching them crash and fail. There was an idea, hours of typing, reading manuals, and misunderstanding APIs. Production was the wall that separated those who could create from those who could only talk about it.

That wall is gone now, rented out, making it easier to create. However, there's a catch: the value you built by learning to navigate that wall doesn't disappear. It moves, and your judgement becomes the new filter. You now decide what's worth keeping based on your taste, not the cost of making it. Taste is that compressed, wordless verdict you reach faster than you can justify.

It's the "no, again" feeling you have about a sentence or function. It's the hardest part of the work because it's the only part that was never mechanical. The friction of learning was the curriculum, teaching you what was worth paying for. When the friction is removed, people can generate fluently from the first day. They won't ship bad versions and learn from failure, as they won't have to experience it.

They will arrive at fluency, having skipped the apprenticeship that used to be required, but they will be more productive than before. However, they won't be able to tell if their work is good or not. Removing the friction that taught them taste has led to a situation where people ship at the same speed as those who can't. Taste is slow, saying "no, again" and sending plausible things back because they're not right.