The craft of software development has fundamentally shifted. Historically, the challenge lay in bringing an idea to fruition, a process involving extensive typing and debugging. Production was the barrier separating creators from talkers. Now, thanks to advancements, a plausible version of an idea can be generated almost instantly, collapsing the idea-to-artifact distance.

This ease of creation, however, presents a new problem: the value of learned skills is not lost but transformed. The output is "good enough," which erodes the incentive to strive for better. Previously, the expense of creation acted as a natural filter, rationing output and ensuring that existing work had survived a cost. This economic floor enforced a standard. With its removal, the arbiter of what is worth keeping is no longer cost, but human judgment—taste.

Taste is not about preference but a wordless, intuitive recognition of quality, akin to Robert Pirsig's "Quality." It’s the immediate sense that something is wrong, even before a justification can be formed. This skill, honed through the painful process of making mistakes and living with their failures, was the non-automatable core of development. The friction of creation was the curriculum. Removing this friction for new developers bypasses this essential apprenticeship, leading to fluency without the underlying judgment.

This shift creates an economic disadvantage for those with taste. While they spend extra time refining, those without taste ship faster, their efforts indistinguishable in market time. Taste is invisible on dashboards and in code diffs. The abundance of "slop"—indifferent, "fine" output—now floods the field, making the scarce act not creation, but curation. Deciding what deserves to exist from an infinite sea of plausible options is the new challenge.