A reflective essay by veteran developer Les Orchard is cutting through the noise of AI coding debates by reframing the conversation entirely. Rather than arguing whether AI coding tools are good or bad, Orchard — who has been programming since 1982 — contends that the real story is one of exposed identity: the rise of AI coding agents has simply made visible a divide that was always latent in the developer community. On one side sit the 'craft lovers,' those who code for the tactile, artistic satisfaction of writing elegant software. On the other are the 'result chasers,' developers who have always been motivated purely by outcomes. For the latter group, AI coding agents are not a disruption — they are the next rung on a ladder that has always been climbing toward higher abstraction.
Orchard engages directly with peers who are grieving the transition. Nolan Lawson's widely-read essay mourns the loss of deep immersion in code — the personal craft of a repository, the particular satisfaction of a late-night debugging session. James Randall, coding since 1983, describes the sense of discovery as compressed rather than eliminated, something lost even as something is gained. Orchard does not dismiss these feelings, but argues the camps are mourning fundamentally different things. The craft lover grieves the act itself; the result chaser, if they grieve at all, grieves the surrounding ecosystem — the open web, the shifting career landscape, the consolidation of internet power — losses that exist independently of whether anyone types code by hand.
The framing matters for anyone trying to read the actual pace of AI coding adoption. If a significant share of working developers are latent result-chasers — people who endured the craft because there was no alternative, not because they loved it — the real resistance to these tools may be smaller than the debate implies. Orchard reports that after 18 to 24 months of adjustment, his fears largely did not materialise: not the loss of comprehension over AI-generated code, not the erosion of his ability to judge correctness, not the disappearance of puzzle-solving satisfaction. Decades of reading and reviewing code proved durable. The puzzle migrated upward, to architecture and composition and directing the assistant.
What his account suggests is that the public conversation is probably skewed. Craft lovers have concrete losses and the vocabulary to articulate them; result-chasers who found the transition workable have less reason to write about it. Orchard is a rarer data point — high-context, long-tenured, genuinely candid about what he feared losing — and his conclusion that the other side is recognisable, if not identical, to what came before is worth more than a thousand more takes from people who haven't crossed that threshold yet.