Scott Bolinger's new piece opens with a line that will sound familiar to anyone who has watched a coding agent take over a complex task and not embarrass itself: "Dear software engineer — you still have value, even if AI writes all the code." The GoDaddy Principal Software Engineer isn't writing from an anxious sideline — he's writing from inside daily use of Claude, Amp, and Cursor, and his argument, [published this week on his Substack](https://substack.com/home/post/p-184377592) and the GoDaddy engineering blog, is less reassurance than diagnosis.
The diagnosis is this: current coding agents hit a ceiling in production environments. They loop. They fix one test while breaking another. They don't know that the approach that passed code review last quarter was quietly deprecated in a team meeting, or that the commerce API they're confidently calling has a rate limit nobody documented. An engineer who has been in the codebase for eighteen months carries context that a language model, however capable, has to be told. That's not a temporary limitation — it's a structural one, and it means the job of technical steering isn't going anywhere.
What is changing is everything beneath that steering. Bolinger reaches for the "product engineer" frame that Lee Robinson popularized in [an essay at Vercel](https://leerob.com/product-engineers) — the idea that the old frontend/backend full stack is giving way to a new one: product plus engineering. Robinson's original argument was about working backwards from the desired experience rather than up from the technology stack. Bolinger's version is sharper for the AI moment: when the gap between idea and shipped software collapses, the value of an engineer who can hold the whole product in their head multiplies. He built an internal prototype in a matter of days using OpenAI tool calling against commerce APIs; it became a company-wide product. A colleague's market research initiative surfaced as a quarterly priority. These outcomes didn't happen because AI wrote better code — they happened because someone had the product instinct to point it at the right problem.
Obie Fernandez made the same observation last December in a piece that Simon Willison [surfaced here](https://simonwillison.net/2025/Dec/13/obie-fernandez/): "agents feel like they clear space. They take care of the mechanical expression and leave you with judgment, tradeoffs, and intent. Because truly, for someone at my experience level, that is my core value offering anyway. When I spend time actually typing code these days with my own fingers, it feels like a waste of my time."
Bolinger's warning lands somewhere between encouragement and blunt assessment. Engineers who have anchored their identity in craft — in the satisfaction of clean code and elegant architecture — face real displacement risk if they treat that craft as the primary offering. Engineers who can think in products, communicate across functions, and exercise judgment that only comes from experience are, by his account, about to become significantly more valuable. The execution bottleneck is gone. The judgment bottleneck is where the work is now.