Responding to Matt Shumer's "Something Big Is Happening" piece that's been circulating.
The pace of change is real, but the "just give it a prompt" framing is self-defeating. If the prompt is all that matters, then knowing what to build and understanding the problem deeply matters MORE. Building simple shit is getting commoditized, fine. But building complex systems and actually understanding how they work? That's becoming more valuable, not less. When anyone can spin up the easy stuff, the premium shifts to the people who can architect what's hard and debug what's opaque.
We also need to separate "building software" from "building AI systems", completely different trajectories. The former may be getting commoditized. The latter is not. How we use this technology, how we shape it, what we point it at, that's specifically human work.
And the agent management point: if these things move fast and independently, the operator's ability to effectively manage them becomes the fulcrum of value. We are nowhere near "assign a broad goal and walk away for six months." Taste, human judgment, and understanding what other humans actually need, those make that a steep climb. Unless these systems are building for and selling to other agents, the intent of the operator and their oversight remain crucial.
Like everything before AI: it isn't the tool, but the hands.
Original article: https://www.linkedin.com/pulse/something-big-happening-matt-shumer-so5he
[link] [comments]