Pandora's box has been opened...
AI coding assistants are changing what it means to be a software engineer. Earlier generations were trained across the full lifecycle: gathering requirements, modeling and design, building, testing, deploying, and maintaining systems. Over time, the role narrowed into “just coding.” Now AI can generate code faster than any human, and for many engineers, that’s sparked an identity crisis.
The trade-off
The speed and productivity gains are real, but they come with costs: more duplicate code, lower stability, and less opportunity for deep skill development. As Annie Vella puts it: you can’t prompt your way to intuition. True expertise is built by grappling with the hard problems yourself, not outsourcing all the thinking to a bot.
The path forward
Balance. Engineers need to lean back into systems thinking, product context, and quality, while leaders shift the focus from speed to sustainable outcomes. Treat AI less like a magic tool and more like a teammate: sometimes a junior dev, sometimes a principal on demand — always requiring oversight. Your lil AI buddy.