Software engineers face existential uncertainty as AI tools threaten to automate their roles. A 2026 projection suggests the industry may shrink or transform radically, leaving many questioning their career trajectories. The author, a staff engineer, notes their current work already resembles AI supervision - reviewing code and guiding junior developers - a role likely to expand as AI matures.**

The core debate centers on whether tech companies will undershoot or overshoot AI capabilities. Undershooting means continued human hiring despite AI readiness, preserving jobs but shifting focus to overseeing automated systems. Overshooting risks mass displacement, though companies might later scramble to manage AI-generated codebases. Historically, predictions of software engineering's demise (via high-level languages, outsourcing) failed, but this AI-driven shift feels fundamentally different due to tools' rapid improvement in code maintenance and bug fixing.**

The author challenges the Jevons effect optimism - that increased software efficiency will create more demand. They argue AI agents already outperform many engineers at maintenance tasks and will likely surpass human oversight soon. Their personal experience shows AI tools transitioning from "sometimes faster" to "usually faster" at solving coding problems, eroding the safety net of human-in-the-loop workflows.**

While hoping for a human-centric future, the author acknowledges possible industry collapse. They express bittersweet acceptance of automation's inevitability, recognizing their current privileged position compared to junior peers. The piece ends with cautious hope that uniquely human creativity might still differentiate software development, but prepares for radical career pivots if AI continues its relentless advance.