source: simon willison: why ai hasn’t replaced software engineers, and won’t
level: business
arvind narayanan and sayash kappor argue that ai is not causing mass layoffs in software engineering, a field with few regulatory barriers. in new york, the first state to add an ai disclosure checkbox to warn act filings, not a single company checked the box in over 160 filings during the first year. this suggests that even in a sector ripe for disruption, ai is not leading to widespread job losses.
the real bottlenecks in software engineering go beyond writing code. surveys point to meetings, debugging, and other tasks as major time sinks. digging deeper, the core challenges are deciding and specifying what to build, verifying and being accountable for what is delivered, and maintaining a deep human understanding of the codebase, business, and environment. these tasks resist automation because they require context, judgment, and accountability that ai cannot fully replicate.
simon willison notes that while ai helps with deciding and verifying steps, the deep human understanding remains key to his value. even with advanced ai assistance, the quality of his work depends on how well he grasps the problems and the solutions that agents produce. this highlights that ai augments rather than replaces the critical thinking and contextual knowledge that engineers bring to their work.
why it matters: for ai and data science, this shows that automation tools augment rather than replace human expertise, emphasizing the need for deep domain understanding to effectively leverage ai.
source: simon willison: why ai hasn’t replaced software engineers, and won’t