AI Raises Deskilling Risk for Doctors and Developers

According to The Economic Times, researchers and analysts are raising alarms about AI-driven "deskilling" as tools become routine in healthcare, software development, and other fields. The article reports that multiple studies involving doctors and software engineers found that while AI assistance can improve immediate performance and efficiency, users tended to struggle more when working without the tools and retained less task knowledge over time, per The Economic Times. The Economic Times cites Syracuse University information scientist Kevin Crowston urging professionals to reflect on which skills they want to preserve versus outsource to AI.

Industry-pattern observations: automation historically creates short-term productivity gains while shifting the locus of human expertise to oversight, exception handling, and tool design. For practitioners, that shift typically increases the importance of robust evaluation pipelines, longitudinal skill-assessment metrics, and tooling that exposes model reasoning rather than hiding it behind opaque suggestions.

For organizations deploying assistive models in high-stakes domains such as medicine and safety-critical engineering, the reported findings amplify existing concerns about reliance, degraded mental models, and knowledge erosion. Industry observers note comparable patterns in other automation waves, where deskilling raised costs for retraining and incident response when tools were unavailable or failed.

Signals to monitor include independent studies measuring skill retention under assisted versus unassisted workflows, adoption of explainability features in developer and clinical tools, and experiments that mandate occasional unassisted task performance to quantify competence drift. Observers should also track whether vendors publish longitudinal user-impact data or introduce controls that let organizations limit suggestion scope to preserve human practice.