AI Adoption Creates 'Grief' for Developers
The integration of AI tools is causing anxiety and a sense of loss among developers as traditional workflows change. In a .NET Rocks! podcast, guest Andrew Murphy discussed this emotional transition, referencing the essay “We Mourn Our Craft.” He argued that leaders must acknowledge this "grief" and help teams adapt by upskilling rather than just replacing tasks with AI.
The essay "We Mourn Our Craft" by Nolan Lawson gives a name to this developer anxiety, arguing the role is being reduced to a "glorified TSA agent, reviewing code to make sure the AI didn't smuggle something dangerous into production." Lawson's piece captures a sense of loss for the hands-on artistry of programming, comparing it to a blacksmith's tool that will become a curio for future generations. This sentiment is reflected in industry data, revealing a major paradox. While AI tool adoption surged to 84% among developers in 2025, positive sentiment simultaneously crashed from 72% to 60%, the largest decline ever recorded for developer tools. Trust in the accuracy of AI tools has also plummeted, with more developers now actively distrusting AI output than trusting it. A primary driver of this frustration is the "almost right, but not quite" problem, cited by 45% of developers as their biggest complaint. Many report that debugging and fixing flawed AI-generated code can take more time than writing it from scratch, challenging the technology's core productivity promise. In fact, one 2025 study from METR (Model Evaluation & Threat Research) found a significant gap between perception and reality: while developers using AI *felt* 20% faster, objective measurements showed they were actually 19% slower. The research indicated that 13% of development time was consumed by managing the AI tools themselves. This transition marks a fundamental shift from direct code authorship to a new role centered on engineering and orchestrating AI agents. The focus is moving toward system design, defining intent, and enforcing constraints for AI to execute, rather than hands-on coding. As a result, upskilling has become a critical focus. According to Gartner, 56% of organizations now view AI expertise as one of the most essential skills for developers. The demand for engineers who can bridge traditional application development with machine learning and generative AI has surged, with companies seeking to build AI competency internally rather than relying on expensive external consultants.
