The Great AI Deskilling has begun

As AI technology becomes increasingly integrated into various industries, many professionals are experiencing a disconcerting phenomenon: a gradual deskilling that may undermine their expertise. Josh Anderson, a software consultant with 25 years of coding experience, recently shared his experience of utilizing AI to develop a software application named Road Trip Ninja, which assists families in finding reliable stops during long journeys. Initially, the process felt seamless as he leveraged AI to generate features without writing any code himself. However, as the project progressed, Anderson noticed a marked change. The pace of development slowed significantly, with interactions with the AI chatbot extending from mere minutes to several hours. The initial excitement faded, replaced by a growing sense of frustration as he struggled to navigate the expanding codebase, which ballooned to approximately 100,000 lines. This shift raised questions about the real impact of AI on skill retention and development. Anderson's experience highlights a broader concern among workplace researchers: the risk of deskilling in an environment increasingly reliant on AI. As AI tools become integral to routine tasks, some professionals find that their abilities begin to diminish. The recent outage of Anthropic's Claude AI brought this issue to light, with developers admitting they found tasks considerably more challenging without the assistance of AI, indicating a dangerous dependency. John Nosta, founder of Nosta Lab, refers to this as the "AI rebound effect," where increased productivity disguises a decrease in actual skill levels. The reliance on AI can lead to an overinflated perception of one's capabilities, distorting the way individuals assess their own expertise. Nosta warns that such reliance could lead to a regression in cognitive skills, as AI alters traditional problem-solving processes. Rebecca Hinds, head of the Work AI Institute at Glean, further elaborates on this issue, pointing out that AI can create an illusion of expertise. Workers may find it increasingly difficult to discern the boundary between their own knowledge and the assistance provided by technology. Hinds suggests that while AI can enhance productivity when used thoughtfully, it can also lead to what she calls "cognitive debt" when relied upon reflexively. The implications of this deskilling trend are particularly concerning for early-career professionals. These roles have historically served as training grounds, allowing individuals to develop problem-solving skills and resilience. Without hands-on experience, newcomers may appear competent without ever truly mastering their craft. Experts like Jan Tegze and Ben Eubanks emphasize that those who enter the workforce without a solid foundation of skills may face significant challenges. The gap between theoretical knowledge and practical application is widening, as AI tools provide quick solutions that bypass the learning process. Many companies are inadvertently reinforcing this behavior by prioritizing speed and output over genuine understanding in performance evaluations. As Anderson returned to modify his code after several months of AI-assisted development, he encountered a stark realization: his coding skills had dulled in the absence of practice. He likened the experience to a golfer who understands the mechanics of a swing but struggles to execute it. This sentiment resonates with many professionals who find themselves in a similar predicament, prompting some leaders to consider new training approaches. The future may require organizations to create "mental gyms"—spaces that encourage employees to engage in problem-solving without the aid of AI, fostering skill retention and growth. Anderson's journey serves as a cautionary tale, illustrating the potential pitfalls of over-reliance on AI and the importance of maintaining a balance between technology and skill development.