Bosses think AI will boost productivity — but it's actually deskilling workers, a professor says (www.businessinsider.com)

UC Irvine philosophy professor Anastasia Berg warned that widespread reliance on AI is quietly eroding workers’ core skills, particularly among junior employees who lean on tools from day one and never build foundational knowledge. Citing a body of empirical work and conversations with academics and industry colleagues, she argues AI often speeds tasks and boosts engagement but can sacrifice depth, critical thinking, creativity and long-term skill retention. Supporting concerns about non-work dependence, an analysis of 1.58 million ChatGPT conversations (OpenAI, Duke, Harvard) reported that by June 2025 roughly 73% of adult messages were non‑work related, suggesting pervasive cognitive offloading beyond the office. The technical and organizational implications are stark: early-career coders increasingly rely on LLMs to write and debug code rather than learn underlying logic, leaving them unable to verify or correct model outputs. Skill atrophy removes the “friction” that develops reasoning and problem‑solving abilities, creating workers who appear productive but lack independence, resilience and verification skills—raising risks around errors, security, compliance and innovation. If companies prioritize short‑term efficiency without redesigning training and assessment, Berg warns AI may automate not only tasks but the learning processes that produce competent professionals.