The widespread adoption of generative AI is changing more than how people complete tasks—it is also reshaping how expertise develops. In this Communications of the ACM blog, the author argues that the most significant consequence of the generative AI revolution may not be automation itself, but its influence on the learning process. Traditionally, expertise is built through years of practice, trial and error, and repeated problem-solving. As AI increasingly provides instant answers, code, analyses, and creative outputs, there is growing concern that people may bypass the experiences that help them develop deep understanding and professional judgment.
The article explains that true expertise is more than accumulating information—it involves recognizing patterns, making sound decisions under uncertainty, and understanding why a solution works. While generative AI can dramatically improve productivity and reduce routine work, overreliance on AI-generated outputs may weaken opportunities for learners to build these higher-order cognitive skills. If individuals consistently accept AI recommendations without critical evaluation, they may become proficient at using AI tools without acquiring the deeper knowledge needed to solve unfamiliar or complex problems independently.
Rather than rejecting AI, the author advocates using it as a learning partner instead of a replacement for thinking. Educational institutions, employers, and professionals should encourage practices that require users to question AI-generated responses, validate evidence, explain reasoning, and engage actively with the underlying concepts. AI has the potential to accelerate learning by offering personalized feedback, simulations, and guidance, but only if users remain intellectually engaged rather than delegating all reasoning to the technology.
The article concludes that the long-term success of the AI era will depend not only on building more capable systems but also on preserving humanity's capacity to develop expertise. Organizations and educators should design workflows that combine AI's efficiency with opportunities for deliberate practice, reflection, and independent problem-solving. If used thoughtfully, generative AI can amplify human expertise; if relied upon uncritically, it risks creating a generation of users who are highly productive but less capable of developing the deep skills that drive innovation, leadership, and scientific progress.
