# The future of learning at work: How the medical residency model offers an inspiring solution > Source: > Published: 2026-08-12 11:05:00+00:00 Today’s corporate talent faces a quiet but meaningful disruption. Many of the pathways for professionals to build their skills and capabilities are disappearing. Not because companies are investing less in training, but because the work itself is rapidly changing. The early tasks that once allowed people to “learn by doing” are being reshaped or removed by [AI](https://www.fastcompany.com/section/artificial-intelligence) tools. At first glance, the shift seems manageable. AI takes over routine work, freeing people to focus on higher-value activities. But that logic breaks down over time. The same work being automated today is the work that previously served as a critical training ground and built foundational capability in fields like consulting, law, and finance. Reviewing documents, building models, and drafting analyses were not just outputs; they were learning mechanisms. Without them, organizations face the erosion of human capability and judgment across all tenures. Research reveals that as AI becomes more embedded in daily work, [capability erodes](https://link.springer.com/article/10.1007/s10462-025-11352-1) in two ways: through *deskilling*, when existing expertise fades from lack of use, and through *upskilling inhibition*, when employees never gain exposure to the kinds of challenges that build judgment in the first place. Simply said, the conditions that historically built professional experience and judgment are no longer reliably found in the modern workplace. To understand what happens next, it helps to look to a sector that has confronted this challenge head-on while adopting AI at a rate [2.2 times as fast as the broader economy](https://hitconsultant.net/2025/10/22/healthcare-ai-adoption-is-2-2x-faster-than-the-broader-economy/): medicine. Medicine has long recognized that professional judgment does not develop on its own. It is not simply a byproduct of work. It is the result of intentional and structured experience. Judgement is [deliberately](https://onlinelibrary.wiley.com/doi/10.1002/9781119839446.ch28?msockid=3c2aced83b23695d15b7d99c3af96870) cultivated through apprenticeship with multiyear mastery programs and continuous, comprehensive competency-based evaluations. Trainees move gradually from observation to autonomy. They work on diverse cases under continuous supervision and with frequent feedback. They develop competence not through abstraction, but through exposure to real-world complexity. Responsibility progresses from low-stakes supervised decisions to independent ownership with clear standards of excellence at each stage. What is striking is not just that medicine invests so heavily in apprenticeship and the residency model, but that it has continued to do so even as technology has radically advanced. From imaging systems to AI diagnostics, medicine has absorbed new tools without abandoning the human learning processes that underpin good decision-making. This is because no matter how sophisticated the tools become, they do not replace the need to interpret, prioritize, and make decisions under uncertainty. AI systems can now summarize patient records and surface patterns that would have taken hours for a doctor to assemble manually. Yet the activities of reviewing charts, synthesizing information, and identifying what is relevant have historically been essential to building clinical judgment. When those steps are bypassed, the efficiency gain is real but so is the loss of embedded learning. The response within medicine has not been to slow down technological adoption, but to rethink how learning happens alongside it. One of the most important developments has been the increased emphasis on [cognitive apprenticeship](https://meded.ucsf.edu/sites/meded.ucsf.edu/files/inline-files/pearls-cognitive-apprenticeship.pdf), which focuses not just on what experts do, but on how they think. This approach makes reasoning visible. It encourages experienced practitioners to articulate their decision-making processes, explain how they weigh evidence, and show how they navigate ambiguity in real time. This matters more in an AI-enabled environment than previously. When outputs are instantly available, the differentiator is no longer the ability to produce them but the ability to interpret them correctly. That requires a deeper form of learning, one that goes beyond task execution and into judgment formation. In medical education today, AI is being integrated in ways that reinforce, but do not replace, the apprenticeship and medical training approach. AI tools now allow learners to test their reasoning in real time, receive immediate feedback, and access large bodies of knowledge without leaving the workflow. Medical professionals are also delving into [adaptive learning platforms, AI-powered simulations, automated assessments, personalized coaching, and chatbot-assisted learning](https://pmc.ncbi.nlm.nih.gov/articles/PMC12007958/). The most important part of expertise has always been invisible. In knowledge work, value is created not only through execution, but through judgment in terms of how problems are framed, how trade-offs are evaluated, and how decisions are made under uncertainty. These are not skills that can be learned passively. They must be brought to the surface. As with medicine, reframing apprenticeship for modern work means experts must be encouraged to explain how they reason. Learners must be asked not just to produce answers, but to articulate how they arrived at them. This is where many organizations fall short. They review outputs, but rarely expose the reasoning behind them, or they evaluate results, but do little to develop the underlying decision-making capability. While medical professionals are required to teach and are often driven by altruism and a desire to “pay it forward,” organizations in other fields must establish clear apprenticeship objectives, set accountability metrics, and incentivize senior staff to share their skills and experience to create their own successful “residency” programs. Incentives could include personal recognition, career growth, or professional development opportunities. Meanwhile, to ensure executive leaders support and fully participate in apprenticeship, apprenticeship programs can be integrated into executive job descriptions and annual reviews for leader evaluation, promotion, and remuneration. Today, executives are challenged in ensuring their organizations remain capable of developing expertise in an AI-enabled environment. However, a road map is taking shape. It includes recognizing that some forms of inefficiency—repetition, iteration, and even rework—are not waste but rather how capability is built. Removing inefficiencies entirely may improve short-term [productivity](https://www.fastcompany.com/section/productivity) but create long-term fragility. The road map encourages organizations to be intentional about where AI is introduced into workflows and its broader potential impacts. In some cases, AI should accelerate work. In others, it may need to be deliberately limited with guardrails that preserve learning. The goal is not maximum automation. It is growing and sustaining capability, as well as value creation. And the road map entails rethinking early-career roles. If entry-level tasks are automated, those roles cannot simply shrink. They must be redesigned around interpretation, oversight, and decision-making, with structured opportunities to build judgment rather than just execute tasks. There are signs that organizations are beginning to respond to these new AI realities and develop initial best practices. They include: These approaches are all connected by one important thread: a recognition that learning cannot rely solely on exposure to outputs. As with the COVID-19 pandemic, developing expertise in the Age of AI is a novel and complex organizational challenge that business has yet to fully understand or solve. For companies, finding the right balance among inefficiencies, creating appropriate workflows, and redesigning leadership structures to help build employee capabilities will be difficult in the earlier stages of AI’s implementation. In addition, many organizations today lack the internal skill sets and knowledge base to drive this complex transformation. These companies may benefit from the support and unique perspectives of external experts such as neuroscientists, behavioral economists, and medical directors as they embark on this journey and consider key questions that will need to be addressed, including: For years, companies have developed talent without necessarily constructing or holistically funding the end-to-end professional learning journey. Many informal experiences and costs have been hidden in early-tenure inefficiency, embedded in workflows, and absorbed over time. But AI has rapidly changed this. Organizations now need to design and fund these pathways more intentionally, drawing inspiration from medicine’s approach. As leading companies explore and embrace new best practices and integrate policies, processes, and systems that make them more efficient and competitive, initiating an AI-era learning and skills development strategy is quickly moving from want to need, from innovation to table stakes. Just as the invention of the stethoscope shifted how clinicians evolved from anatomists to diagnosticians, AI represents an inflection point that changes the way work gets done and how expertise is developed. The companies that respond to these challenges early and strategically will maximize AI’s return on investment and be more productive, more resilient, and better able to attract and retain talent. For the long term.