Many organizations are seeing a quiet shift in how people learn at work. It’s not that companies are cutting training budgets. Instead, the work itself is changing.

AI tools are reshaping or removing the early-career tasks that used to be the “practice field” for developing skills. In fields like consulting, law, and finance, activities such as:

• Reviewing documents
• Building models
• Drafting analyses

were never just outputs. They were core learning mechanisms that helped people build judgment over time.

As AI takes over more of this work, research points to two types of capability erosion:

• Deskilling: Existing expertise fades because people use it less.
• Upskilling inhibition: Newer employees never get exposure to the kinds of challenges that build real judgment in the first place.

The result is that the conditions that historically built professional experience and judgment are no longer reliably present in the modern workplace. Without rethinking how people gain experience, organizations risk weaker decision-making and shallower capability across all levels, not just among junior staff.

Medicine offers a useful blueprint for reimagining how people learn at work, especially as AI becomes more embedded in daily tasks. Healthcare has been adopting AI at a pace 2.2 times faster than the broader economy, yet it continues to invest heavily in apprenticeship through medical residency.

Key elements of the residency model that translate well to corporate settings include:

• Intentional, structured experience: Professional judgment is treated as something that must be deliberately cultivated, not a byproduct of doing the job.
• Multiyear mastery programs: Training is planned over several years, with clear milestones and expectations.
• Competency-based evaluations: Progress is measured against defined competencies, not just time in role.
• Gradual progression: Trainees move from observation to autonomy in stages, with responsibilities increasing as they demonstrate readiness.
• Continuous supervision and feedback: Learners work on real cases under oversight, receiving frequent, specific feedback.
• Exposure to real-world complexity: Skills are built through hands-on work with diverse, complex situations—not just theoretical training.

For companies, the takeaway is that as AI reshapes tasks, you can’t assume judgment will “just happen.” You need to design structured pathways—much like residency—that intentionally build capability over time.

While most companies won’t copy medical residency directly, they can adapt its principles to fit corporate roles. Some practical moves include:

1. Define clear capability stages
   Map out what “early,” “intermediate,” and “advanced” judgment look like in key roles, and link them to specific skills and behaviors.
2. Design structured apprenticeship paths
   Pair less-experienced employees with more senior colleagues on real work, not just shadowing. Make the progression from observing, to assisting, to leading explicit.
3. Use AI without removing learning opportunities
   When you automate routine tasks, keep some structured exposure for juniors—e.g., having them review AI outputs, simulate the work, or handle selected cases end-to-end under supervision.
4. Build in continuous feedback
   Move beyond annual reviews. Create regular, competency-based check-ins tied to actual projects and decisions.
5. Track capability, not just performance
   Measure how people are developing judgment over time, not only whether they hit short-term targets.

By borrowing these elements from the residency model, organizations can reshape learning at work so that employees continue to build deep skills and sound judgment—even as AI takes on more of the day-to-day tasks.