If learning is infrastructure, it needs to be designed with the same seriousness as any other operational system. It must be accessible, adaptable and able to serve different users at different moments, often simultaneously.

The modern learning environment is distributed through a network of settings that supports everything from deep concentration to mentoring, from structured teaching to spontaneous knowledge exchange. Increasingly, the critical spaces are not the formal classrooms but the third places: the in-between environments where learning becomes self-directed, unscheduled and often more memorable.

In practice, learning infrastructure looks like spatial layering with neighbourhoods and hubs, quiet libraries and project rooms, writable walls and digital labs, social staircases and town hall spaces, touchdown zones where a five-minute peer exchange is as valuable as a formal session.

Architecturally, this means designing workplaces less as collections of fixed departments and more as campuses in miniature. Circulation becomes more than movement but places of exchange where atriums, stairs and shared thresholds become opportunities for visibility and connection.

AI is the third accelerating driver of design and as tools become more intelligent, learning becomes more frequent and more personalised, but also more fragmented. People can access information instantly, but that information then has to be competently translated, absorbed and integrated.

This is the paradox we see repeatedly: as education becomes more digitally delivered and dispersed, the physical learning space becomes more, not less, important. It provides presence and shared context, crucial ingredients for collaboration and the kind of tacit learning that happens by watching how others work.

Design must respond by shifting focus from content delivery to practice environments. The most valuable learning spaces in an AI-enabled workplace are often those that support rehearsal, simulation, prototyping and demonstration. They are spaces where people are consuming knowledge and then able to test it, apply it and build confidence through shared experience.

Our work for Medtronic illustrates this clearly with a training centre designed to integrate high-fidelity simulation with social and collaborative space, supporting thousands of professionals annually. The building’s central atrium acts as a spatial connector, encouraging interaction and exchange, while adaptable teaching environments support evolving technologies and methods.

A similar philosophy applies in corporate settings. At Google Ananta in Bengaluru and EY’s Lisbon headquarters, flexible neighbourhoods, shared amenities, collaboration spaces and quieter settings allow people to select the right environment not just for the task, but for the learning moment within it.