What Happens to Mastery in the Age of AI?

Over the weekend, I finished reading Mastery by Robert Greene. What struck me most was not simply the discussion around expertise itself, but how many of the conditions Greene argues are necessary for mastery appear increasingly at odds with the direction of AI. Throughout the book, he repeatedly returns to the importance of long apprenticeships, deep immersion, repetition, cognitive struggle, and the gradual development of pattern recognition through years of experience. Much of mastery, in Greene’s framing, comes not from shortcuts, but from sustained exposure to complexity over time. And AI removes much of that friction.

And that raises an interesting question for international schools, and perhaps even more so for the organisations and advisors operating around them. What happens to expertise in a world increasingly optimised for instant answers?

The education conversation around AI often focuses on classroom use, assessment, academic integrity, or productivity gains. Those are important discussions, and schools clearly need to engage with them seriously. But I suspect there is a wider organisational question emerging underneath them. AI is making competent outputs dramatically easier to produce. Strategy documents can now be drafted in minutes. Market summaries generated instantly. Reports synthesised rapidly. Presentations structured automatically. Communications polished almost immediately.

At one level, this is incredibly valuable. Most organisations, including schools, are overloaded operationally, and productivity matters. There are many low-value administrative and analytical tasks where AI can create genuine efficiencies. But there is also a risk that speed begins to replace depth, particularly when organisations become focused on outputs rather than capability development.

Historically, expertise was often built through the process of wrestling with complexity. The analyst who manually worked through data developed instincts about anomalies and weak signals. The school leader who spent years handling difficult parent conversations, staffing decisions, regulatory issues, and enrolment pressures gradually developed judgement. The advisor who immersed themselves deeply in markets developed contextual understanding that could not easily be written into a framework or dashboard.

AI can often replicate the output shape of expertise. It cannot automatically replicate the underlying judgement behind it.

That distinction may become increasingly important across international education. In the data and intelligence space, this is already starting to appear. Access to information is rapidly becoming commoditised. AI can summarise markets, generate benchmarking commentary, identify trends, and synthesise reports very effectively. Increasingly, many organisations will have access to similar information and similar analytical tools.

But information was rarely the real scarcity.

Interpretation was. Knowing which data matters. Understanding what is missing. Distinguishing noise from signal. Identifying second-order implications. Recognising when a trend is structural versus temporary. Understanding the political, regulatory, or cultural context surrounding a market. Those capabilities are significantly harder to automate because they rely heavily on accumulated contextual experience and judgement.

International education is particularly complex because so much of the sector operates through nuance, relationships, reputation, and local context. Two markets can appear structurally similar on paper while functioning entirely differently operationally. Two schools with similar enrolment numbers may have completely different sustainability profiles. Two expansion strategies may look equally attractive in a spreadsheet while carrying very different execution risks in reality.

AI can describe the market. Reading the market is something else entirely.

The same tension exists within strategic advisory work. AI can already produce competent strategic plans, and many are structurally sound. Some are surprisingly sophisticated. But advisory value was never really about producing slides or frameworks. The real value sits in prioritisation, sequencing, tradeoffs, implementation realism, governance dynamics, stakeholder management, and judgement under uncertainty.

In many ways, AI may actually increase the value of high-quality advisory. If every organisation can generate strategy-shaped outputs quickly, the differentiator becomes knowing which direction is actually right, what is realistic operationally, and what unintended consequences may emerge from decisions over time. Faster outputs do not necessarily produce better decisions. In some cases, they may simply accelerate poor ones.

This may also create an interesting challenge for international schools themselves. Many schools are rightly investing heavily in AI integration and future-focused learning models. But there is perhaps a deeper question underneath the innovation narrative. How do schools continue developing deep thinkers in environments increasingly built around cognitive outsourcing? How do future school leaders develop judgement if many of the apprenticeship experiences that historically built expertise become compressed or bypassed? How do organisations avoid confusing polished outputs with genuine capability?

I do not think the answer is resisting AI. That would be unrealistic, and probably unhelpful. The strongest organisations and leaders will almost certainly be those who combine deep expertise with AI leverage. AI is clearly becoming an essential tool across education, business, research, and strategy.

But I do think the AI era may force a reconsideration of where value actually sits.

For years, access to information created advantage. Increasingly, interpretation may become the scarcer capability. And in a world filled with surface-level competence, genuine mastery may become even more valuable.

Source:

Tony Atkinson
The Weekly Education Pulse