The Illusion of Mastery
How AI and the Relentless Pursuit of Efficiency Are Creating a Generation of Hollow Product Managers
Somewhere along the way, in our relentless pursuit of efficiency, we forgot that real mastery isn’t about getting things done quickly; it’s about knowing why and how to do them in the first place. The tech industry, in its obsession with streamlining everything, is at risk of turning product managers into fast-moving but shallow executors, rather than strategic thinkers who drive meaningful innovation. When organizations flatten hierarchies and outsource key decision-making to AI, they create PMs who look like they know what they’re doing but lack the deep, nuanced understanding required for long-term success. Perhaps this should not be a surprise; as Clayton Christensen pointed out, companies that focus too much on short-term efficiency often sacrifice their ability to innovate in ways that truly disrupt the market.
But how do PMs actually develop expertise? It turns out, there are (at least) two compelling ways to think about this journey. The first, Shuhari, comes from martial arts and describes the stages of individual mastery. The second, Lev Vygotsky’s Zone of Proximal Development (ZPD), focuses on how learning thrives in a social, collaborative environment. Both offer powerful insights into why today’s AI- and efficiency-driven product culture is falling short.
Shuhari: The Road to Mastery
In Japanese martial arts, Shuhari describes three stages of skill development, which, when applied to product management, provide context for the journey from following established frameworks to innovating beyond them. Although it may read like a linear path, in truth it is a transformation of thought, a continuous rewiring of how one approaches problems, decision-making, and leadership.
Shu (守, “Protect”) – Learning the Rules
The Shu phase is where discipline meets repetition. It’s about absorbing the fundamentals, following best practices, and executing with precision. In martial arts, this means practicing the same kata (form) over and over until muscle memory takes hold. In product management, it means mastering Agile ceremonies, roadmapping, stakeholder communication, and feature prioritization.
At this stage, PMs operate within predefined structures. They take Jira tickets at face value, run retrospectives by the book, and rely on standard OKRs to measure success. Their decision-making is primarily rule- or framework-based. If a backlog needs prioritization, they might instinctively apply a RICE or MoSCoW framework without questioning whether those methods are the right fit for their specific context.
The key challenge here is that many PMs (and organizations) mistake competence at this level for true expertise. If a PM is highly efficient in executing tasks, it can create the illusion of mastery when, in reality, they are still heavily reliant on external guidance and frameworks. This often reveals itself in the much-bemoaned trend of PMs leaving big tech, and its cushy tools, processes and aligned values, for startups, where they subsequently flounder without the prior support. AI-driven tools promise to exacerbate this further by making execution easier, reinforcing a cycle where PMs remain stuck in the mechanics of the role rather than engaging in deeper strategic thinking.
Ha (破, “Break”) – Questioning and Adapting
The Ha phase is where true learning begins. Here, the PM starts to see the cracks in rigid methodologies and begins adapting frameworks to fit their team’s needs. Instead of blindly applying Agile, they might modify sprint cadences based on engineering cycles or merge Lean and Scrum elements to better suit their product’s evolution. They start seeing patterns across projects and drawing on their own judgment rather than relying solely on industry “best practices.”
At this level, decision-making becomes less about strict adherence to frameworks and more about evaluating trade-offs. A Ha-level PM no longer asks, “What does the framework say we should do?” but instead asks, “What are we really trying to solve, and what’s the best approach given our constraints?”
However, this phase is also where insecurity can creep in. Many PMs, especially in high-pressure environments, find themselves second-guessing their instincts. The absence of a clear “right answer” can be unsettling, leading some to retreat to the safety of rigid execution rather than embracing the discomfort of ambiguity. Organizations that prize efficiency over learning tend to punish this kind of exploration, reinforcing a culture where PMs are encouraged to stay at the Shu level indefinitely.
Ri (離, “Transcend”) – Mastery and Innovation
Reaching Ri is like Neo seeing the Matrix. It’s the moment where rules no longer define your thinking, but rather serve as reference points in a much broader landscape of possibility. A Ri-level PM doesn’t just use frameworks; they create new ones. They don’t just manage roadmaps; they redefine how roadmapping is done.
At this stage, intuition plays a dominant role. But this intuition isn’t magic, rather it’s built on years of deep learning, pattern recognition, and experience. A Ri PM can navigate ambiguity not because they have a perfect answer, but because they trust their ability to make sense of uncertainty. They know when to abandon a standard playbook and when to double down on discipline.
Interestingly, this is also the point where PMs become most resistant to the very systems that helped them learn. They question traditional ways of doing things, challenge company-wide strategies, and push organizations to rethink their approach to product development. But in highly structured companies that prioritize efficiency and execution over long-term vision, Ri-level PMs can be seen as disruptive rather than valuable. When AI is used to automate decision-making and flatten organizations, there’s little room for a PM who operates beyond predefined and predictable systems, leading to a stagnation of innovation at the highest levels.
If Shuhari has a fatal flaw in application to product management, it is that it treats learning, by nature, as an individualistic journey. It assumes mastery is an internal process, something one achieves through personal discipline and experience. But modern product management doesn’t exist in a vacuum — it’s a team sport. And that’s where ZPD comes in.
Zone of Proximal Development: The Power of Learning Together
Vygotsky’s Zone of Proximal Development (ZPD) turns the concept of mastery on its head. Instead of seeing expertise as something an individual achieves, it views it as something that emerges through collaboration and guided learning.
The core idea is simple: the best learning happens in the gap between what someone can do alone and what they can accomplish with guidance from a more knowledgeable peer. This approach is especially effective because the peer does not need to be an expert in everything, only in a specific area that helps the learner progress. The key is to be consistently challenged just beyond one's current abilities while receiving the right support to close the gap.
Why ZPD Matters for Product Management
Traditional corporate learning assumes that people move from junior to senior roles in a fairly linear fashion, gaining expertise through experience. But real growth rarely happens in isolation. PMs don’t suddenly “level up” because they’ve been in the job long enough; they grow by working alongside more experienced colleagues, tackling increasingly complex problems with guidance.
Research on “cognitive apprenticeship” highlights that learning embedded in real-world practice and reinforced through social collaboration is significantly more effective than isolated study. In the context of product management, this approach translates to:
Engaging directly with senior leaders to refine product strategies, allowing for the exchange of insights and experiential knowledge.
Participating in live discussions where decisions are debated in real-time, fostering critical thinking and adaptability.
Receiving hands-on mentorship tailored to specific challenges, as opposed to generic career coaching, to address immediate learning needs.
Yet, as organizations flatten their hierarchies and remove layers of management, these learning opportunities disappear. PMs are expected to “own” their work independently, which sounds empowering but often leaves them without the support structures they need to actually develop expertise.
The Risk of Isolated Learning
Without mentorship, PMs are left to figure things out on their own. Some might manage to navigate the Shuhari journey independently, but many will stagnate, stuck in the Shu phase because they lack the right challenges and feedback loops to push them forward.
And this is where AI creates a dangerous paradox. By automating analysis, prioritization, and even decision-making, AI removes the need for PMs to engage in deep problem-solving, effectively cutting off access to the very struggles that drive growth. If a PM never has to wrestle with ambiguous data, defend a controversial product decision, or navigate cross-functional conflicts, they never develop the instincts needed to operate at the Ri level.
The problem isn’t the accuracy of applying Shuhari or ZPD to the PM’s learning journey; it’s that modern organizations are stripping away the very conditions that allow any learning model to function. On a fundamental level, mastery requires both individual discipline (Shuhari) and social learning (ZPD), but AI-driven efficiency models prioritize neither.
If companies want PMs who are more than just execution engines, they need to reinvest in mentorship, structured learning, and environments where people actively challenge each other’s thinking. Without these, we’ll be left with a generation of PMs who can run a sprint perfectly but have no idea why their product is failing; or worse, no idea how to even ask the right questions.
AI and “Flat” Organizations Are Making PMs Shallow
Right now, the tech industry is making two big mistakes:
Killing mentorship – Flat organizations might sound great in theory (less bureaucracy! more ownership!), but they also eliminate natural mentorship structures. Without strong mentors pushing them beyond their comfort zones, PMs struggle to reach true mastery. As Peter Drucker pointed out, great leadership isn’t about removing hierarchy, but rather ensuring that knowledge and expertise are continuously passed down.
Over-reliance on AI – AI tools are great at surfacing insights, but overusing them creates “shallow experts”: people who can read the AI’s output but don’t really understand the underlying dynamics. If you don’t understand why something works, you won’t be able to solve problems when the AI inevitably fails.
What Can We Do About It?
If we want to reverse this trend and create PMs who are more than just well-dressed API endpoints, we need to rethink how they grow:
Bring back structured mentorship. Implement formal structures where senior PMs actively guide junior PMs through complex challenges, helping them move beyond basic execution.
Foster collaborative learning. Encourage learning environments where PMs constantly engage with new perspectives, refining their thinking through debate and shared experience.
Develop AI literacy - and skepticism. Train PMs to critically evaluate AI insights, ensuring they remain decision-makers rather than passive executors.
Create space for experimentation. Give PMs opportunities to work on high-stakes, ambiguous projects that force them to develop independent judgment. True innovation comes from organizations that deliberately cultivate risk-taking and deep thinking.
Mastery isn’t about following steps faster. It’s about understanding the game deeply enough to change it. If we keep prioritizing efficiency over expertise, we won’t just lose great PMs, we’ll lose the ability to build truly innovative products. And if that happens, AI won’t need to take over. We’ll have already automated ourselves into irrelevance.
