Takeaways: Lessons learned during the session

The Illusion of Progress

     We opened the conversation with Duolingo, the language app with 500 million users where almost nobody becomes fluent. Someone mentioned Anki, a less polished competitor that actually produces language acquisition. The difference? Anki has no streaks, no gamification, minimal UX polish. Duolingo optimizes for engagement. Anki optimizes for learning. The market chose Duolingo.

     This pattern repeated across domains. Meditation apps that reward opening the app, not developing a practice. Dating apps where swiping feels productive while rarely leading to relationships. The formula is always the same: optimize for adoption metrics, not actual outcomes. These apps wouldn't survive if they optimized differently, the market rewards engagement, not results.

     Herbert Simon called this Bounded Rationality: people make optimal decisions within their constraints. When the constraint is maximize engagement, actual learning becomes irrational. When the constraint is look productive, depth becomes a liability. The tragedy isn't moral, it's structural.

The Healthcare Mirror

     While digital media's guilty optimization is well-documented, what emerged during our discussion was more unsettling, healthcare suffers from identical structural biases.

     Someone from the field described the reality, hospitals are rewarded for getting patients through as quickly as possible. Not for outcomes. Not for prevention. For throughput.

     This wasn't an individual hospital's choice. The Dutch healthcare system has GPs acting as gatekeepers, a model that evolved through the 20th century and was formalized in major reforms. By 2006, a regulated competition system emerged, but the underlying reimbursement structures remained tied to volume metrics. The same measurement framework applied universally creates the same perverse optimization everywhere.

     Even a forward-thinking hospital wanting to optimize for cure rates instead of volume won't get monetarily rewarded under the current system. The infrastructure determines what's possible.

     Here's where it gets interesting. One exception emerged: the Martini-Klinik in Hamburg, Germany. It's famous for prostate cancer surgical outcomes. They've built the world's largest database on prostate cancer surgery, tracking patient outcomes for decades. They're rewarded not for throughput, but for cure rates, continence, and quality of life metrics.

     How did they survive in systems typically optimized differently? They attracted similar people. Surgeons who cared about outcomes over volume. Administrators who supported long-term tracking over quarterly metrics. It became self-reinforcing.

     This illuminated something critical. The system doesn't just shape behavior, it selects for specific types of people. And once you have the right concentration of people, you can optimize differently within the same external constraints. But this raised an uncomfortable question: what happens in systems that select for the wrong people?

When System Selects Against Value Creation

     The healthcare observation unlocked a pattern. Someone noted that autism care and addiction treatment attract completely different clinicians. Autism services draw structured, process-oriented people. Addiction treatment draws flexible, relationship-focused ones. The systems weren't neutral containers. They were filtering mechanisms.

     This pattern showed up everywhere we looked. An American participant flagged Dutch corporate culture: endless meetings, slow decisions, building consensus even when the answer is obvious. We traced this to the Netherlands' historical development waterschappen (water boards) established in the 12th-13th centuries for collective water management.

     When flooding threatened everyone equally, coordination and buy-in mattered more than speed. These water boards became some of the oldest democratic institutions in the Netherlands, operating independently and requiring collective decision-making. The motto was simple: "Whom the water harms, stops the water." But achieving consensus across competing interests was essential or everyone would drown.

     This governance model didn't just shape Dutch politics. It shaped Dutch organizational culture. The insight struck hard: Applied to wrong environments, this doesn't just slow decisions, it filters for people who prefer process to decisiveness.

     The effect compounds over time and often times invisibly. Another participant described Amazon's two-pizza teams (Jeff Bezos' rule that teams should be small enough to feed with two pizzas, roughly 5-7 people), to maximize autonomy and minimize coordination costs. They looked autonomous on paper. But the people who are usually attracted by the idea want ownership, not freedom within guardrails. The structure often promises one thing but delivers another.

     Niklas Luhmann's Organizational theory applies: organizations reduce complexity through categorization. But those categories don't just organize work, they organize who is attracted to that work, and it is of a prime importance to ensure that we categorize cautiously.

Performers Beat Craftsmen when Visibility is Currency

     The most uncomfortable realization emerged gradually. In most modern organizations, people who deeply understand their craft get passed over, while people who perform expertise advance.

     This isn't limited to any single domain. Writers who chase visibility over quality. Clinicians who optimize for legible metrics over patient outcomes. Leaders who master presentation over decision-making. When LinkedIn endorsements matter more than deployed work, people optimize for endorsements. When quarterly decks matter more than long-term outcomes, people optimize for decks.

We shape our tools, and thereafter our tools shape us. Marshall McLuhan

     The conversation turned philosophical. How do craftsmen survive in these systems? The answer felt bleak at first: they don't aim for the reward.

     But this reframed everything. The question isn't how to succeed in broken systems. It's how to design systems where craft and reward align. Where doing excellent work and being recognized for it become the same thing.

     Several participants mentioned seeking out environments that value depth over performance. Non-profits were raised, then quickly dismissed: people there are looking for meaning, but the systems are often just as dysfunctional. Low salaries, poor structure, burnout. The feeling of doing good substitutes for actually doing good.

     Then someone articulated what many were thinking: Many people would trade high salaries for genuine community and meaningful work. Truth being told, there are no straight-forward recipes, but only through cautious evaluation by stating our core objectives and acknowledging given constrains (both from the perspective of institutional rewards and human nature) can we build a space which optimizes for the long-term value.

Conclusion

     The question was: How do we create value when incentives pull us toward shortcuts? The answer: Design systems where doing the right thing is the obvious thing.

The question isn't how to succeed in broken systems. It's how to design systems where craft and reward align, where doing excellent work and being recognized for it become the same thing.

     You can't do this by fighting existing systems. You build subsystems (teams, projects, companies) where incentives align. The Martini-Klinik didn't convince German healthcare to change. They created a microenvironment where outcomes mattered and attracted people who cared.

     This is what system architecture actually means, not optimizing broken constraints or exiting the system, but finding the path within the given constraints.

The Power of Personal Examples as Catalysts

     A double realization crystallized during this session: personal, concrete examples serve as the most reliable entry points to genuinely new ideas. When someone shares a specific moment (a frustrating meeting, a surprising success, a moment of clarity), the group doesn't just listen; it begins to see patterns that abstract discussion alone rarely uncovers.

     Encouraging participants to reflect on their own experience proved extraordinarily generative. Each story functioned like a probe dropped into unknown territory: it revealed hidden structures, unexpected parallels, and counterintuitive mechanisms that the theory by itself could not reach. The more specific and honest the example, the richer the subsequent analysis became.

     Yet the second half of the insight was equally important: the conversation must remain open enough for people to ask clarifying questions of one another. Without that permission and habit, without genuine curiosity directed toward the person who just shared, personal stories risk becoming isolated anecdotes rather than shared data points for collective sense-making.

     The most productive moments occurred when someone paused and asked: "What exactly did that feel like?" or "What did the other person say right after?" or "Have you seen that pattern show up somewhere else?" These small interventions turned individual reflection into group discovery.

Personal examples unlock new territory. Curiosity and open questioning turn those examples into shared theory. When both are present, reflection becomes rocket fuel for insight.