Teaching What You Are Still Learning
Also known as:
The practice and ethics of sharing genuine current learning — not only polished expertise — with intellectual humility and transparency about the edges of one's knowledge.
The practice and ethics of sharing genuine current learning — not only polished expertise — with intellectual humility and transparency about the edges of one’s knowledge.
[!NOTE] Confidence Rating: ★★★ (Established) This pattern draws on Learning / Intellectual Honesty.
Section 1: Context
Multi-generational systems — families stewarding land, organizations building institutional memory, movements sustaining across decades, open-source communities, public agencies — face a peculiar fragility: the gap between what is known and what is actually transmitted. The polished expert, the seasoned official, the senior activist often teaches only the resolved answers, the proven methods. But living systems evolve faster than expertise consolidates. A forest manager trained in fire suppression faces climate patterns that render that knowledge incomplete. A product team inherits architecture decisions no one fully understands. A movement’s second generation learns tactics but not the reasoning that made them fail last time.
This creates a knowledge desert at the edges — precisely where adaptive capacity lives. The system becomes dependent on institutional memory held by individuals rather than woven into practice. When those carriers leave or fail, the system loses not just answers but the way of thinking that generated them. Teaching What You Are Still Learning addresses this by making the learning process itself the inheritance — not a finished product, but a living practice of inquiry that new participants can join, extend, and steward forward.
Section 2: Problem
The core conflict is Action vs. Reflection.
The pressure to act pulls practitioners toward closure: finish the analysis, settle on the answer, teach what you know. Reflection wants the opposite — to stay open, to name uncertainty, to examine the edges where knowledge breaks. In multi-generational contexts, this becomes acute: Do you teach the doctrine (action wins — faster, clearer, more confidence-building) or do you teach the live question (reflection wins — more honest, adaptive, but slower and messier)?
Unresolved, this tension produces systems that are either brittle or paralyzed. Brittle: practitioners inherit dogma. They apply yesterday’s answers to today’s problems, then blame the problems for not fitting. The forest burns because suppression doctrine never questioned itself. Paralyzed: in the name of humility, nothing gets shared. “I’m still learning” becomes an excuse for silence, for hoarding insight, for failing to apprentice the next generation into the actual work.
The keywords expose the core: teaching demands clarity and authority; still learning demands openness and edge-naming; ethics asks which honesty serves the system better. These three are rarely in harmony. An organization under deadline pressure teaches finished answers. An activist movement rightly distrustful of hierarchy teaches almost nothing, losing hard-won tactical knowledge. A tech team optimizes for shipping velocity, teaching only the code that works today.
Section 3: Solution
Therefore, the practitioner structures regular, bounded occasions where genuine current learning — including failure, uncertainty, and open questions — is explicitly shared with those who will steward the work next, framing these moments not as remedial or informal, but as core practice.
This pattern works by shifting the ontology of teaching. Instead of “teaching = transfer of finished knowledge,” it becomes “teaching = transmission of thinking process.” The shift is subtle but regenerative.
When you name what you’re actually learning right now — “Here’s where last month’s approach failed and I don’t yet understand why” — several things happen simultaneously. First, you create permission. The people learning from you see that the boundary between learning and doing is permeable, not a wall they must cross to earn the right to contribute. Second, you inoculate against brittleness. The next practitioner inherits not a law but a lineage: “Here’s the question people have been working on for three iterations. Here are the dead ends. Here’s where your fresh eyes might see differently.”
Third, you build resilience into the commons itself. When institutional knowledge lives only in individual expertise, the system is one retirement or departure away from starting over. But when the process of learning is visible and shared, new participants can join the inquiry at any point. They don’t repeat mistakes; they extend them. They don’t replace expertise; they apprentice into the practice that generates it.
This draws directly from intellectual honesty traditions — from the scientific method’s commitment to publishing null results, from craft apprenticeship’s “learn by watching and making mistakes alongside the master,” from indigenous knowledge transmission that embeds learning into ritual and repeated practice.
The pattern also dissolves a false choice: you need not choose between action and reflection. The teaching moment becomes the place where both happen together. You act (you’re implementing a policy, shipping a feature, running a campaign), you reflect (you notice what’s working and what’s breaking), and you make that reflection visible in real time to the people who will inherit the work.
Section 4: Implementation
For organizations (corporate & government): Establish a “Working Notes” practice: the person responsible for a major initiative (product launch, policy implementation, organizational change) maintains a shared document or monthly standup where they explicitly surface current confusion, failed experiments, and half-formed hypotheses. Not in a confidential executive summary, but in a teaching-directed way: “Here’s what we thought would work; here’s where it cracked; here’s what we’re testing next; here are three interpretations and we don’t know which is right yet.” Make this a formal deliverable, not an optional reflection. The Department of Commerce might require that every new regulation proposal include a “Learning Record” where implementers name what they expected to happen vs. what actually is, visible to the team that will maintain it in five years.
For movements (activist contexts): Structure “Debrief Circles” after major actions or campaigns. These are distinct from celebration or blame. Someone who led the initiative talks through: “We designed it this way because [reason], and it partly worked because [mechanism], and it failed in [specific way we didn’t predict], and here’s the question nobody answered.” These circles are recorded (with consent) and archived. New organizers don’t learn “here’s how we do actions”; they learn “here’s how we reasoned through the last three iterations and where we still disagree.” Movements that do this—the Movement for Black Lives organizing training, climate action networks—transmit tactical intelligence that survives across generational turnover.
For government: Build “Practice Journals” into civil service onboarding. When a senior official hands off, they don’t write a memo; they create a living document that names: the decisions that seemed obvious at the time and feel less obvious now; the stakeholders whose approval you thought mattered but didn’t; the data you wish you’d collected; the mental models that broke down. A city planning director handing off to a successor doesn’t teach “this is how zoning works”; they teach “here’s what I was wrong about, here’s what surprised me, here’s what we still don’t understand about how developers actually respond to our incentives.”
For tech/products: Implement “Architectural Journals” and “Design Decision Records” that include not just the decision but the thinking-in-progress. When you choose a tech stack, a UI pattern, or an API design, document not only the final reasoning but the alternatives you nearly chose, the risk you’re still unsure about, and the question you’d ask your successor: “We picked this database for concurrency; we still don’t know if it’s the right call for distributed writes at scale. Watch for this.” This becomes the knowledge commons your replacement learns from, not a set of rules they follow blindly.
Across all contexts: Create a teaching rhythm. Monthly or quarterly, the practitioner leads a 90-minute session with the next layer (junior staff, emerging leaders, new team members, apprentices). The structure is fixed: 20 minutes on something that worked and why, 30 minutes on something that broke and what they think happened, 20 minutes on an open question the group tries to think through together, 20 minutes for the learners to name what surprised them or pushed back against their assumptions. Make attendance non-optional for successors. Pay attention to who gets space to teach in this way — ensure it’s not only the loudest or most confident voices.
Section 5: Consequences
What flourishes:
Systems develop what might be called “thinking resilience.” When new practitioners inherit not just answers but the reasoning process, they can adapt faster when conditions change. A tech team that understands why a particular architecture was chosen can modify it; a team that just knows “this is how we do it” freezes. Relationships between generations deepen because the teaching moment becomes genuine exchange, not instruction-from-above. A senior practitioner saying “I don’t know yet” gives permission to everyone else to stay curious rather than pretend to certainty. The commons itself becomes more vital — it doesn’t just preserve knowledge; it actively renews it because each new practitioner brings fresh eyes to the open questions.
What risks emerge:
There is real risk of performative learning — where “teaching what you’re still learning” becomes a box to check, a confession without genuine uncertainty. Teams perform vulnerability while defending the actual decisions. This appears most in corporate contexts (stakeholder_architecture = 3.0) where hierarchy makes real intellectual honesty risky. A product manager says “I’m still learning” about a decision that’s already been locked in; nothing actually changes.
There is also the risk of endless reflection, where action stalls because the learning moment never closes. Activist groups can dissolve into endless debrief-and-rethink. Governments can treat “ongoing learning” as cover for avoiding hard decisions.
Most critically: this pattern has modest scores on stakeholder_architecture (3.0) and ownership (3.0). If the people who do the learning are not the ones who own the decision about what happens next, teaching becomes extraction — capturing knowledge without genuine power transfer. A junior developer learning what’s broken in the codebase has little agency to fix it. Their learning serves the system but does not strengthen their capacity to steward it.
Section 6: Known Uses
The Permaculture Movement’s “Observation Year”: The Permaculture Institute teaches design not as a set of principles to be followed, but as a thinking process you learn by doing. New permaculture designers are expected to spend a full year observing and working alongside experienced practitioners, explicitly focusing on the failures and unexpected outcomes. A designer managing a degraded hillside shares not only what worked (terracing, native plantings) but the specific ways the soil behaved differently than predicted, how neighbors responded in surprising ways, what they’d do differently. This knowledge is recorded in case studies and teaching gardens. The pattern sustains because new designers don’t inherit dogma; they inherit a way of paying attention.
Linux Kernel Development’s “Commit Message and Code Review Practice”: When Linus Torvalds or maintainers merge code, they don’t just accept or reject it. They write detailed responses that make visible their thinking: “This is clever but it creates a problem we discovered five years ago when we tried this approach; here’s that story.” The code review process is explicitly a teaching moment. New developers learn not by reading documentation but by watching how experienced maintainers reason through problems in real time. This is why newer kernel developers, even those working alone, often produce higher-quality work — they’ve inherited the process of thinking through edge cases, not just the finished product.
The Annie E. Casey Foundation’s “Evaluation as Learning Practice”: Rather than treating evaluation as a final audit, they structured their grantmaking to require that grant recipients document their evolving understanding throughout the work. A youth development organization didn’t just report results at the end; they maintained a quarterly “Learning Narrative” describing what they thought would happen, what actually happened, and what question emerged. The Foundation’s staff read these not to judge but to learn alongside the practitioners. When a grantee left, the next team inherited not a success story but a real inquiry: “Here’s where the model worked, here’s where it cracked, here’s what we’re curious about next.” This practice, across hundreds of organizations, created a commons of distributed learning that shaped the field itself.
Section 7: Cognitive Era
The emergence of AI-assisted coding, research synthesis, and decision-making transforms this pattern in two opposing ways.
On one hand, the pressure to act accelerates. When an AI system can generate code, policy options, or campaign messaging in seconds, the temptation intensifies: choose the best output, move fast, teach only what worked. The cost of genuine reflection seems higher. Tech teams shipping LLM-powered features often do exactly this — they externalize the learning to “the model learns from user feedback” and stop cultivating human reasoning in public. This is brittle. When the model fails in edge cases (and it will), the team has no inheritance of thinking to fall back on.
On the other hand, AI itself becomes a teaching partner. A practitioner using an AI system to help reason through a problem can now make that reasoning visible. “Here’s what I asked the AI; here’s why I trusted its answer; here’s where I overruled it and why.” The dialogue with AI becomes a teaching artifact. A product team using AI to generate architectural options can teach the evaluation process — how they decided which option fit this system’s constraints. A government analyst can teach “here’s what the data said, here’s what the AI suggested, here’s my judgment call, and here’s why it matters that you understand both.”
The new risk is epistemic outsourcing: when AI produces the answer, practitioners no longer ask “how do I think about this?” They ask “does the AI seem confident?” Teaching what you’re still learning becomes difficult when you’re not sure what you’re learning versus what you’re deferring to the system. The pattern survives only if teams treat AI as a thinking partner, not a thinking replacement. The “Working Notes” must be explicit about where the practitioner is still in genuine inquiry with the tool, not just applying its outputs.
Section 8: Vitality
Signs of life:
Newcomers can articulate not just what to do, but why the approach might be wrong and what they’d change. A junior developer says, “We chose this database architecture because of assumption X, but I notice the load patterns don’t quite match X anymore—should we reconsider?” That’s a sign the teaching is alive. People actively ask questions in public, and the response is not “we’ll handle that offline” but “great question, here’s how three of us think about it differently.” The debrief sessions produce genuine disagreement — not about facts, but about interpretation — and these disagreements persist as live questions, not settled doctrine. Practitioners leave and the knowledge doesn’t vanish with them; new people already understand how to ask the questions that person was holding.
Signs of decay:
Teaching becomes a performance of humility. Senior people say “I’m still learning” but their behavior signals “do what I decide.” The learning moments are scheduled but desultory — people show up, go through the motions, nothing actually changes in how work gets done. The “open questions” are never revisited; they exist in documents that nobody reads. Newcomers learn the myths but not the actual thinking. They hear “we tried that once and it failed” without understanding the conditions that made it fail, so they avoid the approach even when conditions have shifted. The fastest sign of decay: when someone says, “That’s just how we do things here,” and they can’t explain why or admit uncertainty about whether it still makes sense.
When to replant:
Replant this practice immediately after a major failure or surprise — when the inherited approach clearly broke down. This is when the system is already in a learning posture and the teaching moment has real force. Also replant when you have a generational transition: a new leader, a new team, a new context where “that’s how we’ve always done it” no longer applies. If the pattern has calcified into ritual (the forms are there but nothing genuine is shared), stop the scheduled teaching for two months and instead ask the newest practitioners: “What do you actually want to understand?” Design new teaching moments around their real inquiry. This often regenerates the practice’s authenticity.