Thinking like a gardener not a builder, organizing teams like slime mold, the adjacent possible, and other unconventional product advice | Alex Komoroske (Stripe, Google)
This episode features Alex Komorosky, a former Google and Stripe leader, discussing how LLMs and Generative AI will reshape product development. He advocates for a "gardener" mindset, designing for emergence, and leveraging AI as an "electric bike for ideas," alongside practical advice on productivity, organizational dynamics like "kayfabe," and cultivating "nerd clubs."
Deep Dive Analysis
15 Topic Outline
Introduction to Alex Komoroske and His Bits and Bobs Document
Alex's Compendium and Personal Note-Taking Process
Impact of AI and LLMs on Product Development
Essential Skills for Product Managers in the AI Era
How AI Enhances Individual Productivity
Understanding Organizational Kayfabe and its Dynamics
The Gardening vs. Building Mindset in Product Development
Designing for Emergence and Bottom-Up Innovation
Slime Molds as an Organizational Structure Metaphor
Identifying and Cultivating Resonant Ideas and Metaphors
Creating and Sustaining Strategy Salons (Nerd Clubs)
Productivity Hacks and Self-Control Strategies
Life Philosophy: Energy, Pride, and the Hallmark Card Fallacy
Navigating Strategy with the Adjacent Possible
Reflections on Aggregators and the Future of Software
10 Key Concepts
Bits and Bobs
This is Alex Komoroske's personal, 600-page Google Doc where he collects and reflects on ideas, finding patterns and meaning from his notes. It's a tool for self-reflection, not designed for easy consumption by others.
Magical Duct Tape (LLMs)
This metaphor describes Large Language Models (LLMs) as a disruptive technology that can make writing 'shitty' software cheaper and run certain software cheaply, operating at a cost structure between a human and plain computing. However, LLMs are 'squishy' and not always fully accurate, requiring product design to account for their imprecision.
Taste (in AI Era)
In an era of abundant, often 'slop' information production by AI, 'taste' refers to having a distinctive perspective that stands out from the background noise and is compelling to others. It's about being different from what an LLM would generate and finding what resonates uniquely with people.
Organizational Kayfabe
Derived from professional wrestling, this concept describes a phenomenon in organizations where everyone knows something is fake (e.g., an overly optimistic status update) but acts as if it's real. This can compound through layers of management, leading to significant deviations from ground truth and potentially bad decisions, as disconfirming evidence is suppressed to maintain perceived stability.
Gardening vs. Building Mindset
The 'builder' mindset involves having a plan and manipulating things to match it, creating value proportional to effort. The 'gardener' mindset, by contrast, involves cultivating things that can grow on their own, directing energy, and curating, looking for compounding potential that can yield disproportionate value and appear magical.
Emergence-Oriented Thinking
This approach focuses on creating opportunities for bottom-up growth and unexpected solutions by planting small, cheap 'seeds' (ideas or projects) and responding to those that show viability, rather than imposing top-down control. It embraces the idea that great things come from high-trust environments where people lean into their unique strengths.
Slime Molds (Organizational Analogy)
This analogy suggests that organizations, especially those emphasizing employee autonomy, behave more like slime molds, finding solutions to complex problems without central control. Fighting this natural tendency leads to dysfunction, while embracing it allows for powerful, anti-fragile systems that can discover solutions not initially sought.
Always Rules vs. Sometimes Rules
For self-control and habit formation, 'always rules' (clear, black-and-white rules without exceptions) are more effective than 'sometimes rules.' Always rules reduce decision fatigue and help maintain streaks, making it easier to stick to desired behaviors.
Hallmark Card Fallacy
This fallacy describes the experience of discovering a deep, personal insight that profoundly changes one's perspective, only to realize it's a common cliché (like 'the friends we made along the way'). The insight feels profound because one was finally ready to truly understand and internalize its meaning, not because it's new information.
Adjacent Possible
This concept refers to the set of small, safe, and reasonable actions that are immediately within reach. By taking these incremental steps, the world reconfigures, opening up new possibilities and allowing one to arc towards wildly different, impactful outcomes without taking large, risky leaps into the unknown.
13 Questions Answered
It's a 600-page personal Google Doc where Alex collects notes and reflects on ideas daily, distilling them into insights. It serves as a tool for his own self-reflection rather than a polished publication for others.
He uses an open-source tool called 'The Compendium' to quickly capture notes during meetings, process them daily with added context and similar idea embeddings, and then on weekends, he distills resonant notes into a long-form Google Doc for weekly publication.
AI and LLMs are disruptive technologies that change the cost structure of software, making it cheaper to write and run certain types of software. This shifts the basis of competition, requiring product builders to design for 'squishy' and not fully accurate outputs, and to focus on 'AI-native' problems rather than just slapping AI onto existing playbooks.
Product managers should lean into curiosity and play, experimenting with new possibilities enabled by LLMs. The most important skill will be 'taste' – having a distinctive perspective that stands out from AI-generated 'slop' and resonates compellingly with others.
AI, particularly LLMs, makes individuals significantly more productive by acting as a 'magical duct tape' or an 'electric bike for idea spaces,' allowing them to think through problems faster and cover more ground. This individual empowerment often happens 'below the level of awareness' of the organization.
Organizational kayfabe is when everyone in a company pretends something is real even though they know it's fake, often stemming from a desire to appear optimistic or avoid negative scrutiny. This can lead to a compounding effect where ground truth is obscured across management layers, resulting in bad decisions and a 'zombie' state for large organizations.
The gardening mindset involves cultivating things that can grow on their own, directing energy, and curating, rather than strictly following a plan like a 'builder.' It focuses on planting small 'seeds' with compounding potential and responding to signals of usefulness, allowing for growth that appears magical.
Leaders can foster emergence by dedicating a portion of team effort (e.g., 30%) to planting small, experimental 'seeds' that might seem silly but have potential. They should create a high-trust environment, provide 'cover fire' by ensuring the team delivers on core value (70% effort), and protect these nascent ideas from being prematurely 'dug up' or stopped.
Organizations, especially large tech companies that prioritize employee autonomy, often behave more like slime molds than traditional hierarchical structures. They can find solutions to problems without central control, and leaders should embrace this emergent, decentralized nature rather than fighting it, recognizing that it can lead to powerful and anti-fragile systems.
To identify resonant ideas, one should test frames or concepts by talking to diverse people. If an idea or metaphor resonates with individuals from different backgrounds, skill sets, and social subgraphs, it implies a much larger potential audience and indicates a compelling insight.
Effective hacks include playing yourself like a 'fiddle' by structuring your day to leverage activation energy and bursts of energy from completing tasks, using 'always rules' for self-control (clear, black-and-white rules), and employing the 'Seinfeld trick' of tracking streaks to build momentum.
The adjacent possible refers to the small, safe, and reasonable actions immediately within reach. It informs strategy by guiding incremental steps that, when taken, reconfigure the world and open new possibilities, allowing one to arc towards a low-resolution 'North Star' without needing to predict the entire path upfront.
It requires both: a low-resolution 'North Star' (3-5 years out, plausible, and inspiring) to provide coherence, and then taking incremental steps from the 'adjacent possible' that have the steepest gradient towards that North Star. This avoids random walking while also preventing the creation of impossible 'castles in the sky' by not trying to jump too far ahead.
21 Actionable Insights
1. Adopt a Gardener Mindset
Approach problems and product building with a “gardener” mindset, focusing on cultivating things that can grow on their own, rather than a “builder” mindset of rigid plans. This allows you to “farm for miracles” by directing and curating efforts, potentially creating more value than the effort invested, and responding to what naturally thrives.
2. Design for Emergence
Create opportunities for bottom-up emergence within organizations and product development, rather than relying solely on top-down control. Emergence is a powerful force that can lead to solutions you didn’t even know you were searching for, and allows for greater impact by leveraging collective agency and unexpected growth.
3. Cultivate Curiosity and Play with AI
In the early stages of AI technology, prioritize curiosity and play by actively experimenting with LLMs and Gen AI tools. This helps discover “odd, interesting, weird, provocative, generative things” and new patterns, as existing playbooks may be outdated due to changed cost structures.
4. Develop Distinctive Taste
Focus on developing “good taste,” meaning a unique and compelling perspective that differentiates your work from the average or AI-generated “slop.” In an era of reduced information production costs, having a distinctive taste is crucial for standing out and resonating with others.
5. Leverage LLMs as an Electric Bike
Use LLMs (like Claude) as a personal thought partner to explore problem domains, name concepts, generate examples, and critique ideas. This allows you to cover vast “idea spaces” much more quickly and without the social or financial cost of engaging human experts for early-stage exploration.
6. Create Space for Deep Reflection
Intentionally carve out dedicated time and space for deep thinking and reflection, such as weekly note processing or a meeting-free day. This practice is essential for distilling insights, finding patterns, and identifying high-impact actions that can save significant future effort and lead to “oh my God” moments.
7. Document Repeated Learnings
If you find yourself repeatedly explaining the same strategic concept or change to multiple people, memorialize it in a document. This saves time by allowing others to access the information independently, and ensures valuable insights are preserved for future reference.
8. Understand Organizational Kayfabe
Recognize that organizations often operate with “kayfabe,” where everyone knows something is fake (e.g., overly optimistic status reports) but acts like it’s real. Acknowledging this emergent dynamic helps you navigate systemic pressures, make grounded decisions, and understand why direct challenges to the false narrative can be dangerous.
9. Cultivate “Nerd Clubs”
Initiate or join small, optional, and “secret” groups with explicit norms for collaborative, “yes, and” debate, and a focus on intrinsic enjoyment. These low-stakes environments foster open exploration of half-formed ideas, leverage diverse perspectives, and stochastically generate game-changing insights.
10. Build Community Momentum
When starting a “nerd club,” begin with a small, actively engaged core, incrementally add diverse members, actively prune unproductive interactions, and encourage sharing with positive reinforcement. This approach ensures the community remains generative, self-sustaining, and a safe space for experimentation, leading to valuable insights and growth.
11. Implement “Always Rules” for Habits
For self-control and habit formation, establish “always rules” (clear, black-and-white, sustainable actions) rather than “sometimes rules.” Always rules are easier to adhere to, prevent breaking streaks, and reduce decision fatigue, making it simpler to maintain productive behaviors like daily workouts.
12. Play Yourself Like a Fiddle
Understand your personal energy levels and work patterns, then structure your day to leverage them, such as tackling small, easy tasks for energy boosts or creating artificial time constraints. This maximizes your flow state, makes daunting tasks more manageable, and helps you become “wildly more productive” by aligning with your natural working style.
13. Capture Ideas Immediately
When an interesting idea or “muse hits,” immediately capture it, even if it’s just a rough 30-minute draft. This “first act of creation” is crucial for preserving insights, as delaying capture often leads to the idea being lost or forgotten.
14. Do What Energizes and Makes You Proud
Prioritize activities that intrinsically give you energy and that you would genuinely be proud to reflect on in the long term. This alignment with personal values provides “infinite energy,” makes you indefatigable, and helps ensure your work is meaningful and prevents becoming a “husk of yourself.”
15. Embrace the Adjacent Possible
When planning strategy or product roadmaps, focus on the “adjacent possible”—small, almost certainly workable actions directly within reach. This approach reduces risk, allows for full agency in decision-making, and enables you to “arc to wildly different outcomes” through a series of safe, reasonable steps, rather than risky “flying leaps.”
16. Combine North Star with Incremental Steps
Define a low-resolution, plausible “North Star” 3-5 years out that inspires universal high-fives, then take incremental steps from the “adjacent possible” that have the steepest gradient towards it. This approach balances long-term vision with actionable, low-risk steps, preventing aimless “random walking” (only incremental) and unachievable “castles in the sky” (only long-term).
17. Challenge False Precision
Be skeptical of demands for “false precision” in strategic planning or forecasting, especially in uncertain environments (e.g., exact numbers for distant future outcomes). Such demands often serve as a “comfort blanket” to mask uncertainty at great expense, distracting from the true “orders of magnitude” impact of strong ideas.
18. Hold Expectations Lightly
Manage your expectations by holding them lightly and being willing to adjust them based on reality. Happiness can be viewed as “reality minus expectations”; since reality is hard to change, adjusting expectations is an easier path to contentment and reduces disappointment when things don’t go perfectly.
19. See Future Humor Now
When facing frustrating or difficult situations, try to envision how the event will be a funny story in 10 years and find the humor in it now. This perspective helps to reframe immediate challenges, reduce current stress, and cultivate a more resilient mindset.
20. Find Seeds of Greatness in Others
Assume everyone you interact with possesses “seeds of greatness” and approach them with respect, seeking to “steel-man” their most compelling ideas. This fosters a compassionate and generative environment, maximizes direct and indirect value creation, and encourages others to stretch their agency and be more receptive to feedback.
21. Embrace Awe and Curiosity
Actively cultivate and embrace feelings of awe, wonder, and curiosity in your life and work. These emotions are crucial for being open to disconfirming evidence, seeing beyond current limitations, and preventing your ego from constructing a world that avoids challenging information, ultimately leading to personal and systemic strength.
8 Key Quotes
So much of the way that we tackle problems and build products is this builder mindset. It's like, I have a plan, I then manipulate things to match my plan and make it happen. And this is a way you can create tons of value. Part of the problem though is it can't possibly create more value than the effort that you put into it.
Alex Komoroske
I think LLMs are truly a disruptive technology. In fact, I would argue that what we're seeing in the industry is us trying to use mature playbooks from the end stage of the last tech era in one that doesn't really fit yet.
Alex Komoroske
If you are very powerful, you will find all the confirming evidence you need. And if it doesn't exist, it will be created for you without your knowledge.
Alex Komoroske
A community with zero people speaking is dead and a community with one person speaking is, doesn't yet realize it's dead.
Alex Komoroske
Do things that give you energy that you are proud of.
Alex Komoroske
The secret to life is things you've heard a million times already. You just weren't ready to hear them.
Alex Komoroske
Happiness is reality minus expectations.
Alex Komoroske (attributed to Tim Urban)
To me, software is alchemy is the ability to extend human agency beyond ourselves to create something that can then combine with what others have created in unexpected and unforeseen ways to create the commentatorial possibility of human agency. And somehow in the past decade, we've come, become convinced that all of this potential should be squeezed into about a dozen little boxes on your phone.
Alex Komoroske
2 Protocols
Alex Komoroske's Note-Taking and Reflection Process
Alex Komoroske- Take notes very quickly in meetings, often on a phone or by typing, indicating interest in what is being said.
- Every day or two, process these raw notes by putting them into the Compendium (an open-source tool) as 'working notes,' correcting misspellings and adding context for future understanding.
- Utilize the Compendium's embedding feature to find similar ideas from past notes, connecting new thoughts to existing ones.
- On Friday afternoons, review all notes added that week and select (check) those that still resonate or seem interesting.
- During the weekends, distill the selected notes into a more long-term format within a Google Doc.
- On Monday mornings, publish the distilled reflections and insights publicly.
Setting Up and Sustaining a 'Nerd Club' / Strategy Salon
Alex Komoroske- Create a secret, optional group with a name that attracts intrinsically motivated participants (e.g., 'Naval Gazers') to ensure positive, 'yes, and' energy.
- Explicitly set norms for collaborative debate: encourage building on ideas, allow non-engagement with uninteresting ideas, and frame critiques as personal wonderings ('I wonder if that applies to you') to avoid making it about the other person.
- Dribble in new perspectives by adding 1 to 3 new people per week, ensuring diversity and minimizing the chance of one person disrupting the norms.
- Maintain community momentum by ensuring there's always an interesting conversation, even when the facilitator is not present; a community with only one person talking is considered 'dead'.
- Actively 'garden' the community by gently pruning back unhelpful interactions, for example, by advising members on how to phrase strong feedback more constructively.
- Encourage sharing by responding to direct pings with, 'That's a really interesting idea, you should share that in the group,' and then engaging positively with it publicly in the group to establish a norm of sharing.
- Periodically suggest live conversations (e.g., an hour-long lunch chat) only after confirming quorum with key participants, ensuring the event always has attendance.
- After live sessions, send 'FOMO' (Fear Of Missing Out) inducing notes or summaries highlighting amazing insights to encourage those who didn't attend to come to the next one.