Playing Infinite Games: Rethinking Talent, Education, and Progress

(from zero-sum, elite machinations to promoting infinite-game thriving)

Jun 02, 2025

“Imagine you want to know how to most effectively select and train the most talented students…knowledge about how best to do so is dispersed across a very long list of different fields…thinkers in these disciplines don’t necessarily attend the same conferences, publish in the same journals, or work together to solve shared problems.”
—Patrick Collison & Tyler Cowen, “We Need a New Science of Progress” (2019)

What if we treated talent selection not as an elite, zero-sum lottery—“Who gets into Stanford? Who is awarded a prestigious MacArthur grant?”—but as a distributed, infinite-game system: “How do we unlock curiosity, grit, and ingenuity in every corner of the globe?” In practice, our current pipelines reward “mathletes” over mathematicians, trophy hunters over team builders, résumé jugglers over authentic problem solvers. Until we confront the ways we “die by elitism” and “die (a slow death) by zero-sum games,” we’ll keep producing Ivy-League-branded automata who are ultimately fine living a lucrative and comfortable life (many people I encountered in investment banking, no shade but just calling a spade a spade; or this piece’s candid author) rather than world-changing collaborators.

Below, I sketch a three-part argument:

Why the current talent identification/cultivation and education system is broken.
How the web of value networks and organizational models locks us into that brokenness.
Why AI’s disruptive potential now presents a once-in-a-generation opportunity—and how we could seize it.

I close with a set of anti-patterns to avoid: false hacks that look like innovation but reproduce old flaws. Think of this as an “infinite-game playbook” for educators, parents, policymakers, and progress-minded entrepreneurs.

1. Why the Current Talent System Is Broken

1.1 Die by Elitism: “Mathlete vs. Mathematician”

Benedict’s “Harry vs. Benedict: A Math Team Story” (benexdict.io/p/math-team) lays it bare: “Mathlete” and “mathematician” sound related, but they’re fundamentally different pursuits. A mathlete hunts trophies—ranking first in a timed contest, memorizing tricks, acing AMC exams. A mathematician wrestles with open-ended problems—crafting proofs, exploring uncharted theory, inventing new questions. Mathletes chase scores; mathematicians chase discovery.

And yet, most Ivy-League pipelines, scholarship programs, and youth accelerators reward “mathlete metrics”: SAT/ACT scores, perfect GPAs, gold medals in math olympiads. We call these “proxies for talent,” but as Patrick Collison and Tyler Cowen warn, they’re often “fake measures”—“substituting a fake measure as a proxy for what we actually want to find.” We say we want future climate scientists, civic innovators, or breakthrough artists… yet we filter by AP scores, résumé bullet clusters, glossy brand names.

Gatekeeper Siloing: Psychometrics in educational psychology; sportometrics in athletics; personality psychology in hiring; sociological network studies in internships. None of these experts talk regularly to each other—so no one sees the full picture.
Gaming the System: Kristopher Abdelmessih’s “Cheating Just Can’t Stop, Gotta Cheat-Cheat” (thezvi.substack.com/p/cheaters-gonna-cheat-cheat-cheat) shows how any pipeline with codified rules (40 hours of private SAT prep, two summer institutes, five leadership roles) invites gaming. Why learn to think critically when a test prep tutor can teach “SSAT answer patterns”? Why invent a novel research project when copying last year’s winning essay nets easier bragging rights? We “die by elitism” because we prize exclusivity and scarcity over genuine potential.

1.2 Die (Slow Death) by Zero-Sum Games

Competition has its place: healthy rivalries push us to excel. But when the entire system becomes a zero-sum battle—“less than 4% of applicants get into Stanford,” “only 20 MacArthur “genius” grants per year”—we cultivate scarcity mindsets and endless anxiety.

Take Sujata Bhatt’s insight in “How to Design Schools to Grow Thriving Adults” (gettingsmart.com/2024/09/05/how-to-design-schools-to-grow-thriving-adults-a-provocation/): when her son played competitive soccer, it wasn’t about soccer skill alone, but about meta-skills:

Team vs. Self Trade-offs: Pass the ball back on the right flank so the left-side striker can sprint free—even if it means sacrificing a goal-scoring opportunity and some personal glory.
Cross-Cultural Collaboration: Teammates came from wildly different backgrounds; success required empathy, communication, and adaptability.
Relentless Effort & Resilience: Facing defeat, picking yourself up, showing up next practice exhausted yet determined.

Yes, that’s a zero-sum soccer championship: one team wins, all others lose. But the meta lessons live in the infinite-sum domain: resilience, empathy, adaptive decision-making. We measure goals scored because it’s easy. But we don’t measure “How did passing back from the right enable a winning play?” or “Which teen learned to respect cross-cultural teammates?” Those untallied virtues—creative confidence, grit, and collaborative awareness—fuel innovation beyond the pitch. When we structure everything as a win/lose scoreboard, we suppress infinite-sum growth. We “die by zero-sum games.”

2. Untangling the Value Networks That Lock Us In

Clayton Christensen’s research on disruptive innovation (including in education) teaches that value networks, not just value chains, shape who thrives. Disrupting a value chain (e.g., bundling/unbundling, integration vs. modularization) is relatively well-trodden and works in a profit-driven north star paradigm. But value networks—the ecosystem of suppliers, customers, regulators, financiers, peer institutions—are stickier.

2.1 Components of an Organizational Model

Christensen’s 1997 framework (summarized via ChatGPT and my own reading) shows that an organization’s capacity to adopt innovation depends on:

Value Propositions: What you offer stakeholders—quality instruction, grade-point averages, campus prestige, alumni network access.
Resources: Faculty, libraries, buildings, endowments, lab equipment, brand equity.
Processes: How you deliver—curriculum design, admissions committees, lecture hall formats, standardized testing protocols.
Financial Formula: Tuition revenue, research grants, alumni donations, endowment returns (higher edu); for public K12 education, typically local tax revenue funds per pupil, with possible state and federal subsidies, program funding (e.g., Title I), and for private schools and micro schools there’s tuition revenue with choice-based vouchers and systems providing publicly subsidy/funding to enable families to choose where to attend (to subsidize tuition pyments)

Together, these four define a school system’s capabilities. To innovate, a school must align new ideas with all four. But even if the innovation “fits” these components, it must also satisfy the broader value network: parents, accreditation boards, teacher unions, local districts, government regulators, testing agencies, philanthropic foundations, and labor markets.

Example: At Streetlight Schools in inner-city Johannesburg, our value proposition is a blended-learning model that integrates CAPS-aligned curriculum with adaptive digital platforms (Siyavula, Khan Academy, PlayLabAI) and a focus on social-emotional learning. Our resources include small-group tuition, a teacher-intern pipeline that trains facilitators in project-based and self-directed learning, and data dashboards to track mastery. Our processes hinge on weekly coaching cycles for new teachers, personalized learning playlists for each student, and project-based modules. Financially, we operate within the government’s per-learner funding range supplemented by partnerships and modest fees to ensure sustainability. Yet the broader value network—provincial education mandates demanding rote instruction and high matric pass rates, parents fixated on traditional grade-point averages, and NGOs prioritizing standardized literacy benchmarks—continually pressures us to conform or risk being forced back into the conventional, exam-driven school model.

2.2 Breaking Free Requires “Choice”

When the grammar of school demands—“cover this content, rank this test, graduate this GPA”—direct confrontation often fails. Instead, we need to ride the growing choice movement like water: using jujitsu to design alternatives that rebuild value-network incentives, and then build traction and trust over time through brand, results, and movement building.

Choose Modular Credentials Over Standard Transcripts: If employers agree to hire based on verified project portfolios—“This candidate built a solar-powered water pump for remote villages; here’s the code, here’s the lab data”—then schools can shift from “teach to the SAT” to “teach to real challenges.” The value network (employers, donors) sees that projects matter more than test scores.
Offer Parallel Pathways That Tap Hidden Demand: Parents fed up with “test prep factory” will pay for after-school “Progress Labs” where kids iterate AI hacks or prototype low-cost sensors. Initially, these labs serve the affluent—building brand equity—until success stories (e.g., “3 high schoolers scaled their open-source malaria-detection device to 20,000 tests per day”) attract philanthropic and governmental backing.

In other words, we don’t try to force Harvard or the state board to accept new models. We create “infinite-game” alternatives that align with real-world value networks: employers need skilled problem solvers, funders want measurable impact, parents want kids who can navigate uncertainty, communities want local solutions. As those external stakeholders demand the new models, traditional incumbents reluctantly adapt or wither.

3. The True Opportunity: AI’s Disruptive, Transformational Moment

Today’s “big code” and Generative AI revolution do more than automate tasks: they rewrite the skills economy. Suddenly, rote memorization and narrow job training are obsolete. Instead, meta-skills—judgment, creativity, ethical reasoning, cross-disciplinary synthesis—become paramount.

Building this to be universally accessible is my goal, but not universally implemented. Read that again. While I think everyone should be given an opportunity to take this path, they should not be required. We don’t need everyone, but just more than we have now, and to build a parallel system (it doesn’t need to completely disrupt and replace the current system, not initially)

Some people work to live, and that’s fine.

Some people live to work. These are the people we want.

3.1 ”Choice” Is the Way

In Star Wars parlance, “This is the Way.” AI can handle data analysis, draft presentations, simulate models, even propose policy briefs. But AI can’t:

Determine which problems are worth solving (meta-crisis awareness).
Balance trade-offs between community and self (Sujata’s soccer metaphor).
Navigate cultural, political, or moral nuances.

Choice lets us simplify the value networks alignment required, as well as on enabling AI and human to play to their respective strengths. By providing learners multiple pathways—modular badges, project sprints, community lab experiences—we drastically reduce friction in adopting edge innovations. Rather than overhaul “grammar of school,” we embed AI tools into progress labs, after-school micro-schools, and homeschools that already have flexible curricula.

A rural Progress Lab repurposes second-hand laptops to run open-source AI tutors. A local nonprofit funds micro scholarships for top performers to join “Talent Sprints.”
A microschool in São Paulo partners with a fintech to let students prototype local payment solutions—AI handles data, but students craft user interviews, business models, and deploy pilots.

Choice jujitsu: we don’t storm the local district or board’s castle; we build thriving ecosystems that render the old pipelines increasingly irrelevant, especially for aspiring leaders and families who want a different approach.

3.2 Alternative Models Aligned with Progress & Meta-Crisis

Under the banner of “progress studies” and confronting the meta-crisis of meaning, we need an umbrella term that encapsulates real skills, real portfolios, real leadership—measured by impact, not by brand. Below are design principles and existing precedents:

Measures That Matter (Enabled by AI & LLMs)
- Skill Assessments: Instead of SAT curves, use AI-driven adaptive assessments (see ETS’s “Charting the Future of Assessment”). Conduct portfolio audits in real time: Did this student’s prototype lower local waterborne disease incidence? AI can track usage metrics, social impact surveys, and cost savings.
- Behavioral & Psychographic Mapping: Use ethnographic interviews, AI-analyzed reflection journals, and gamified tasks to surface hidden values (e.g., perseverance, communal empathy) rather than just GPA.
Build Elite Initially to Capture Mindshare, Then Democratize
- “Barbell Strategy”: High-end tier = flagship Progress Labs (e.g., African Leadership Academy; Polymath University). Charge full price, brand builds prestige, early success attracts media.
- Scholarship Arm: Simultaneously carve out 20–30% seats for under-resourced learners (paid by impact investors, grantmakers). As brand equity grows, philanthropic dollars increase, funding new cohorts.
Position Under a Unified Umbrella
- Super-Schooling (Beyond “Unschooling” & “Microschools”): Call it Progress Schools or Infinite Academies—emphasize building durable meta-skills through core processes (Sujata’s “process over content” model).
- Progress Labs: Physical and virtual hubs where learners aged 14–35 prototype solutions to SDGs (Sustainable Development Goals), iterate lean start-up cycles, and document every step on open registries.
  1. Distinct cohorts
    1. In high school cohort
    2. In college cohort
    3. Early professional cohort
    4. Any age cohort
  2. Cohort-staging makes it much easier, because similar broad life experience, understanding, life stage and maturity, and also schedules/scheduling logistics are more consistent
  3. But we will also intentionally enable cross-pollination across these and inter-generational connection opportunities.
Leverage “Inner Development Goals” (IDGs)
- Self-awareness, emotional intelligence, adaptive resilience, reflective habits. We must teach how to learn not just what to learn—mirroring the “meta-crisis is a crisis of meaning” imperative to ground education in purpose, not just credentialing.
Examples & Precedents
- Thiel Fellowship / Emergent Ventures / Progress Fellows: Each offers micro-grants to high-potential misfits. They build brand and test innovative criteria—often outside academic silos.
- African Leadership Academy & Polymath University: Blend rigorous selection with co-creation labs, then funnel alumni into high-impact ventures.
- MacArthur “Genius” Grants & Mandela Washington Fellowship: Highlight mid-career risk-takers. But selection remains opaque; we need more transparent, values-aligned rubrics.

By unifying these into a coherent Progress Ecosystem, we create a parallel track—no need to topple existing structures immediately. Let them continue, but let us build alternatives that thrive.

4. A Bold Agenda: From Pilot to Planetary Progress

If education is the crucible where human potential is discovered, shaped, and released, then we must rethink its architecture. Here is a six-part action framework for forging “infinite-game” education and talent ecosystems aligned with progress:

4.1. Unbundle & Modularize Learning

Why: Traditional schools still operate on 19th-century “factory rows.” Students sit in age-segmented cohorts from 8 am–3 pm, dutifully copying notes, memorizing facts for tests.
How: Create modular learning credentials—“micro-badges”—that attest mastery of discrete skills:
1. “Engineering Design Sprint: 6-Week Cycle”
2. “Social-Emotional Reflection Journal: 50 Entries”
3. “AI Ethics Research Fellowship: Independent Project” Each badge involves a short project, mentor review, and peer-feedback loop. By stacking badges, anyone can assemble a multidisciplinary portfolio: “AI-for-Cardiac-Diagnostics + Community-Organizing for Health Equity + Data Visualization for Climate Justice.”

4.2. Decentralize Credentialing Through “Progress DAOs”

Why: Degrees and diplomas remain the gatekeepers of hiring, despite being poor proxies for actual skills and mindsets.
How: Build a blockchain-oriented credential registry—a “Progress DAO”—where mentors, peers, and external reviewers can endorse badges, portfolios, and micro-fellowships. Every endorsement is timestamped, immutable, and tied to verifiable project artifacts (GitHub repos, video presentations, white papers). Recruiters query the DAO and see a transparent “scorecard” of impact.

4.3. Forge “Infinite-Game” Fellowships & Sprints

Why: Year-long fellowships (e.g., Thiel, Gates, Rhodes) demand high stakes: drop-outs, huge time commitments, geographic relocation—sacrifices many cannot afford.
How: Launch “Talent Sprints”—three-month micro-fellowships that:
- Provide $5K–$10K seed grants.
- Pair fellows with cross-disciplinary mentor clusters (psychometricians, cultural anthropologists, network scientists, AI ethicists).
- Focus on concrete deliverables: “Build a prototype of a low-cost community water filter,” “Publish an open-access dataset on local air quality,” or “Develop a peer-mentoring curriculum for neurodiverse learners.”
- Host a public showcase (live stream + digital archive) where fellows pitch outcomes to a decentralized panel. Votes from that panel unlock subsequent funding tranches.

4.4. Build “Infinite-Game” Community Hubs

Why: Authentic learning emerges from communities of practice, not solitary screens.
How: Establish regional “Progress Labs”—physical and virtual spaces—where learners of all ages can gather to:
- Prototype hardware (3D printers, low-cost computing kits).
- Collaborate on open-data civic projects (mapping local flood risk, designing microcredit algorithms).
- Host live salons—mixers that pair local artisans, coding mentors, venture builders, and policy advocates—so that the lines between “student,” “teacher,” and “entrepreneur” blur.

4.4. Reorient Policy to Reward Infinite Games

Why: Current education policy often mandates standardized accountability—high stakes for test scores, compliance checks, and funding formulas tied to annual assessments. This stifles risk-taking and innovation.
How: Advocate for “Progress Zones”—jurisdictions (districts, provinces, entire small countries) that receive waivers from certain compliance requirements if they commit to:
- Outcome-Oriented Funding: Replace “money per seat” with “money per demonstrable skill gained”—e.g., “$200 per learner who masters advanced problem-solving” (verified by external panel).
- Open Data Mandates: Schools in Progress Zones publish monthly learning-outcome dashboards (peer-reviewed by researchers), accelerating meta-science.
- Public-Private Partnerships: Local governments, philanthropies, and impact investors co-fund experimental programs (AI tutors, community labs) with rigorous outcome tracking.

4.6. Teach “Infinite Game” Mindsets from Day One

Why: Most K–12 curricula still emphasize rote memorization and ranking—“Who can memorize the Krebs cycle?” or “Who can solve 40 algebra questions in 30 minutes?”—rather than fostering intellectual autonomy.
How: Introduce “Infinite Game Workshops” in middle school:
- Students pick an open-ended challenge (“How can we reduce household food waste in our neighborhood?”).
- They form teams, conduct user research, prototype low-fidelity solutions, gather feedback, and iterate.
- At each step, lessons underscore the difference between “winning a contest” and “advancing a problem into a new frontier.”
- Reflection journals and group debriefs help students internalize that value sometimes lies in the questions, not just the answers.

4.7. Center Core Processes to Develop Real-World Competence (link)

Why: Graduates need more than isolated facts; they need habits of inquiry and iteration. By embedding fundamental processes—rather than only discrete subjects—students learn how to learn, adapt, and create value across any context.

How: Integrate these four core processes into the curriculum as organizing pillars, each with its own activities, competencies, and career pathways:

Scientific Method
- Activities: Observe & research, form hypotheses, conduct experiments, analyze data, draw conclusions, share results.
- Competencies: Critical thinking (evaluating evidence), collaboration (lab teamwork), creativity (designing experiments), ethical awareness.
Engineering Design
- Activities: Identify problems, research requirements, brainstorm solutions, build prototypes, test & refine, implement, communicate outcomes.
- Competencies: Systems thinking (solving technical challenges), creativity (innovating designs), teamwork (multidisciplinary collaboration), clear communication.
Design Thinking
- Activities: Empathize with users, define needs, ideate solutions, prototype rapidly, gather feedback, iterate.
- Competencies: User-centered creativity, analytical refinement, co-creation with stakeholders, persuasive storytelling, community impact orientation.
Lean Startup (Entrepreneurship)
- Activities: Explore market needs, plan strategically, build minimum viable products (MVPs), measure real-world feedback, pivot or persevere, scale solutions.
- Competencies: Hypothesis testing (validating assumptions), business model innovation, stakeholder collaboration, pitch communication, social responsibility.

By repeating these processes in varied contexts—solving environmental, social, or technical problems—students acquire a toolbelt for real-world problem solving. They learn how to learn new content and adapt when faced with novel challenges. Over time, they graduate not just with knowledge, but with a proven ability to create value, collaborate, and iterate—traits that mirror adult professional workflows.

5. Anti-Patterns to Avoid

Below are practices that masquerade as innovation but merely repackage old failures. Avoid these traps at all costs.

Value Chains Without Value Networks
- Anti-Pattern: Treat education as a linear “content → test → diploma” pipeline.
- Fix: Embrace value networks. Forge partnerships among community leaders, employers, NGOs, local government, and funders. Build feedback loops so each stakeholder’s needs inform curriculum design.
Process Lite, Content Heavy
- Anti-Pattern: “Teach X chemistry formula, but we’ll get to why later.”
  Fix: Front-load core processes (see Section 4.7): scientific method, engineering design, design thinking, lean startup. Embed those processes into every lesson—contextualizing content.
Opaque Selection Metrics
- Anti-Pattern: “Apply to our fellowship; good luck figuring out how we’ll score you.”
- Fix: Publish explicit rubrics—quantifiable, peer-reviewed criteria for creative confidence, social impact, cross-disciplinary fluency. Let applicants tailor portfolios to real outcome metrics.
Hero Founder Myths
- Anti-Pattern: “We only accept ‘rockstar’ founders who’ve already built unicorns.”
- Fix: Celebrate team breakthroughs. Give equal weight to “ensemble cast” stories—peer-mentored cooperatives, open-source community projects, distributed research networks.
Curricular Silos
- Anti-Pattern: “You learn history on Monday, biology on Tuesday; never the twain shall meet.”
- Fix: Use process crosswalks to show how core processes map to competencies and careers. A “water-purification” module teaches chemistry, civic policy, social design, and lean startup in one cohesive arc.
One-Size-Fits-All Delivery Models
- Anti-Pattern: “If you can’t access high-speed fiber, tough luck—no remote labs for you.”
- Fix: Build offline-first kits (Raspberry Pi AI tutors, printed workbooks) and train local mentors. Create hybrid models so urban and rural hubs share curriculum, iterate together, and co-publish outcomes.
Short-Lived Fellowships Without Ecosystem
- Anti-Pattern: “Congrats—you’re a Fellow! Now here’s $50K. Good luck navigating alone.”
- Fix: Attach ongoing incubators—virtual Progress Hives, peer cohorts, alumni mentorship, micro-grant follow-ups. Ensure every Fellow remains embedded in a collaborative network long after the award.

6. Epilogue: From Zero-Sum Comfort to Infinite-Game Courage

I started as a typical “Ivy-obsessed” kid and ended up through prep, will, social capital and luck breaking into Wall Street without an Ivy diploma. As a Chinese American, I “won the opportunity lottery of education” and by age 22, I had achieved by any conventional definition — I was working on Wall Street as an investment banker making bank. Yet I felt like I had lived my entire life to satisfy the expectations and hoops that others had put in front of me, without really finding my particular infinite game. I realized: comfort ≠ impact; brand ≠ brilliance. So I set off on a quest (that’s 12 years and counting) to carve my journey.

We face a paradox:

On most metrics, humanity has never had it better (Pinker’s Enlightenment Now, Crawford’s “Progress Studies as a Moral Imperative”).
Yet we suffer a meta-crisis of meaning, polarization, and climate emergency (Ernesto’s “The Metacrisis Is a Crisis of Meaning”).

Education mirrors both trends: it can bounty human flourishing or entrench zero-sum scarcity. I choose flourishing. I choose the infinite game.

Not for the trophy.
Not for the résumé zig-zag.
Not for the brand stamp.

But for the question: “What problem will I solve that no one else has?”

Imagine every community hosting a Progress Lab. Children, teenagers, and adults prototyping: “How to purify our river water cheaply?” Or, “How to deploy AI to predict local crop yields?” Or, “How to co-design a peer-mentoring app for seniors experiencing loneliness?” They document every iteration publicly, earn modular badges, earn decentralized endorsements, and build living portfolios.

When employers, universities, policymakers, and philanthropies see those portfolios—when they demand those outcomes—the old pipelines crumble. SAT prep factories lose relevance. Turf wars over textbook mandates vanish. We build new value networks—networks that value ingenuity over brand, collaboration over competition, impact over pedigree.

“When the cost of reducing risk is directing all your lifeforce to satisfy an endurance test, what are we even doing here? If I need it for me, and I do, then how much effort should I expend proving it meets someone else’s criteria—rather than exploring my own frontier?”

The infinite game demands courage: to die by elitism and zero-sum metrics, then to live by thriving—to cultivate real skills, forge real portfolios, and spark real progress.

Join me. Start a local Progress Lab. Mentor a young tinkerer. Fund a micro-grant that seeds a solar-pump prototype. Help rewrite this story so that every learner, everywhere, can play the infinite game—and blaze new trails for the next century of human progress.

Learning By Design

Discussion about this post