The Lab
A small research lab (5–15 people) studying one question:
How do small, decentralized networks coordinate, compete, and survive against centralized systems — and what are the mathematical structures that govern those dynamics?
Principles (from Moonshot)
- One object. The math of network coordination dynamics. Everyone orbits it. No fragmentation into separate “research” and “product” divisions.
- Hiring is the management system. People who understand dynamical systems AND write production code AND have taste. Generalists who move between theory and implementation.
- No layers between people and the problem. Flat. Direct communication. No middle management.
- AI agents as infrastructure. 5 people with agents operate like 30. Agents handle literature review, data pipelines, simulation runs. Humans handle judgment.
- Commercial pressure funds the research. The research produces the commercial value. Tools work because the theory is right. Theory improves because the tools generate data.
- Accountability = reality. Does the tool work? Does the simulation predict what actually happens? No proxy metrics.
The two outputs
- Theory. Dynamical systems, graph theory, mechanism design applied to decentralized network coordination. Phase transitions in cooperative behavior. Topological conditions for network survival.
- Tools that embody the theory. Software, algorithms, simulations. Things people pay for. The commercial pressure forces the theory to be correct, not just elegant.
Commercial applications
- Platforms bootstrapping network effects. When does a marketplace tip? What’s the minimum node density for cooperative equilibrium? What topology predicts coordination vs. fragmentation?
- Coordination mechanism design. Allocation systems, voting systems, resource sharing protocols. Backed by dynamical systems theory, tested on real data.
- Simulation engines. The TTRPG work is a bounded-rule-system testbed for network dynamics. Game studios, training environments, policy simulators.
- Decentralization infrastructure. Indie retail (Guild), local food systems, cooperatives, credit unions, open source governance. The math transfers. The tools transfer.
Phases
Phase 0 — Now
The lab is one person. Guild is the first tool. The bookstore is the first dataset. The voter model and network dynamics work are the first publications. Revenue: bookstore + Guild subscriptions.
Phase 1 — Post-WithFriends exit (late 2026)
2–3 people. First hires are theorists who build. Output: tools + papers. Revenue: tool licensing, consulting, grants, possibly Mana investment. The Mana pitch reframes from “loyalty SaaS” to “applied network dynamics research, indie retail as first market.”
Phase 2 — Tools have users, papers have citations
5–10 people. Lab has a reputation. SFI, Barabási network, NFX — collaborators or funders. Revenue diversifies across verticals because the math transfers.
The discipline
Every tool, every paper, every commercial engagement must advance understanding of the same question. If it doesn’t, don’t do it. The moment applications become separate product lines with separate teams, you need hierarchy, and the frontier structure dies.
Intellectual lineage
- Whitney Tabor — coordination games on network topologies, emergence of structure in dynamical systems (undergrad advisor, contributed to research 2017–2018)
- Voter model / statistical physics — opinion dynamics, metastability, fixation probability on complex networks
- Mechanism design — designing rules that incentivize truthful reporting in decentralized systems
- NFX — taxonomy of network effects (the lab’s theory is upstream of their framework — studying the dynamics that determine whether network effects emerge or collapse)
- Hanseatic League — historical model of decentralized coordination that preserved autonomy while achieving collective power
Key relationships
- Whitney Tabor — former advisor, warm connection, works at the exact intersection
- Leonard Scheidemantel — independently validated cross-shop intelligence thesis, startup advice
- Phil (Victory Point) — first external validation of coordination mechanism (distributor allocation), potential early tool user
- Jimmy Lee / Mana Inc — potential funder, aligned aesthetics
- Maggie Tang / Loyalist — parallel thesis in different vertical (fine dining), $10M raised, peer founder
The anchor
The bookstore doesn’t depend on the lab to survive. It’s the stable base — revenue, data, and a reason for the first collaborators to connect — while the research finds its footing.