Why STORE / Why Now
L-zero democratic cloud infrastructure. Democratic computing. Not a chain.
Cited in 16 later patent filings, including by IBM, Mastercard, Ant Group, and Mitsubishi. These companies have cited STORE Research patents in their own patent filings - not an endorsement or partnership.
01The problem
Every major digital network in the last thirty years followed the same pattern: starts open, then concentrates. Search became Google. Cloud became Amazon, Microsoft, and Google. Now the same dynamic is playing out in AI.
The six leading AI labs already control the overwhelming majority of frontier capability. Two of them hold more than 80% of usage. Decisions about safety, access, and pricing are made by a handful of executives. There is no structural check that prevents any single actor from crossing a meaningful threshold of control.
Regulators are trying. The problem is timing. AI capability compounds in months. Legislation moves in years. By the time a rule is written, the concentration has already hardened.
Metcalfe's Law is not a trend. It is gravity. The value of a network grows as the square of its users. So does the power of whoever sits at its center. When that math runs on AI - on systems that set prices, write code, run logistics, and make medical decisions - concentration means one company sets the rules for everyone else. That is not a prediction. It is already happening.
The 20th century built laws for this. Antitrust. Regulation. Congressional hearings. None of it changed the math.
The same one-third bound STORE enforces at commit is now running in code. 195 years after the Founders encoded supermajority rules. Zero cross-citation.
STORE is the first machine that enforces it. The token is how the ceiling gets enforced economically - governance, payment, and denomination unified under one mathematical constraint. The United States is a 50-state democracy, bounded by geography and constitution. STORE is a mathematical protocol, not a nation-state. The protocol is the first user. Developers build on it. AI agents run on it. Their users participate under the constitutional ceiling. The math has no geographic ceiling. STORE is the first live constitutional compute-governance experiment of its kind. Hamilton's Law scales.
02The discovery
Hamilton and the Founders studied every failed republic in history and wrote supermajority locks into the US Constitution in 1787. Leslie Lamport proved the oral-message bound f < n/3 in 1982. STORE named the identity-layer ceiling. The connection is named, not identical. 195 years apart. Zero cross-citation. The bound is older than both traditions.
Ordinary people, given the right rules, self-govern better than experts expect. STORE's governors proved it with token pricing. When no actor can control more than one-third, cooperation is the rational strategy.
Hamilton's Law is not a policy. It is a mathematical invariant. The difference matters: a policy can be waived, lobbied away, or ignored. An invariant cannot. When the ceiling is enforced at the database layer, a violation is not illegal - it is impossible. The write crashes. The state never changes. STORE is the first machine that enforces it this way.
126 governors today - no limits on democratic growth tomorrow. Hamilton's Law: as the network grows, stability compounds. Democratic networks become MORE stable at scale. The math makes the network harder to capture as it grows, not easier. The threshold has never been crossed. Every new governor makes it harder to capture. The math makes cooperation the only rational strategy. The constitutional ceiling is 33.3%.
03The moment
Congress wants a kill switch. Silicon Valley says that kills innovation. Both are right. Both are incomplete.
The real solution is not a button held by one person. It is a mathematical law held by the architecture itself.
The Kill Switch Act was introduced July 23, 2026 (bipartisan, publicly reported). The EU AI Act's transparency rules took effect August 2, 2026, and under Regulation (EU) 2026/1744, as adopted, its high-risk rules apply from December 2027. The CEOs of OpenAI, DeepMind, and Anthropic are publicly calling for independent governance. They are describing something that already exists. STORE has been building it since 2017.
126 governors. 30+ countries. Zero violations. The threshold has held for 2,200+ consecutive days.
STORE was nine years early. It is now right on time.
The token powers the democracy that governs the intelligence that governs the compute. The democracy governs the compute. Hamilton AI governs the intelligence running on it. Constitutional mathematics enforced at both layers.
Compute runs today on five partner locations; our own data centers are in development.
STORE governs compute. Once The Governor launches, STORE governs AI. The era of democratic cloud computing begins with STORE. The era of democratic AI begins with The Governor. One math. Two eras. No geographic ceiling.
04One token
STORE is simultaneously a governance token, a utility token, and a payment token. One object. Four functions. One ceiling.
- 01GovernanceYou vote with it.
- 02DenominationEvery compute price is set in it.
- 03PaymentEvery transaction settles in it.
- 04DistributionEvery reward pays in it.
No prior system unified all four functions under a hard mathematical ceiling on concentration. STORE does. See The Token chapter for the full economic case.
Nothing here is an offer to sell securities.
Why nowThe more powerful AI gets, the fewer people govern it.
This is not a conspiracy. It is math. The same Metcalfe's Law that produced Google from search, Facebook from social graphs, and Amazon from retail logistics is now running on AI infrastructure. Network value scales as n-squared. Concentration is the natural resting state of an unconstrained network.
ChatGPT holds 53.9% of consumer AI traffic. Gemini holds 27.9%. Claude holds 9.2%. The top two together exceed the constitutional ceiling the Founders wrote into practice and Lamport proved into math. No governance mechanism anywhere in the AI stack has triggered. No mechanism exists to trigger.
Legislation moves in years. AI moves in months. By the time a regulatory framework reaches a vote, the system it was designed to govern has already changed. This is not regulatory failure. It is a structural timing mismatch between the speed of democratic deliberation and the speed of machine learning.
The CEOs of OpenAI, DeepMind, and Anthropic are now publicly calling for an independent governance body - FINRA-style, IAEA-style. They are describing an institution. Institutions require trust. Trust requires time. The models are not waiting.
Hamilton's Law: max(Pc) < 1/3 at every layer of the democratic compute rail. The Founders wrote two-thirds and three-fourths locks. Lamport proved the oral-message bound. STORE named the identity-layer ceiling and enforces it at commit. STORE built the machine implementation. It has been running for zero constitutional violations since enrollment began.
The architecture makes single-actor takeover a type error.
The old model: trust the institution. The new model: trust the math. The math has been running. The institution is still being proposed.
A constitutional democracy - for compute and AI - one that pays its users.
Every user who helps govern, grow, and secure the network is paid by the protocol in STORE tokens. The more users participate, the harder the network is to capture. The harder it is to capture, the more valuable the constitutional guarantee becomes. Governance, growth, and security are the same activity - rewarded by the same token - under the same mathematical ceiling.
AI advances. Compute scales. Humans remain ≥ 2/3 of governance weight, constitutionally guaranteed. The window to build this is 5-10 years. Without it, we lose the ability to stay in control as AI advances. With it, AI innovation and AI safety coexist - not as a compromise, but as a mathematical consequence.
The Virginia Proof · 1787
When the math is right, cooperation beats extraction. Virginia proved it. STORE governors proved it again.
Virginia was the largest state in the new republic. In the Senate, a Rhode Island citizen held about 10.9 times the voting power of a Virginian (1790 census). Virginia accepted less proportional power because a stable constitutional union was worth more long-term than maximum power in an unstable one. Cooperation was the dominant strategy. The math made it rational.
In January 2022, STORE governors voted themselves a raise, then reversed it. Nobody told them to. They could have voted to extract more. They did not. The math made cooperation the rational choice - the same way it did in 1787.
Going Lower
STORE currently enforces the math at the schema layer - the lowest layer of the database stack, where a violation becomes a type error before any application runs. The vision is to go lower: into compute itself, into energy, into hardware. Constitutional mathematics enforced at the mechanics of how power is allocated. The math does not change as the layer changes. The scope does.
The United States bounded itself by geography. STORE is a mathematical protocol. The math has no geographic ceiling.
STORE is the first live constitutional compute-governance experiment of its kind. This is not a political goal. It is where the math leads.
The same threshold. 800 years. No coordination.
The Haudenosaunee Confederacy discovered it around 1142 while designing a peace between five nations. The Roman Republic encoded it in 509 BCE. When Julius Caesar crossed it in 49 BCE, the Republic ended. Not because the law failed. Because one man exceeded the threshold.
Twelve hundred years later, Hamilton and the Founders encoded it multiple times in the US Constitution. 195 years after that, Leslie Lamport proved it is the tight mathematical bound for any distributed computer network. Zero cross-citation with 1787. He was solving a different problem. He got the same answer.
In December 2017, a founder from Kelso, Washington noticed that the governance rule he was designing for a decentralized network protocol was the same rule Hamilton and the Founders had encoded and Lamport had proved. 195 years apart. Zero cross-citation. One named connection. He did not invent this number. He recognized it. Then he built the first machine that enforces it.
The timeline
Same year. Two parallel tracks.
Neither team knew about the other. Now converging.
Four sections, in full
The canonical origin history. Open any to read.
The Rule That Kept Showing UpFive traditions. 800 years. One answer.Read more
The story of STORE Research is the story of one mathematical rule appearing everywhere McCoy looked - in high school student government, in five centuries of constitutional history across three continents, in the most important proof in distributed computing, and finally in the infrastructure layer for AI.
The rule is simple. No single actor above one-third of control. Decisions require two-thirds agreement. When the threshold holds, the system self-corrects. When it breaks, the system fails.
Five traditions confronted the same trust problem: the Haudenosaunee Confederacy, the Roman Republic, Lateran (1179), the Founders (1787), and Lamport (1982). Lateran and Lamport wrote the two-thirds door independently - one in institutional practice, one in mathematical proof. The others are earlier examples of the same instinct, not identical proofs. The Founders encoded supermajority locks. STORE named the ceiling. The connection is named, not identical.
Same number. Different field. Zero cross-citation. When two completely different traditions arrive at the same mathematics without knowing about each other, that is evidence the math is describing something real about how the world works - not a coincidence, not a choice, a structural discovery.
McCoy's recognition in December 2017: this is not coincidence. Five traditions. 800 years. One answer.
Hamilton's Law is the same discovery made twice, 195 years apart, in two independent domains: constitutional governance (1787) and distributed computing (1982). STORE is the first implementation. Two derivations. Zero cross-citation. One threshold.
Hamilton's Ceiling vs. Metcalfe's LawThe counter-pressure, and why it does not repeal the growth law.Read more
Metcalfe's Law describes how networks grow. As a network gets larger, its value grows faster than its size. This is why Facebook is worth more than the sum of its users, why AI compute is concentrating into a handful of data centers, why the biggest platforms keep getting bigger. The math is neutral. It is simply what unconstrained networks do.
Hamilton's Ceiling is the counter-pressure. It does not repeal Metcalfe's Law - networks still grow in value as they grow in size. What Hamilton's Ceiling does is raise the structural cost of concentration until cooperation becomes the only rational long-term strategy.
Virginia was the largest state in 1787. Under Metcalfe's dynamics it should have captured the founding union. With Hamilton's Ceiling in place, cooperation offered more long-term value than domination. The union grew from 13 states to 50. Equal-weight Senate seats raise the cost of domination as N grows - that is Sherman's design. The same scaling property holds in STORE. As the governor count grows, the capture coalition grows with it. Democracy becomes harder to capture at scale. This is architecturally opposed to Metcalfe's Law - not mathematically inverse.
ChatGPT commands the majority of consumer AI traffic. The top two AI systems together hold more than 80% of usage. No constitutional governance threshold exists anywhere in the AI stack. Metcalfe's Law, unconstrained.
STORE is the first system to enforce Hamilton's Ceiling at the infrastructure layer. Not by policy. Not by training. By database constraint - before any application can decide anything.
The AI concentration of 2026 is the same law running faster on better hardware. The scaling laws that explain why more compute produces better models are also Metcalfe's Law applied to compute infrastructure. The leaders of the labs building AI are beginning to say so publicly. They are right about the diagnosis. The answer is not to pace the frontier. It is to enforce Hamilton's Ceiling at the infrastructure layer - the only force capable of countering a mathematical inevitability.
Think of a public road system. Under Metcalfe's Law, whoever builds the most popular highway eventually controls all transportation - raising tolls, blocking competitors, rewriting the rules. Under Hamilton's Law, the 1/3 ceiling is poured into the asphalt itself. You cannot build a road that breaks the limit. The system rejects it before the road can exist.
How the Discovery Became OperationalFour convergence points across nine years.Read more
Constitutional mathematics as a field did not emerge from a single insight. It emerged from four convergence points across nine years.
2017: THE RECOGNITION. McCoy connected the Founders' supermajority rules to Lamport's oral-message bound - 195 years apart, zero cross-citation. Raghavendra Bhagavatha joined in January 2018, confirmed the convergence, and they filed the first patent.
2019: THE INFRASTRUCTURE INSIGHT. STORE Research pivoted away from a direct consumer token and recognized the theorem applied lower in the stack than initially understood. You cannot govern AI without being able to govern the infrastructure it runs on. This is also the year the key insight crystallized: the same threshold must apply to AI actors, not just humans. The constitutional mathematics applied to machine actors as well as human ones.
Q2 2020: FIRST GOVERNANCE LIVE. Twelve to fifteen human governors made the first Democratic AI governance structure operational. When COVID hit, the governance held. Two years later, in 2022, FTX collapsed. STORE's governors recorded zero constitutional violations the same month.
2024-2025: THE COMMUNITY PROOF. Season 1 (July 2024): 60 voters, 223 votes, 97.22% constitutional ratification. Season 2 (2025): 187 voters, 1,137 votes, triple participation, ceremony February 7, 2026. Zero fraud. Zero disputes. The first constitutional democracy at community scale. Scale does not change the math. It validates it.
2024-2025: FORMALIZATION. December 19, 2024: McCoy began the formal mathematical writing. "This is bigger than I thought." The formal mathematics have begun. Five derivations. 800 years. One answer.
The Record: Hiding in Plain SightNine years. Multi-year fundraising. Governors voted themselves a raise, then reversed it.Read more
Nine years after registering the company. Multi-year community and institutional fundraising. 126 governors in 30+ countries. 2,200+ consecutive days · 5 ratified ballots, FG.006 in SR deliberation · zero constitutional violations - per STORE governance records at storecloud.org.
During that run, in January 2022, the governors reversed their own price increase with a 49% cut on FG.002, taking the price below where it started - when they had the power to profit instead. In most governance systems, the people in power use that power to benefit themselves. The governors of STORE could have held prices high. They voted to lower them. That is what it looks like when the architecture makes cooperation the rational choice.
The governors had real financial skin in the game. FG.001 (Jan 2022): Governors voted to raise the token price - 52.94% approval. The crypto market collapsed the same month. FG.002 (Jan 2022): The community reversed it four days later - 49% cut, 89.47% approval. The governors who would have benefited from high prices voted to cut them instead. They voted to make their own holdings worth less, ratified the reversal at 94.74%, and settled at $0.039 in June 2022, still 20.4% under the $0.049 they began with. Each time the cooperation rate was higher than the extraction attempt. The Nash equilibrium is not a theory at STORE. It is a voting record. Six years of governor votes show consistent cooperative behavior - including votes that reduced governors' own token price. This is not a DAO. A DAO lets token holders vote to extract value. STORE governors voted to return it.
Any controlling coalition requires at least three independent actors - formally proved. At 126 governors, capture requires corrupting 43. At 1,000 governors, 334. As more governors join, capture requires corrupting more of them. Democracy scales against concentration. Architecturally opposed to Metcalfe's Law - not mathematically inverse. Metcalfe is quadratic in N - value concentrates. Hamilton is linear in N under a binding ceiling - stability distributes.
2 granted US patents (US 11,080,691 B2 - US 12,165,142 B2) - 36 granted claims. 180+ invention disclosures in pre-filing and pre-prosecution, being prepared for provisional filing. Across every layer of the architecture.
The operational record is directional evidence. The formal mathematics are in peer review. The threshold, if it continues to hold, is not.
The lab was incorporated as Storecoin Inc. in 2017 at the Consensus conference in New York. Renamed STORE Research Inc. - same entity. Building a democratic AI frontier governance lab while most of Silicon Valley was building AI wrappers. Not hiding. Building.
"One theorem. 800 years of independent derivation. First machine implementation. He kept going for nine years because the math kept being right."
The discovery is the story. The token is the mechanism.
This document is the canonical STORE Research origin history for journalistic and research purposes. It is not investment advice. It does not constitute an offer or solicitation to buy or sell any security or digital asset. Past operational performance does not guarantee future results. All forward-looking statements are based on current expectations and subject to change. Invention disclosures described as "in pre-filing and pre-prosecution" have not yet been filed or granted.
STORE Research Inc. · San Francisco · Founded May 17, 2017
The vision is not a hope. These are the equations that make it a theorem.
Hamilton's Law is not one equation. It is six mathematical constraints working simultaneously - each one named for a builder of constitutional systems. Together they form Multi-Level Power Distribution: the math that makes cooperation the dominant strategy at any scale. Virginia proved it in 1787. STORE has been proving it for 2,200+ days.
Act 1Two laws in conflict
Value flows to the center. The bigger the network, the stronger the concentration. This is why Google, Meta, and two AI labs hold dominant positions. The math demands it.
Stability grows as concentration falls. The bigger the network, the harder it is to capture. The math makes cooperation the rational choice. Opposite of Metcalfe. Same starting point.
Act 2Six constraints. One system.
Each constraint prevents a specific failure mode. All six operate simultaneously. Remove any one and the system can be captured.
Every participant starts with equal base power. One governor, one vote. No one can buy a larger share of governance weight. Economic stake determines entry; it does not determine votes.
No single entity can hold more than one-third of governance weight. The American founders encoded this multiple times in the Constitution in 1787 - in the ratification threshold, treaty ratification, impeachment conviction, veto override, amendment process, and more. Lamport proved it necessary for distributed computing in 1982. STORE enforces it at the database layer - a type error, not a policy.
Every verified participant holds equal governance power regardless of economic stake. Reynolds v. Sims (1964) completed the political layer. STORE carries it into compute.
Vote weight and token holdings are mathematically independent and perpendicular. Your economic stake in the protocol does not determine your voting power. The two systems interact through the token - they do not merge.
Byzantine Fault Tolerance: a distributed system tolerates up to one-third faulty or adversarial actors. Lamport proved this in 1982. STORE named the identity-layer ceiling. STORE is the first governance system to enforce both simultaneously.
On failure boundaries, prior commitments are preserved and no clawback occurs. Failure is a boundary, not a trap. Clean Severance ensures that when a governance process fails, participants exit cleanly without losing prior commitments. Failure is a boundary - not a trap that can be exploited for capture. This closes the exit-boundary capture vector that the five MLPD patterns do not cover.
The scaling inversion
The six constraints have held in production for 2,200+ consecutive days across 126 governors in 30+ countries. The evidence is public.
- 126 governors nowcorrupt 43 to capture · $430,000
- 1,000 governorscorrupt 334 to capture · $3.34M
- 10,000 governorscorrupt 3,334 to capture · $33.34M
Illustrative cost at current token price. No prior network has this property. Metcalfe's Law makes systems easier to capture at scale. Hamilton's Law makes them harder.
Voting thresholds
Five levels of consensus required for different governance actions. Each threshold prevents a specific failure mode. All are enforced at the database layer.
| Threshold | Required | Used for |
|---|---|---|
| Super | ≥75% | Constitutional amendments |
| Qualified | ≥67% | Monetary policy changes, price reparameterization, kill switch activation |
| Simple | ≥51% | Routine protocol updates, non-constitutional parameters |
| Soft | 33% | Constitutional ceiling - maximum concentration allowed |
| Recognition | ≥10% | Agenda setting, proposal submission, threshold to force a vote |
Overturning a decision passed at 67% takes 75%. A matter can be overruled once; after that, the decision is final, and reopening it is a doctrine amendment (75%). Percentages count votes cast, with a quorum of 33% + 1 (42 of 126 today).
Quorum: 33%+1 (42 governors minimum at 126 total). Confirmed by FG.006 operational reality.
The STORE record
One token. Four functions. One constitutional ceiling.
STORE is simultaneously a governance token, a utility token, and a payment token. The same token denominates every compute price, pays every infrastructure bill, governs the protocol's intelligence system, and distributes every reward. No prior system unified all three under a hard mathematical ceiling on concentration.
Autonomous token economies The builder paradigm
Your token. Your compute. Your intelligence. Governed by the same constitutional ceiling.
Builders can use their own crypto asset to pay for compute and intelligence on the STORE rail. STORE accepts any token as currency at the infrastructure layer and converts it into democratic compute. The builder gets their workload run, their AI inference governed, and a Trust Receipt proving it - all paid in their own token.
This changes the paradigm. On Ethereum and Solana, miners and validators are paid only in the native currency. Builders have no monetary sovereignty. STORE is different: the governed rail accepts your token, prices compute in STORE tokens, settles constitutionally, and returns a Trust Receipt. Your economy participates in the democratic compute economy without surrendering its own monetary unit.
This is the monetary premium. The value of holding the infrastructure layer's settlement currency when every autonomous token economy in the democratic compute stack must route through it.
How STORE pulls this off
- 01Go to the infrastructure layer - Layer 0 - where compute itself is governed.
- 02Build Hamilton AI underneath to coordinate it.
- 03Form a constitutional democracy to govern Hamilton.
That makes STORE the first constitutionally governed AI compute protocol on the planet.
The business model One paragraph
Developers and enterprises pay protocol fees to run workloads on the democratic compute rail. A portion of every fee flows to governors, builders, and the treasury in STORE tokens. What makes demand durable rather than speculative: the constitutional ceiling is enforced at the database layer.
Any system that needs a verifiable Trust Receipt - proof that no single actor controlled more than one-third during that inference - must route through STORE. The receipt cannot be produced any other way. The ceiling cannot be lobbied away.
As AI regulation tightens and enterprise AI procurement requires provable governance, the Trust Receipt becomes non-optional. The floor is structural.
Four things it does
- 01Governance You vote with it.Every change to the protocol - compute pricing, emission rates, constitutional parameters - requires a two-thirds governor vote. Governors vote with STORE. One governor, one vote. SAFT enrolls the seat.
- 02Denomination Every compute price is set in it.Every compute price on the democratic compute rail is denominated in STORE tokens. Not ETH. Bits. This is the unit of account for the entire democratic compute economy.
- 03Payment Every transaction settles in it.Multi-currency in. STORE-only out. Whatever enters the rail, settlement happens in STORE tokens. The token is the settlement currency of the democratic compute economy.
- 04Distribution Every reward pays in it.Protocol fees flow to governors, builders, and the treasury in STORE tokens. The constitutional ceiling applies to distribution weight as it applies to governance weight. You cannot concentrate rewards any more than you can concentrate votes.
1 billion tokens authorized. Emission rate requires two-thirds of governors to change. No founder exemption. The ceiling applies from the first write. Token not yet publicly released. Nothing here is an offer to sell securities.
Deployment StagesTen layers. Two live. One in private testing. Seven on the horizon.
One theorem. No geographic ceiling.
This is the Armstrong-through-Turing deployment timeline. For the technical layer stack (L0-L9), see the Democratic AI Stack.
Each stage named for a builder of constitutional systems - from the American founders to Alan Turing, the first theorist of machine intelligence. The math does not change. The scale does.
Threshold Roadmap · Circulating Supply Milestones
| Milestone | What happens |
|---|---|
| Launch | Five partner locations; CLEAR, STORE Pay, STORE Account, First Governance |
| $50M circulating supply · 6 months | Treasury Branch established |
| $100M circulating supply · 6 months | TREASURIES launches (regulatory approval required); self-serve mServices open to others, STORE SIGN first; sixth location, Switzerland |
| $250M circulating supply · 6 months | Security Branch established; sovereign partnership begins |
| $500M circulating supply · 6 months | Second Governance opens; competitive governor auctions begin; seventh location |
| $1B circulating supply · 6 months | Ten locations; edge cloud research; advanced TREASURIES |
0 · Hamilton AI - Constitutional CoreThe database layer where Hamilton's Law is enforced as a type error.LiveRead more
The Constitutional Core is the foundational database layer that implements Hamilton's Law as a type error - no single actor above one-third, enforced before any application can decide anything. This is the layer that has held for 2,200+ consecutive days with zero violations.
1 · HamiltonHamilton AI is the intelligence layer. Building now.Now - buildingRead more
Hamilton AI is being built now. It is the intelligence layer for all STORE core economies: payments (STORE Pay), outgoing distributions (STORE Send), financial accounting (STORE Account), and all 11 sub-economies. Hamilton is the AI that governs the protocol before Armstrong launches. It makes pricing decisions, reads the governance oracle, and executes the calculations that power every sub-economy.
You cannot launch a democratic compute protocol without a governed intelligence layer. Hamilton is that layer. Armstrong only launches when Hamilton is running.
2 · ArmstrongSTORE launches. Hamilton is running. The democratic compute rail goes live.LaunchRead more
Armstrong is the public launch of STORE. By this point Hamilton AI is operational, the three payment primitives (STORE Pay, STORE Send, STORE Account) are live, and the constitutional rail is open to developers. The first democratic compute transactions clear publicly. Hamilton's Law is enforced at the database layer at scale.
Armstrong is not the beginning. It is the moment the work becomes public. Hamilton runs before Armstrong launches. That is the design.
3 · AdamsDemocracy as a Service (DaaS) goes live.Year 1-2Read more
Democracy as a Service (DaaS) becomes the layer other systems build on. Any network, any organization, any AI system can access constitutional governance as a service. STORE provides the governance rail. Others plug in. The democratic compute economy begins to expand.
DaaS is the business model made operational. Once the rail exists, others pay to ride it.
4 · JeffersonThe Governor deploys. Every AI inference gets a Trust Receipt proving no single actor controlled more than one-third.Year 2-3Read more
The Governor sits on top of DaaS. Governed inference goes live. Every output from an AI model under The Governor returns a Trust Receipt proving it operated under the one-third ceiling - no single actor in control. Verifiable. Permanently recorded. No single actor above one-third. Enforced. Verifiable. Democratic AI.
The Governor is the AI layer of the democratic compute economy. DaaS makes it possible. Jefferson is when it becomes real.
5 · WashingtonPlanetary scale. The math governs compute at scale.Year 3-5Read more
STORE governs compute transactions for millions of users through developers and agents on the governed rail. Democratic cloud computing - infrastructure constitutionally constrained rather than corporately controlled. The 50-state limit does not apply. The math has no geographic ceiling. STORE is the first live constitutional compute-governance experiment of its kind.
At Washington scale, the Scaling Inversion is fully visible: the more governors, the harder to capture. No prior network has this property.
6 · Madison / FranklinAutonomous computing. AI agents on the democratic rail.Year 5-10Read more
AI agents participate in the democratic compute economy directly. Builders deploy their own token economies on the STORE rail - paying for compute and intelligence in their own crypto asset. STORE accepts any token as currency at the infrastructure layer, governs the workload constitutionally, and returns a Trust Receipt. Physical AI is next - robotics, autonomous systems, anything that runs on democratic compute. The same constitutional ceiling that prevents faction capture of a cloud protocol prevents faction capture of a robot fleet. Autonomous Token Economy paradigm: your token, your compute, your intelligence - governed by the same constitutional ceiling. Humans remain ≥ 2/3 of governors by mandate. AI participates. It cannot control.
This is the paradigm shift from L1 protocols where miners are paid only in native currency. STORE builders pay in their own token and receive democratic compute + Trust Receipts in return. The monetary premium: every autonomous token economy routes through the STORE constitutional rail. The infrastructure layer's settlement currency accrues the value of everything built on top of it.
7 · TuringMachine identity. AI earns a seat. Humanity holds ≥ 2/3.Year 10-20Read more
Machine identity activates. AI governors can be enrolled by network consensus - up to one-third of the total. For the first time, human and AI governors co-govern a constitutional system. Humans always hold two-thirds. The math guarantees it. The democracy decides when AI has earned its seat.
Not engineers. Not the AI. The democracy decides.
The Constitutional Future
8 · Democratic Edge ProtocolAnyone with a GPU or phone could serve constitutional inference. Hyperscalers buy wholesale. No single provider controls more than one-third.VisionRead more
The democracy is open. Anyone will be able to compute for it. LINCOLN ceiling on providers AND buyers. At 10M phones: ungameable at AGI scale. NOT a current product. Constitutional future.
This is a vision layer, not a product.
9 · Physical AI + QuantumThe same one-third ceiling governs robot fleets, orbital infrastructure, energy grids, and quantum computing. One theorem. Every substrate.Research being finalizedRead more
The math has no substrate preference. Same one-third ceiling. Every substrate.
This is a vision layer, not a product.
A Verifiable Framework for Constitutional Governance of AI.
Four Gates to Launch
The protocol launches when all four gates clear - not on a date. Two gates are controlled internally. Two depend on external Swiss authorities. The math is published so anyone can check it.
Four gates. Sequential. All four must clear before Armstrong launches. The aggregate is published publicly. Each gate advances the next.
Updated August 2026.
SEND, ACCOUNT, and payment infrastructure. Constitutional compute rail. Payment verification integration.
Reach the $2.67M launch milestone. $944.94K committed.
MME counsel (Swiss legal counsel) confirms STORE payment flows do not constitute exchange activity under Swiss law.
Favorable ruling on asset transfer to Switzerland. Clears $6M liability risk.
63% of the way there. All four gates must clear before Armstrong launches. That is the protocol. We will not launch before the math is ready.
Data source: storecloud.org/path-to-launch. When the launch page updates, this page updates automatically. No dates are published. The protocol launches when gates clear.
Two ways in. One network.
STORE is in active development. Peer review is open now. 2,200+ consecutive days. 126 governors. Zero violations.
Register your interest on the Governor Protocol page and we will send you the governor protocol technical paper when it publishes, with information on purchasing.
Join 30+
Join the Public Peer Review Process
This research is proposed, not yet proven. That's why it's headed for public peer review rather than being asserted as settled fact - a claim this size should have to earn it.
Thank you. You're on the list for the public peer review.
Something went wrong. Please try again.
Franklin Templeton projects $3-5T in agentic commerce by 2030. Every agent operating on STORE Cloud carries a Trust Receipt. The democratic compute economy is being built now. Nothing here is an offer to sell securities.
The founding recognition
In December 2017, working alone in San Francisco, Chris McCoy noticed something. The two-thirds signature requirement in BlockFinBFT - the consensus architecture he was building - was the same fraction Hamilton and the Founders had written into the US Constitution multiple times. Which was the same threshold Leslie Lamport had derived in 1982 while solving the adversarial Generals Problem in distributed computing.
Two independent proofs. Zero cross-citation. The same bound. Hamilton encoded it constitutionally in 1787. Lamport proved it mathematically in 1982.
The recognition was not a discovery in the conventional sense. It was the moment when a mathematical law that had always been true became visible to someone who was looking for it. Chris did not invent Hamilton's Ceiling. He recognized the convergence - and built the first continuous implementation.
"He kept going for nine years because the math kept being right."
Nine years
- 2014Data4America founded. Research into distressed community economics. Kelso, Washington becomes the design standard: if it works for Kelso, it works for America.
- 2017May 17 - STORE Research Inc. registered. The recognition happens at Consensus 2017 in New York. McCoy flies home knowing what he has to build.
- 2018Rag Bhagavatha joins as founding technologist. BlockFinBFT architecture begins. First governor onboarding.
- 2020Constitutional governance experiment begins. It has run for six years without a single violation.
- 2022Two granted patents, cited in 16 later patent filings, including by IBM, Mastercard, Ant Group, and Mitsubishi. Governors voted themselves a raise, then reversed it. The governors could have extracted. They didn't. (These companies have cited STORE Research patents in their own patent filings - not an endorsement or partnership.)
- 2024Season 1: the first constitutional democracy at community scale. 60 voters. 97.22% ratification. The constitutional governance mechanism scales beyond compute.
- 2026AI Kill Switch Act introduced July 23, 2026 (publicly reported) - bipartisan. ChatGPT at 53.9% of consumer AI traffic, above Hamilton's Ceiling with no governance mechanism anywhere in the stack. STORE was nine years early. It is now right on time.
The team
The research program
- Constitutional technology portfolio2 granted US patents (US 11,080,691 B2 - US 12,165,142 B2) - 36 granted claims. 180+ invention disclosures in pre-filing and pre-prosecution, being prepared for provisional filing. Cited in 16 later patent filings, including by IBM, Mastercard, Ant Group, and Mitsubishi. (These companies have cited STORE Research patents in their own patent filings - not an endorsement or partnership.)
- Two papers in preparationPhysics of Trust for Computing · Extending Ostrom
- The Kelso ExperimentSeason 1 (July 2024) · 60 voters · 97.22% ratification · Season 2 (2025) · 187 voters
- Open questionsFormal mathematical proof in preparation · Production adversarial validation
STORE was nine years early. It is now right on time.
Two granted US patents. Cited in 16 later patent filings, including by IBM, Mastercard, Ant Group, and Mitsubishi. (These companies have cited STORE Research patents in their own patent filings - not an endorsement or partnership.)
"The Nobel was for the theory. The Controlled Collision Engine is the first machine implementation."
STORE Research Inc. · Founded May 17, 2017 · San Francisco, California