The Infrastructure Loop: Digital Asset Ownership and Economic Resilience

01What this research is about and why it matters

Executive Summary

Working Paper EAD-2026-01 documented that independent businesses across economies pay the same nominal USD price for AI infrastructure but absorb materially different income-adjusted costs. EAD-2026-02 (Judgment Layer Theory) identified the missing institutional architecture through which organisations synthesise information into accountable decisions. Both papers converge on a shared observation: the organisations that face the greatest structural vulnerability are those that do not own what they depend on. EAD-2026-03 examines whether that vulnerability is individually or collectively remediable -- and concludes that while individual firm-level resilience is achievable through the Specialised Digital Asset hierarchy, structural change requires the Infrastructure Loop: a coordinated but non-political mechanism through which ownership and capability distribute outward from concentrated sources to independent builders.

If some portion of Specialised Digital Asset revenue is directed back into the infrastructure layer -- through ownership participation, community education, and cooperative infrastructure development -- what changes in the structure of the AI economy?

Primary Finding

The Infrastructure Loop is a self-reinforcing mechanism by which independent builders can shift the structural composition of the AI economy over time. EAD-2026-01 established that independent businesses face asymmetric infrastructure costs and proposed the Specialised Digital Asset hierarchy as the firm-level response. EAD-2026-03 extends that finding to the collective level: when enough builders direct a portion of Specialised Digital Asset revenue into infrastructure ownership participation and community knowledge transfer, the proportion of economic activity controlled by external infrastructure providers decreases and the proportion accessible to independent builders increases. Across eleven assessed economies, the consistent pattern is that small business participation is lowest and external dependency is highest in the sub-categories most critical to the Loop -- foundation model access, infrastructure IPO access, and venture capital access. This concentration is the structural problem the Loop is designed to address. It cannot be dissolved by individual action alone. It can be changed by coordinated collective action that does not require political coordination to begin.

Economies Assessed

US, China, India, UK, Germany, France, Brazil, South Korea, Singapore, UAE, Sub-Saharan Africa (representative economy)

Participation Categories

Infrastructure, Digital Assets, Capital Participation, and Competence -- assessed with three metrics per sub-category across all economies

~5%

SME Foundation Model Access

Small business participation in foundation model ownership across all assessed economies, including the US (Source economy)

~45%

SME Data Centre Construction

SME share of global data centre construction market -- the sub-category with highest independent builder participation (Coherent Market Insights, 2025)

USD 99B

CoreWeave Revenue Backlog

Q1 2026 revenue backlog, alongside USD 28B combined financing -- the scale of the infrastructure IPO wave currently in progress

Ecosystem Groups

Settlers (Group A1: infrastructure ownership), Connectors (Group A2: knowledge transfer), Circulators (Group B: Specialised Digital Asset builders)

Who this research serves

Independent Founders

To understand where Specialised Digital Asset revenue can be directed to generate both individual and collective resilience -- and which sub-categories of infrastructure participation are most accessible in their economy.

Community Builders and Educators

To understand their structural role as Group A2 Connectors in the Infrastructure Loop, and why knowledge transfer is the highest-leverage, lowest-barrier entry point for any independent builder.

Investors and VCs

To identify the infrastructure IPO wave as the current window for independent builder participation in ownership -- and to assess portfolio exposure to the participation gap documented across all eleven economies.

Policymakers

To understand the three-level governance framework -- individual builder, community, institutional -- and identify which level is most tractable for policy intervention in their jurisdiction.

02What is being lost and why

The Problem

Working Paper EAD-2026-01 ended with a prescription for individual builders: use general-purpose AI as productive infrastructure to build owned Specialised Digital Assets that compound in value and remain under the builder's control. That prescription is correct for the individual. It does not change the structural condition. The AI economy's infrastructure -- foundation models, cloud compute, AI chip fabrication -- remains concentrated in five to six entities in two economies. Independent builders using open-weight models and cloud compute are still building on infrastructure they do not own. The income-adjusted cost asymmetry EAD-01 documented is still present. The five financial mechanisms are still active. The question EAD-01 does not answer: does the existence of Specialised Digital Assets, in sufficient quantity and diversity, change anything at the infrastructure layer? This paper examines that question through the Economic Participation Framework applied across eleven economies.

The infrastructure IPO wave is active. CoreWeave raised USD 28B in combined financing and entered Q1 2026 with a USD 99B revenue backlog. Nscale raised USD 2B at a USD 14.6B valuation in March 2026. The capital formation and public ownership events that will determine the infrastructure ownership structure of the next decade are occurring now. The participation gap in these events is not a future risk -- it is a present condition. The consistent pattern across all eleven economies assessed by the Economic Participation Framework is that small business participation is lowest and external dependency is highest in the sub-categories most critical to the Infrastructure Loop: foundation model access (~5% SME participation even in the US), infrastructure IPO access (~32% retail allocation in the US, declining to near-zero in Critical-EADI economies), and venture capital access (restricted to accredited investors in most economies). This concentration is not an accident of market structure -- it reflects decades of capital accumulation in entities that are, by definition, not accessible to most builders.

Existing responses to AI infrastructure concentration operate at the geopolitical or macroeconomic level: national AI strategies, sovereign compute initiatives, export control policies, and multilateral digital trade frameworks. These responses are necessary but structurally slow and politically dependent. They require government coordination, budget allocation, and implementation timelines measured in years. The Infrastructure Loop operates at a different level: it does not require political action to begin. Any independent builder who directs a portion of Specialised Digital Asset revenue into infrastructure ownership participation, and who teaches one other person to do the same, is already participating in the Loop. The 1% Education Mechanism -- a proposed norm rather than a regulation -- is the knowledge transfer component of the Loop: each builder who has benefited from community knowledge transfer contributes 1% of their time to teaching the next builder. This scales without institutional overhead.

Even in the United States -- the Source economy with EADI 92 and the most favourable structural position for independent builders -- Foundation Model Access shows ~95% established ecosystem control and only ~5% small business participation. OpenAI, Anthropic, Google, and Meta own all commercial-scale US foundation models. Independent builders are almost entirely consumers at this layer. This is the most concentrated sub-category in the Economic Participation Framework across all eleven economies. Open-weight models (Llama, Mistral, Falcon) represent a partial structural exception, but fine-tuning and deployment still depend on cloud compute infrastructure that remains hyperscaler-concentrated. The Infrastructure Loop does not require independent builders to compete with these entities -- it requires them to participate in the capital structures that fund the next generation of infrastructure.

The current infrastructure IPO wave is the first opportunity in a decade for independent builders to participate in the ownership of AI infrastructure at any scale. CoreWeave (USD 99B revenue backlog, Q1 2026) and Nscale (USD 14.6B valuation, March 2026) represent the first commercial-scale AI infrastructure companies accessible to retail investors. In the US, retail investors received approximately 32% of IPO allocations in 2025. In Critical-EADI economies (India, Brazil, Sub-Saharan Africa), this access narrows dramatically: currency risk, accredited investor restrictions, and minimum investment thresholds combine to exclude most independent builders from participation entirely. The Infrastructure Loop proposes that independent builders in accessible economies treat infrastructure company ownership as a standard component of Specialised Digital Asset returns -- not as speculation but as a structural response to the ownership concentration that EAD-01 documented.

SMEs comprise approximately 45.3% of the global data centre construction market (Coherent Market Insights, 2025). This is the highest small business participation percentage in any infrastructure sub-category across all eleven economies. Independent builders participate extensively in constructing the infrastructure they then pay to use at above-proportional income-adjusted costs. This is the Infrastructure Loop in reverse: labour participation in building infrastructure, combined with zero equity participation in owning it. The pattern is consistent: independent businesses bear construction and operational costs, while capital appreciation and revenue from the infrastructure accrues to the entities that own it. The Infrastructure Loop does not claim this will reverse quickly. It claims the reversal begins when individual builders accumulate enough Specialised Digital Asset revenue to move from labour participation to ownership participation.

03The precise question this research addresses

Research Question

If independent builders direct a portion of Specialised Digital Asset revenue into infrastructure ownership participation and community knowledge transfer, does this create a self-reinforcing mechanism -- the Infrastructure Loop -- that expands independent builder participation in the AI economy over time?

Context

The question emerged from examining what EAD-01 left unanswered. EAD-01 established the structural cost asymmetry and proposed the Specialised Digital Asset hierarchy as the individual firm-level response. The natural next question is whether individual responses, in aggregate, change anything structural. The Infrastructure Loop is the proposed mechanism by which they do. The three-group ecosystem (Settlers, Connectors, Circulators) provides the analytical structure for understanding how builders at different stages of the Loop interact.

Scope

Eleven economies assessed using the Economic Participation Framework: United States, China, India, United Kingdom, Germany, France, Brazil, South Korea, Singapore, United Arab Emirates, and Sub-Saharan Africa (representative economy). Japan and Canada are retained for EADI reference but not assessed in full in this iteration. The analysis covers the period 2024-2026 with primary data from 2025-2026. The unit of analysis is the independent builder in the context of their economy's participation structure across Infrastructure, Digital Assets, Capital Participation, and Competence.

Assumptions

Specialised Digital Asset revenue can be generated by independent builders operating in high-EAD environments, consistent with EAD-01's hierarchy.
Infrastructure company equity is accessible to retail investors in at least some economies -- a condition that the CoreWeave and Nscale IPOs have partially validated.
Community knowledge transfer has a measurable effect on the number of independent builders capable of building Specialised Digital Assets.
The 1% Education Mechanism, if adopted as a community norm rather than a regulation, can sustain a knowledge transfer cycle without institutional overhead.
The three-group ecosystem (Settlers, Connectors, Circulators) is a useful analytical structure even where the boundaries between groups are not crisp.
The Infrastructure Loop produces systemic market change only at scale -- individual participation is necessary but not sufficient.

Out of scope

Enterprise AI adoption economics and the strategy of large-firm infrastructure ownership.
Geopolitical dimensions of AI infrastructure sovereignty at the state or multilateral level.
Technical specifications for cooperative infrastructure ownership vehicles.
Regulatory design for open-weight model access or infrastructure IPO participation requirements.
Intra-economy variation in participation barriers across regions, income groups, or demographic sub-populations.
Empirical validation of the Infrastructure Loop as a mechanism -- this paper proposes and analyses it; empirical testing requires longitudinal data not yet available.

04How the research was conducted- four phases

PaaF Methodology Journey

The first phase required constructing a measurement framework that could capture economic participation across eleven diverse economies in a comparable way. The Economic Participation Framework organises the AI economy into four categories: Infrastructure (the physical and software stack that AI runs on), Digital Assets (the owned outputs that AI enables), Capital Participation (the financial instruments through which ownership is accessed), and Competence (the knowledge and skills that enable building).

Each sub-category within these four categories is assessed using three percentage metrics: Established Ecosystem % (share controlled by governments, large enterprises, dominant vendors), Small Business Participation % (share attributable to SMEs, startups, and independent builders), and External Dependency % (share dependent on products, services, IP, capital, or expertise originating outside the assessed country).

Evidence classification codes distinguish between direct empirical evidence from named primary sources [A], estimates derived from primary data using explicit methodology [B], and proposed mechanisms with no prior empirical test [C]. The distinction is critical: many infrastructure sub-category figures are [B] derived, reflecting the absence of systematic measurement of small business participation in AI infrastructure -- which is itself a finding about how rarely this question has been asked.

The mapping phase revealed a structural pattern consistent across all eleven economies: Small Business Participation % is lowest and External Dependency % is highest in the three sub-categories most critical to the Infrastructure Loop: Foundation Model Access (~5% SME participation even in the US), Infrastructure IPO Access (~32% retail allocation in the US, declining to near-zero in Critical-EADI economies), and Venture Capital Access (~8% non-accredited investor participation in the US). This pattern holds regardless of a country's overall EADI score.

Questions asked

Which sub-categories of the AI economy have the lowest independent builder participation across all assessed economies?
Is the participation gap in foundation model ownership a US-specific condition or universal?
How does external dependency correlate with small business participation across the four Economic Participation Framework categories?
Does the consistent pattern of low SME participation in critical sub-categories suggest a structural barrier or merely current state?

Patterns identified

Foundation Model Access shows ~95% established ecosystem control even in the US, making it the most concentrated sub-category globally.
Infrastructure IPO Access is the sub-category where the gap between US participation (~32% retail) and Critical-EADI economy participation (~5%) is most acute.
SME Construction participation (~45% globally) is highest in the infrastructure layer, creating the paradox of builders who construct but do not own.
Community Knowledge Transfer consistently shows the highest SME participation percentages across all economies, suggesting the Connector role is most accessible.

The participation gap mapping established that the structural barrier to the Infrastructure Loop is not primarily a competence gap -- SME participation in Knowledge Transfer and Entrepreneurship is high -- but a capital and access gap at the ownership layer. Independent builders have the skills to participate; the ownership structures are designed in ways that limit their access.

The second phase applied the Economic Participation Framework identically across all eleven assessed economies. The assessment covers four categories, each with multiple sub-categories, producing a structured picture of where independent builder participation is accessible, where it is blocked, and where it is absent due to missing data.

The United States presents the baseline case: meaningful large-firm concentration in infrastructure, but genuine small business participation across digital assets, capital markets, and competence. Retail investors receive approximately 32% of IPO allocations and account for approximately 20.5% of daily equity trading volume. For the Infrastructure Loop, US independent builders occupy the most favourable structural position: they can invest in infrastructure company equity without currency conversion costs and face no regulatory exclusion from IPO access.

Critical-EADI economies (India EADI 12, Brazil EADI 7, Sub-Saharan Africa EADI 5) show the participation pattern in most acute form. Brazil has near-total import dependency for AI hardware (95% external dependency), near-zero domestic foundation model presence, and infrastructure IPO access restricted to an estimated 5% of retail investors due to regulatory barriers, currency risk, and minimum investment thresholds. Sub-Saharan Africa, while having the lowest EADI score, shows a distinctive structural asset: the highest community knowledge transfer distribution percentages in the study, reflecting a mobile-first infrastructure base and strong community learning networks.

The assessment reveals one consistent finding across the EADI spectrum: economies with the lowest EADI scores show the highest Community Knowledge Transfer distribution percentages. This inversion suggests that the Connector role -- Group A2 in the three-group ecosystem -- has its most natural conditions precisely where the Infrastructure Loop is most needed. Knowledge transfer requires no regulatory reform, no capital access, and no infrastructure ownership to begin.

Questions asked

Which economies have structural conditions that make Group A1 (Settler/infrastructure ownership) participation accessible to independent builders?
Does the mobile-first infrastructure of Sub-Saharan Africa create structural advantages or disadvantages for the Infrastructure Loop?
Is the currency risk associated with USD-denominated infrastructure equity the primary barrier to Capital Participation in Critical-EADI economies?
What structural asset in each economy is most underutilised in the context of the Infrastructure Loop?

Patterns identified

EUR-denominated infrastructure company access (Nscale, Mistral, OVHcloud) partially resolves currency risk for European independent builders but not for Critical-EADI economies.
Singapore and UAE show EADI scores (26 and 24 respectively) that provide material mitigation of currency risk -- SGD/USD stability and AED/USD peg -- not available in Critical-EADI economies.
Germany's Mittelstand represents the highest concentration of Level 2-3 Specialised Digital Asset potential in any assessed economy -- deep domain expertise across industrial sectors accumulated over decades.
The infrastructure IPO window is currently open only in the US market; EUR-denominated equivalents are emerging (Nscale, Mistral potential IPO) but not yet at comparable scale.

The country assessments confirm that the Infrastructure Loop's three groups have different structural conditions across economies. Group A1 (Settler) participation is most accessible in the US; Group A2 (Connector) participation is most active in Critical-EADI economies; Group B (Circulator) conditions are most favourable in economies where domain expertise is deep and Specialised Digital Asset costs are declining.

The third phase synthesised the participation gap mapping and country assessments into the Infrastructure Loop mechanism. The Loop has four sequential activities: (1) Build Specialised Digital Assets using general-purpose AI as productive infrastructure. (2) Generate revenue from those assets. (3) Direct a portion of revenue into infrastructure ownership participation -- equity in infrastructure companies, cooperative infrastructure structures, open-weight model investment. (4) Transfer knowledge to the next builder.

The three-group ecosystem describes how builders participate in the Loop at different stages. Group B (Circulators) are the primary building group -- they use AI to create Specialised Digital Assets and generate the revenue that can flow into the other groups. Group A1 (Settlers) own infrastructure and provide the base on which Group B builds -- they are the destination for ownership participation. Group A2 (Connectors) are educators and community builders -- they transfer knowledge from experienced builders to new ones, expanding the population of Group B participants.

The Specialisation Flywheel extends EAD-01's Specialised Digital Asset hierarchy into a dynamic model. As Group B builders develop higher-level assets (moving from Level 1 Reusable Workflows toward Level 5 Specialised Domain Products), their revenue increases and their dependence on general-purpose AI infrastructure decreases. Revenue from Level 4-5 assets provides the capital buffer to participate in Group A1 infrastructure ownership and to spend Group A2 knowledge transfer time. The Flywheel connects individual asset development to ecosystem-level participation.

The 1% Education Mechanism is the knowledge transfer norm that sustains the Loop without institutional overhead. Each builder who benefits from community knowledge transfer contributes 1% of their time to teaching the next builder. At scale, this produces a self-sustaining educational cycle that expands the population of capable builders faster than any formal educational institution could, because it operates on current, practical, specific domain knowledge rather than generalised curriculum.

Questions asked

At what point in the Specialisation Flywheel does a builder accumulate sufficient capital buffer to participate in Group A1 infrastructure ownership?
Is the 1% Education Mechanism sufficient to sustain knowledge transfer at the pace required to outrun the adoption gap between large and small firms?
Which economies have the structural conditions for Group A1 participation to exist in a domestically accessible form?
Does the Infrastructure Loop require all three groups to be simultaneously active, or can it operate with only two of the three?

Patterns identified

The Infrastructure Loop does not require political coordination to initiate -- it begins when any builder directs a portion of Specialised Digital Asset revenue into ownership or education.
The 1% Education Mechanism is structurally self-reinforcing: each builder who teaches produces a new teacher, not just a new builder.
Group A2 (Connector) activity has the lowest barrier to entry and the highest leverage in low-EADI economies -- it requires only time and knowledge, neither of which is currency-denominated.
The Specialisation Flywheel produces compound returns: Level 5 assets generate revenue from ownership; Level 1-2 assets generate revenue from services -- both contribute to the Loop, at different volumes and timelines.

The Infrastructure Loop is a mechanism, not a policy. It does not require regulatory action to begin. It requires enough builders to direct enough Specialised Digital Asset revenue into ownership participation and knowledge transfer that the proportion of economic activity accessible to independent builders increases measurably over time. The evidence from eleven economies suggests the conditions for this are present in every assessed economy at the Connector layer, and in US and EUR-zone economies at the Settler layer.

The fourth phase translates the Infrastructure Loop mechanism into a structured framework for action at three governance levels: individual builder, community, and institutional. The levels are not sequential -- they operate simultaneously -- but each level has actions it can take without requiring action at the other levels.

At the individual builder level, the primary actions are: (1) Progress along the Specialised Digital Asset hierarchy from Level 1 to Level 5, using the revenue from earlier levels to fund development of higher levels. (2) Direct a portion of revenue into infrastructure company equity where accessible (US market directly; EUR market via Euronext for European builders). (3) Contribute 1% of time to the knowledge transfer cycle -- writing documentation, running community sessions, mentoring one builder. (4) Treat the Infrastructure Loop as a financial planning component, not a charitable act -- ownership participation is the mechanism through which individual resilience becomes collective structural change.

At the community level, the primary actions are: coordinating Group A2 (Connector) networks that transfer current, domain-specific knowledge rather than generalised curriculum; creating shared tooling and resources that reduce the cost of Level 1-2 asset development for new builders; and surfacing accessible Group A1 (Settler) participation opportunities (ETFs, fractional infrastructure equity, cooperative structures) for community members who have generated sufficient Specialised Digital Asset revenue.

At the institutional level, the most tractable interventions are: expanding accredited investor definitions to increase infrastructure IPO access for non-institutional retail investors; designing AI tool cost subsidy programmes calibrated to income-adjusted cost burden rather than nominal price; and developing national Specialised Digital Asset development programmes that reduce the time-to-first-asset for independent builders in Critical-EADI economies. The paper notes that institutional action at this level is most tractable in EUR-zone economies, where domestic infrastructure alternatives (Mistral, Nscale, OVHcloud) provide policy leverage unavailable to Critical-EADI economies.

Questions asked

What action at the individual builder level produces the highest leverage for the Infrastructure Loop in a Critical-EADI economy where Group A1 participation is structurally inaccessible?
Can the community level sustain the Infrastructure Loop in the absence of institutional action?
Which institutional interventions are most tractable within existing legal and regulatory frameworks?
Does the three-level governance framework require all three levels to be operating simultaneously for the Loop to function?

Patterns identified

The community level can sustain the Loop independently of institutional action -- Connector activity requires no regulatory change to begin.
Individual builder action in Critical-EADI economies is most productive at the Connector and Circulator roles -- Settler participation requires capital access that is structurally restricted.
Institutional action is most valuable at the Capital Participation layer (accredited investor access, IPO allocation requirements) -- these are the highest-barrier sub-categories across all assessed economies.
The three levels reinforce each other but do not depend on each other: the Loop begins at whichever level first accumulates sufficient builders, revenue, and knowledge transfer activity.

The governance framework confirms that the Infrastructure Loop is not a top-down initiative. Institutional action accelerates the Loop but does not initiate it. The Loop begins at the individual and community levels, where the barriers are lowest and the knowledge transfer cycle can begin immediately. Institutional action then becomes the mechanism that opens the higher-barrier ownership participation sub-categories as the Loop scales.

05The Infrastructure Loop

Interactive Framework

Four-stage self-reinforcing mechanism connecting Specialised Digital Asset development to infrastructure ownership participation and community knowledge transfer.

Click any node to see details.

Input / Process

Output

Constraint

Ecosystem

Module

Tier

06Browse and filter the evidence base

Evidence Explorer

Evidence underlying the research findings, organised by PaaF phase and evidence strength. Use the filters to focus on the evidence most relevant to your evaluation.

Claim

Source

Strength

Phase

Foundation model access shows ~95% established ecosystem control even in the US, making it the most concentrated sub-category across all 11 assessed economies.

CNAS Sovereign AI Index (2026); Stanford HAI AI Index (2026)

Strong

Retail investors received approximately 32% of IPO allocations in the US in 2025; access narrows to near-zero in Critical-EADI economies due to currency risk and regulatory barriers.

Barclays Private Bank (January 2026); SQ Magazine (October 2025); Forge Global (2026)

Strong

Accredited investor definition restricts VC access to approximately 13% of US households; ~92% institutional/accredited control, ~8% non-accredited investor access.

SEC (2023); PitchBook/SVB (2025)

Strong

SMEs comprise approximately 45.3% of the global data centre construction market -- the highest small business participation in any infrastructure sub-category across all 11 economies.

Coherent Market Insights (2025)

Strong

Sub-Saharan Africa shows approximately 70% community knowledge transfer distribution outside large-firm structures -- the highest in the study despite the lowest overall EADI score (5).

GSMA (2025); Oxford Insights GAIRI (2025)

Moderate

Brazil has zero domestic foundation model at commercial scale, with ~97% external dependency -- the highest foundation model external dependency of any economy assessed.

CNAS Sovereign AI Index (2026); Stanford HAI AI Index (2026)

Strong

CoreWeave entered Q1 2026 with a USD 99B revenue backlog and USD 28B in combined financing, marking the scale of the active infrastructure IPO wave.

CoreWeave company filings; Sacra Q1 2026

Strong

The performance gap between proprietary and open-weight models narrowed from ~8% to ~1.7% between 2023 and 2025, reducing switching costs at the model layer for Circulator builders.

Stanford HAI AI Index (2025)

Strong

07Research maturity, validation status, and roadmap

Research Dashboard

Research Maturity

Evidence Strength

Analytical Confidence

Commercial Applicability

Scores out of 100. Based on internal research assessment criteria. Not independently validated.

Validation stage: Working Paper -- July 2026

Implementation status: Published. Empirical validation of the Infrastructure Loop mechanism is pending.

Known limitations

▪Country-level Economic Participation Framework figures are predominantly derived ([B]) -- primary data on SME participation in AI infrastructure does not systematically exist.

▪The Infrastructure Loop mechanism is classified [C] (proposed, no prior empirical test). Evidence supports the participation gaps; evidence that the Loop produces the claimed structural changes is pending.

▪EADI scores are author-derived prototype estimates (v1.0), not independently validated.

Open questions

?What distributed infrastructure equity ownership threshold produces a measurable change in AI infrastructure concentration?

?What is the minimum builder population required for the 1% Education Mechanism to sustain a self-reinforcing cycle?

?How does the Infrastructure Loop interact with government sovereign AI programmes in Critical-EADI economies?

Research roadmap

Economic Participation Framework data refresh and methodology publication (2027)

Empirical tracking of infrastructure IPO retail participation rates across assessed economies

EAD-2026-04: Collective intelligence layer (tentative)

08What this research means for different stakeholders

Commercial Implications

The Infrastructure Loop determines whether Specialised Digital Asset revenue stays within the individual or cycles back into the ecosystem. Founders who direct a portion of revenue into infrastructure ownership and knowledge transfer participate in the mechanism that expands the builder population -- and therefore the demand for Specialised Digital Assets over time.

Opportunities

Direct a portion of Specialised Digital Asset revenue into infrastructure company equity while the current IPO wave is open.
Operate as a Group A2 Connector through the 1% Education Mechanism -- each hour of teaching is a structural contribution to the Loop.
Use open-weight models to reduce infrastructure cost as a proportion of revenue, creating a larger capital buffer for ownership participation.
Build Level 4-5 Specialised Digital Assets whose revenue scale enables meaningful Group A1 participation.

Risks

The infrastructure IPO window is time-limited -- the current wave reflects a specific capital formation phase unlikely to recur at the same scale.
Currency risk for non-USD builders is not resolved by the Loop -- it must be managed alongside the EAD financial mechanisms from EAD-2026-01.
Premature ownership participation before generating sufficient SDA revenue weakens the individual development base.

Questions to ask

What proportion of your current Specialised Digital Asset revenue can you direct into infrastructure ownership without constraining development investment?

Which infrastructure company equity is accessible from your economy without prohibitive currency risk?

Are you currently operating as a Group A2 Connector? What is the cost of contributing 1% of your time to knowledge transfer?

09Limitations, alternative perspectives, and validation gaps

Critical Review

Research Limitations

Derived country-level figures

The majority of Economic Participation Framework percentage figures are derived ([B]) rather than directly measured ([A]). Primary data on small business participation in AI infrastructure sub-categories does not systematically exist -- the absence of this data is itself a finding, but it means the country assessment tables are indicative rather than precise.

Proposed mechanism without empirical test

The Infrastructure Loop as a causal mechanism is classified [C] -- a proposed pathway with no prior empirical test. The evidence supports the existence of the participation gaps the Loop is designed to address, but does not demonstrate that the Loop, if initiated, produces the structural changes claimed.

EADI score methodology

EADI scores are author-derived from available institutional sources. The index is a prototype (v1.0) and has not been independently validated. Scores should be interpreted as order-of-magnitude indicators of structural position rather than precise measurements.

Cross-economy comparability

Primary data sources vary by economy -- some figures draw on government statistics, others on commercial research, and others on derived estimates. The methodological heterogeneity across sources limits the precision of cross-economy comparisons.

Alternative Perspectives

Claim: Individual Specialised Digital Asset accumulation is sufficient -- the Loop is unnecessary.

Response: EAD-2026-01 established the individual-level response. EAD-2026-03 examines whether individual responses, in aggregate, change anything at the infrastructure layer. These are sequential questions, not alternative positions. Individual resilience is necessary; the Loop adds the collective dimension.

Claim: Government AI sovereignty programmes are more effective than the Infrastructure Loop.

Response: Government programmes and the Infrastructure Loop operate at different timescales and require different enabling conditions. Government programmes require political coordination and budget allocation measured in years. The Loop begins when any builder directs any portion of SDA revenue into ownership or education. The two approaches are complementary, not competitive.

Claim: The 1% Education Mechanism is too small to produce meaningful systemic effects.

Response: The 1% figure is proposed as a starting norm, not a cap. The compound dynamics of knowledge transfer -- each builder who teaches produces a new teacher, not just a new builder -- mean that even small initial adoption rates can produce non-linear effects over time. The mechanism requires no capital and no regulatory change to begin.

Validation Gaps

No empirical test of the Infrastructure Loop mechanism exists at time of publication.
The relationship between distributed infrastructure equity ownership and structural changes in AI infrastructure concentration has not been studied.
The uptake rate required for the 1% Education Mechanism to produce a self-sustaining cycle is unknown.
Country-level EADI scores require independent validation by researchers with access to primary data for each assessed economy.

Future Work

Longitudinal tracking of infrastructure IPO retail participation rates across assessed economies.
Empirical study of Specialised Digital Asset development rates and revenue profiles in Critical-EADI economies.
Refinement of EADI methodology with independent researchers contributing country-specific data.
Assessment of the 1% Education Mechanism in existing builder communities to estimate adoption and knowledge transfer velocity.
EAD-2026-04: The collective intelligence layer -- whether coordinated research sharing across independent builders creates a compounding knowledge advantage analogous to the Infrastructure Loop for ownership.

10Key terms and definitions used in this research

Glossary

Infrastructure Loop

The self-reinforcing mechanism by which Specialised Digital Asset revenue is directed into infrastructure ownership participation and community knowledge transfer, expanding the base of independent builders and increasing their share of AI economy ownership over time.

Economic Participation Framework

A measurement framework assessing independent builder participation across four categories -- Infrastructure, Digital Assets, Capital Participation, and Competence -- using three metrics per sub-category: Established Ecosystem %, Small Business Participation %, and External Dependency %.

Settlers (Group A1)

Independent builders who own or co-own AI infrastructure -- servers, compute capacity, shares in infrastructure companies, or equity in cooperative AI infrastructure structures. Settlers are the destination for Infrastructure Loop ownership participation.

Connectors (Group A2)

Independent builders who transfer knowledge from experienced builders to new entrants -- through documentation, community sessions, mentoring, and educational content. Connectors expand the population of capable Circulators without requiring capital or infrastructure ownership.

Circulators (Group B)

Independent builders who use general-purpose AI to create Specialised Digital Assets, generate revenue from those assets, and direct a portion of that revenue into Group A1 (ownership) and Group A2 (education) activities. Circulators are the primary building group in the Infrastructure Loop.

Specialisation Flywheel

A dynamic extension of the Specialised Digital Asset hierarchy in which progress from lower-level to higher-level assets generates increasing revenue and decreasing infrastructure dependency, creating a compound effect that accelerates access to Group A1 participation over time.

1% Education Mechanism

A proposed community norm (not a regulation) in which each builder who has benefited from community knowledge transfer contributes 1% of their time to teaching the next builder. At scale, this produces a self-sustaining educational cycle that expands the builder population without institutional overhead.

Builder-Dependent Ratio

The ratio of AI-dependent businesses (those building on externally owned infrastructure) to AI infrastructure builders (those constructing domestically owned infrastructure capacity). Originally introduced in EAD-2026-01; extended here to the Economic Participation Framework.

Infrastructure IPO Wave

The current period (2025-2026) in which the first generation of commercial-scale AI infrastructure companies (CoreWeave, Nscale, Lambda, Mistral) is raising public capital and listing on exchanges, creating the first opportunity for independent builders to participate in infrastructure ownership through retail equity.

11How this research has evolved

Research Timeline

2024

EAD-2026-01 Research Initiated

Q1 2025

CoreWeave IPO Preparation

March 2025

CoreWeave Nasdaq Listing (CRWV)

2025-2026

Open-Weight Model Performance Convergence

March 2026

Nscale USD 2B Raise at USD 14.6B Valuation

April 2026

CNAS Sovereign AI Index Published

Q2 2026

EAD-2026-02 (Judgment Layer Theory) Completed

July 2026

EAD-2026-03 (The Infrastructure Loop) Published

12Download, cite, and track version history

Original Paper

Download

↓ Working Paper EAD-2026-03 (DOCX)

Cite this research

Agarwal, S. (2026). The infrastructure loop: Digital asset ownership, distributed infrastructure participation, and economic resilience. triNetra Research Working Paper EAD-2026-03.

Version history

1.0July 2026Initial publication.

Executive Summary

Who this research serves

The Problem

Research Question

PaaF Methodology Journey

Interactive Framework

Evidence Explorer

Research Dashboard

Commercial Implications

Critical Review

Glossary

Research Timeline

Original Paper

Download

Cite this research

Patterns as a Framework (PaaF)

How We Work

Who We Serve

Eagle Insight Platform – Structural Intelligence Engine

Eagle Insight Platform Documentation: Eagle Framework and RiskOpsBench 1.0 Reference

About triNetra

Research Overview

Frequently Asked Questions

Current Validation Stage

Contact triNetra Research