Abundant Intelligence, Broken Routing
A Giant HAPI Research Essay on Citrini Research’s “The 2028 Global Intelligence Crisis”
With a sector by sector friction rent map, early warning indicators, policy memos for key audiences, and a legitimacy first communications plan
Executive summary
Citrini’s scenario is not “AI takes jobs” in the usual way. It is “AI rewires the economic circulation system.” Output can rise while household income claims fall. If households lose predictable purchasing power, then consumption weakens, credit models built on stable incomes crack, and wage based tax receipts decline just as stabilization demands rise. The scenario’s engine is reflexivity: firms cut labor, fund AI, AI improves, more labor is cut, demand softens, firms cut more, and the loop lacks a natural brake.
HAPI makes the scenario tractable. The system is cognitively aware but institutionally slow. Behavioral agility exists inside firms but can become a tragedy of the commons. Emotional and social adaptability are the brittle points that turn a transition into a legitimacy crisis.
The most realistic route to “10 out of 10” adaptability is not a sweeping redesign. It is a minimal set of high leverage upgrades that repair routing, damp demand collapse, prevent sudden recognition events in credit, and create accountable markets around agentic systems. In practical terms: (1) an earnings shock stabilizer, (2) a small universal dividend rail, (3) credit and mortgage underwriting modernization for income volatility, and (4) national assurance and liability standards for agentic systems. These four do not stop AI. They stop the spiral.
This essay provides the full diagnostic, the counter loops Citrini underweights, the sector exposure map, the leading indicators, and three tailored policy memos for government, investors, and AI labs, plus a communications strategy designed to protect legitimacy while reforms move.
Table of contents
- The core claim: routing failure, not just automation
- The system in one diagram, described in words
- HAPI assessment across five dimensions
- Counter loops and stabilizers Citrini underweights
- Sector by sector friction rent map
- Early warning indicators with thresholds
- Minimalistic path to 10 out of 10 adaptability
- Policy memo to government and regulators
- Policy memo to investors and boards
- Policy memo to AI labs and hyperscalers
- Legitimacy first communications strategy
- Appendix: scenario stress tests and implementation notes
1. The core claim: routing failure, not just automation
Most “AI disruption” narratives focus on substitution: machines do tasks humans did, jobs shift, new jobs appear. Citrini’s scenario focuses on circulation: who receives claims on output, how those claims translate into spending, and whether the demand base that sustains the consumer economy, credit quality, and tax receipts remains intact.
The scenario’s phrase “Ghost GDP” is a rhetorical wrapper for a simple macro idea: measured output can be high while household income claims are low. If households do not receive claim checks, they cannot spend. If they cannot spend, firms that sell to humans see revenue pressure. If those firms respond by cutting labor and buying more AI to protect margins, they intensify the original mechanism. A loop emerges where rational micro behavior produces fragile macro outcomes.
This framing matters because it identifies a failure mode that standard policy tools struggle with. Rate cuts can reduce financing costs. They cannot create household purchasing power if the income channel is structurally impaired. Quantitative easing can support asset prices. It cannot restore legitimacy if the public sees gains concentrating while job identity collapses.
Citrini’s scenario is ultimately a test of adaptability under three simultaneous transitions:
- A labor market transition in high earning cognitive roles
- A credit transition in underwriting assumptions tied to career stability
- A governance transition in tax bases and stabilizers tied to wage income
HAPI is designed to evaluate exactly this kind of multi system adaptation problem.
2. The system in one diagram, described in words
Think of the economy as a loop of claims and spending:
- Firms deploy AI and reduce payroll costs
- Payroll savings fund more AI adoption
- AI improves capability, lowering the cost of substitution
- High earning households lose income or face higher income volatility
- Precautionary savings rises among employed households
- Discretionary consumption falls with a lag but at scale
- Consumer facing firms experience margin pressure and adopt more AI
- Credit models that assumed stable income streams begin to reprice risk
- Mortgage and private credit stress triggers financial tightening
- Governments face falling wage based receipts and rising outlays
- Political conflict slows response, weakening trust
- Trust loss increases precaution and instability, feeding back into demand
There are two kinds of loops here:
- Real economy loop: payroll substitution, demand decline, more substitution
- Financial loop: income impairment, credit recognition, tightening, wealth effect decline, more demand decline
Citrini’s scenario assumes these loops reinforce each other faster than institutions can respond.
3. HAPI assessment across five dimensions
For each dimension, I will go slow and do four passes: good, bad, opportunity, threat.
3.1 Cognitive adaptability
Good
Citrini’s most valuable contribution is cognitive reframing. It asks: what if AI bullishness is bearish because productivity gains bypass the wage channel? That is a strong model update. It also correctly highlights reflexivity, where each firm’s rational response strengthens the collective mechanism. This is real systems thinking.
It also captures a subtle but powerful idea about bargaining and belief. You do not need perfect agentic capability to compress SaaS pricing. Procurement needs a credible alternative and enough tolerance for transition pain. The belief itself becomes economically causal.
Finally, the scenario identifies friction rents as a macro vulnerability. A large share of value capture depends on human attention limits and inertia. Agents can attack that by lowering search and negotiation costs, pushing rents toward zero.
Bad
The scenario sometimes treats adoption as smoother than it will be. Enterprise replacement is constrained by integration debt, audit and compliance requirements, liability, exception handling, and organizational inertia. These frictions do not stop disruption, but they alter timing and create unevenness. The macro path depends on timing.
It also treats “white collar” as too homogeneous. Tasks vary in substitutability. Some work is language mediated and pattern heavy. Other work is responsibility heavy, trust heavy, regulated, and anchored in accountability. That work tends to shift, not vanish.
Finally, the scenario leans heavily on a single dominant loop. Real economies often generate counter loops. New demand emerges when costs fall. New aggregation layers appear when old intermediaries collapse. Regulation and insurance markets can reintroduce friction as assurance rather than rent.
Opportunity
Used correctly, Citrini’s cognitive frame helps leaders focus on routing. It shifts attention from “how smart is AI” to “how do claims flow.” It motivates policy and corporate designs that stabilize demand while allowing adoption.
It also suggests an investment and strategy framework: map revenue streams by dependence on inertia, information asymmetry, and attention capture. Those moats degrade in an agent mediated economy. Firms can reposition toward machine readable differentiation: verified quality, reliability guarantees, transparent terms, and auditable service performance.
Threat
The main cognitive threat is a timing trap. High earning households have buffers, so data may look stable while fear rises. Then spending behavior flips, and the downturn appears suddenly. Leaders misread the early phase as contained.
A second threat is clinging to historical job creation analogies without noticing the key difference: new tasks can be done by agents too, reducing labor intensity. That does not mean no jobs. It means wage based routing might shrink faster than institutions adjust.
3.2 Emotional adaptability
Good
Citrini captures the behavioral psychology of fear: employed people spend as if they might be next. That is emotionally accurate and macro relevant. It also recognizes identity shock. High status job loss is not only income loss. It is status and identity disruption, which increases political volatility and reduces tolerance for uncertainty.
The scenario also anticipates moral anger and legitimacy collapse when gains appear concentrated among compute owners while communities face insecurity.
Bad
The scenario underweights emotional stabilization mechanisms. Humans normalize rapidly if two conditions exist: predictable backstops and a believable future narrative. Emotional adaptability is not only individual temperament. It is policy and institutional architecture.
It also underweights how blame mutates. Anger may not remain focused on AI labs. It can diffuse into scapegoating and polarization that damages coordination more broadly.
Finally, it underweights quiet failure modes: shame, withdrawal, depression, reduced civic engagement. These erode the very capacity required to build coalitions for adaptive policy.
Opportunity
Emotional stability can be improved faster than industrial capacity. The high leverage levers are predictability and dignity.
Predictability comes from automatic stabilizers tied to earnings shocks, not only unemployment. Dignity comes from supports that feel like rights, such as dividends and portable benefits, rather than stigmatized assistance.
Corporate practices matter too. Transparent redeployment, credible reskilling tied to real roles, and profit sharing linked to AI productivity can reduce adversarial sentiment and protect trust.
Threat
Low emotional adaptability forms its own loop: fear drives precautionary savings, precaution reduces demand, demand decline drives layoffs, layoffs confirm fear. That loop can outrun traditional macro tools. It also increases the risk of policy whiplash, which raises uncertainty and suppresses investment.
3.3 Behavioral adaptability
Good
The scenario is strong on micro rationality leading to macro harm. Incumbents adopt aggressively because they must. This differs from older disruption patterns. It also highlights mechanical linkages like seat based SaaS revenues falling when customer headcount falls. That is arithmetic, not strategy, and it is a clean propagation channel.
It also accurately describes household adaptation: downshifting, gig work absorption, wage compression in services due to overqualified labor supply, and lagged but large spending declines among high earners.
The “OpEx substitution” point is key: AI spend can rise while total spend falls, so weakening demand does not automatically slow AI adoption.
Bad
The scenario underweights organizational friction and accountability costs. Firms are coalitions with risk controls and liability constraints. Adoption can be slower, or it can be chaotic and failure prone, generating backlash that changes the trajectory.
It also underweights the possibility that revenue becomes the binding constraint, forcing a shift from cost cutting to new product creation. AI can enable that too.
Finally, it underweights re hiring and job creation in assurance functions. As automation increases, audit, compliance, safety, and governance costs often rise. Humans return as accountability.
Opportunity
Small incentive changes can redirect behavior. KPI design is a hidden policy lever inside firms. Reward only margin expansion and behavior becomes labor deletion. Reward verified reliability, customer trust, and new revenue creation, and behavior shifts toward hybrid deployment and product innovation.
At the system level, an earnings shock stabilizer reduces the incentive to cut by stabilizing demand. This is a macro behavioral lever that works through millions of independent decisions.
Threat
The threat is a tragedy of the commons. Each firm cuts to survive. Collectively, demand collapses. No actor is irrational, but the outcome is unstable.
Another threat is cultural contagion. Layoffs become the default corporate playbook. Boards expect it. Markets reward it. Managers benchmark it. The system overshoots.
3.4 Social adaptability
Good
The scenario’s strongest social insight is that time is the villain. AI evolves faster than institutions update. Policy lag becomes an accelerant. The scenario also correctly predicts ideological conflict around redistribution, compute taxation, and industrial policy.
It emphasizes legitimacy risk: if gains accrue to compute owners while households lose bargaining power, trust collapses. Trust is macro infrastructure.
It also highlights fiscal routing: wage tied receipts fall while outlays rise. If true, that compresses state capacity exactly when it is needed.
Bad
The scenario assumes government is slow and therefore fails. Governments can be slow, but coalitions can form abruptly when influential groups feel pain and when a legible policy tool exists.
It also underweights private governance. Insurance requirements, procurement standards, and industry assurance norms can shape deployment and create accountable labor markets even when legislatures stall.
Finally, it underweights international competition effects. If one jurisdiction stabilizes routing and legitimacy, it can attract talent and capital, reshaping domestic politics elsewhere.
Opportunity
The opportunity is to adopt policies that are legible, automatic, and hard to capture. Earnings shock stabilizers, portable benefits, dividend rails, and assurance standards tied to procurement and insurance can stabilize the transition without micromanaging markets.
Narrative alignment is also an opportunity. Frame the project as upgrading the rails of prosperity, not punishing technology. That increases coalition viability.
Threat
The social threat is policy whiplash. Unstable policy is more damaging than slow policy. Whiplash increases uncertainty, reduces investment, and deepens mistrust.
At worst, the system becomes brittle. Small shocks cause disproportionate political and financial reactions.
3.5 Growth potential
Good
The scenario ends with the right meta claim: repricing is not collapse. Abundant intelligence can raise welfare dramatically if claims are routed to households and legitimacy is maintained. The scenario also correctly notes that institutions were built around scarce human cognition, and new frameworks must be built when that assumption changes.
Bad
The scenario underweights how lower costs can create new demand if household purchasing power is stabilized. It also underweights physical constraints as job creation channels: energy, grid, data centers, housing, logistics, and care systems.
It tends to treat abundance as contractionary. Abundance is contractionary only when claims do not reach households.
Opportunity
A high growth potential path looks like this: household purchasing power is stabilized, agentic systems are trustworthy and auditable, education becomes modular and job linked, and investment shifts into physical infrastructure and care sectors. In that world, abundant intelligence reduces cost of living, improves health and safety, accelerates science, and increases leisure without breaking legitimacy.
Threat
The worst equilibrium is high output with low welfare. GDP can rise while household security falls, trust collapses, politics becomes punitive, and innovation becomes socially illegitimate. That is a legitimacy crash, not a technology crash.
4. Counter loops and stabilizers Citrini underweights
Citrini’s scenario is compelling because it has a clean engine. But real systems usually produce stabilizers. Here are the most important counter loops that could bend the curve, plus how they could fail.
4.1 Liability and assurance markets
As agents transact, code, advise, and act, liability becomes a binding constraint. That creates demand for:
- audit trails and logging
- third party verification
- model risk management
- incident response
- regulatory compliance and safety engineering
This can create meaningful employment channels and slow reckless substitution.
Failure mode: if assurance is treated as optional or is captured by incumbents, trust collapses after accidents, leading to policy whiplash.
4.2 Human legitimacy premium and “authenticity markets”
Some categories do not commoditize to pure price optimization because consumers want human connection, social meaning, or legitimacy. Examples: care, education, therapy, live experiences, community anchored services, and premium craftsmanship.
Failure mode: if households lack purchasing power, legitimacy markets shrink into luxury niches, increasing inequality and resentment.
4.3 Physical world constraints as labor sinks
Even in a world of abundant cognition, the physical world remains constrained: energy, housing, transport, grid buildout, and maintenance. These are labor intensive and can absorb workers if pathways exist.
Failure mode: if training and mobility pathways are weak, displaced workers cannot transition, and physical sectors face bottlenecks rather than absorption.
4.4 New products and demand expansion
When costs drop, new products become feasible. If households keep purchasing power, demand can expand and absorb labor in design, operations, and physical execution.
Failure mode: if the wage channel collapses, cost drops do not translate into demand expansion, and the economy becomes high output but low circulation.
4.5 Policy can move abruptly
Institutional adaptation is not always smooth. It often looks like stasis, then sudden action once a coalition forms.
Failure mode: action arrives as punitive backlash rather than stable design, creating whiplash.
5. Sector by sector friction rent map
This map answers: where does value capture depend on human attention limits, inertia, information asymmetry, and switching costs, and therefore where are rents most exposed to agentic optimization?
I will classify each sector by four variables:
- Friction rent dependence: how much revenue depends on human limitations
- Accountability constraint: how much responsibility and liability require human oversight
- Physical coupling: how much delivery requires real world execution
- Likely adaptation path: commoditize, re bundle, or become assurance anchored
5.1 High exposure: habitual intermediation and inertia rents
Subscription renewals and memberships
- Friction rent dependence: very high
- Accountability constraint: low to medium
- Physical coupling: low
- Likely path: agents negotiate, churn rises, pricing becomes usage based or outcome based
Travel booking and aggregation
- Friction rent dependence: high
- Accountability constraint: medium
- Physical coupling: low
- Likely path: re bundling into insurance, disruption management, and concierge level accountability
Price comparison vulnerable retail marketplaces
- Friction rent dependence: high
- Accountability constraint: low
- Physical coupling: medium
- Likely path: shift from attention capture to fulfillment reliability, warranties, verified authenticity, and logistics quality
Basic financial advice and routine tax prep
- Friction rent dependence: high
- Accountability constraint: high because of liability
- Physical coupling: low
- Likely path: automation plus human licensed sign off, with firms competing on auditability and indemnification
5.2 Medium exposure: information asymmetry rents with liability backstops
Insurance distribution and renewals
- Friction rent dependence: medium to high
- Accountability constraint: high
- Physical coupling: low
- Likely path: agents reshop, commissions compress, insurers compete on transparent pricing and service. Brokers shift to complex cases and claims advocacy
Real estate brokerage
- Friction rent dependence: medium to high
- Accountability constraint: medium
- Physical coupling: high
- Likely path: commission compression, rise of flat fee, rise of AI assisted self service plus human closers for liability and negotiation
B2B SaaS with seat based pricing
- Friction rent dependence: medium
- Accountability constraint: medium
- Physical coupling: low
- Likely path: pricing shifts from seats to outcomes, incumbents bundle compliance, security, integrations, and liability guarantees
5.3 Lower exposure: physical delivery, regulated accountability, and scarce real world capacity
Healthcare delivery and elder care
- Friction rent dependence: medium
- Accountability constraint: very high
- Physical coupling: very high
- Likely path: AI augmentation, documentation automation, triage optimization. Human demand remains anchored. Constraint becomes workforce supply and training
Construction, trades, infrastructure maintenance
- Friction rent dependence: low
- Accountability constraint: high
- Physical coupling: very high
- Likely path: demand rises if policy invests in physical buildout. Training pipelines determine absorption capacity
Utilities and grid operators
- Friction rent dependence: low
- Accountability constraint: very high
- Physical coupling: very high
- Likely path: AI helps optimization and planning, but workforce remains. Bottleneck is permitting and capital
5.4 High exposure inside firms: middle layer coordination and process work
These are not sectors but functions:
- procurement, reporting, basic analytics, content drafting, routine legal review, customer support tier 1 and tier 2, project management layers that exist to move information
In these areas, substitution can be rapid and employment can fall quickly, but new roles appear in:
- oversight, escalation handling, audit, toolchain governance, vendor management, data stewardship, and incident response
5.5 Payments and financial rails
Citrini suggests agents route around card interchange via stablecoins and cheaper rails. Whether the exact technology wins is less important than the directional risk: if commerce becomes machine to machine, price sensitivity rises and tolls are targeted.
Exposure varies:
- Networks dependent on interchange: higher exposure
- Networks positioned as infrastructure for new rails: lower exposure
- Banks reliant on rewards funded by interchange: higher exposure
- Entities offering compliance, fraud prevention, identity, and settlement assurance: better positioned
Key HAPI point: the moat shifts from habit and brand to trust, compliance, and reliability.
6. Early warning indicators with thresholds
These indicators are designed to detect the scenario’s engines early. They are grouped into real economy routing, labor and fear, credit recognition, fiscal capacity, and legitimacy.
6.1 Real economy routing indicators
| Indicator | Why it matters | Watch level | Red flag |
|---|---|---|---|
| Real wage growth for top income quintiles | High earners drive discretionary spending | Below 0% for 2 quarters | Below negative 2% for 2 quarters |
| Personal saving rate among high income households | Precautionary shift precedes demand collapse | Rising 1 to 2 points | Rising 3 points quickly |
| Discretionary spend in high end categories | Early signal of high earner pullback | Flat to down | Down 5% year over year |
6.2 Labor substitution and fear indicators
| Indicator | Why it matters | Watch level | Red flag |
|---|---|---|---|
| White collar job openings index | Forward looking labor demand | Down 15% year over year | Down 25% year over year |
| Underemployment among college educated | Downshift signal | Rising steadily | Rising sharply over 2 quarters |
| Wage compression in services | Overqualified labor supply flooding | Flat wages | Wages down plus hours down |
6.3 Credit recognition indicators
| Indicator | Why it matters | Watch level | Red flag |
|---|---|---|---|
| Early stage delinquencies in high FICO metros | Prime stability assumption | Up modestly | Up sharply in tech and finance ZIPs |
| HELOC draws and 401k hardship withdrawals | Hidden stress before delinquencies | Rising | Rising with flat mortgage delinquencies |
| Private credit marks versus public comps | Recognition lag risk | Persistent gap | Gap widens while defaults rise |
| Insurer RBC pressure and asset reclassification risk | Forced selling trigger | Early regulatory scrutiny | RBC tightening plus downgrades |
6.4 Fiscal capacity indicators
| Indicator | Why it matters | Watch level | Red flag |
|---|---|---|---|
| Payroll tax receipts versus baseline | Wage channel health | Below baseline | 10% or more below baseline |
| State and municipal revenue dispersion | Localized stress | Spreads widen | Specific metros show distress pricing |
6.5 Legitimacy and social stability indicators
| Indicator | Why it matters | Watch level | Red flag |
|---|---|---|---|
| Trust in institutions surveys | Coordination capacity | Declining | Collapse in trend plus polarization spike |
| Protest intensity and labor actions in tech hubs | Legitimacy stress | Sporadic | Persistent and growing |
| Policy volatility | Investment deterrent | High rhetoric | Rapid regulatory swings and emergency measures |
These are not predictions. They are gauges designed to detect the spiral early enough to intervene.
7. Minimalistic path to 10 out of 10 adaptability
A 10 out of 10 world is not one without disruption. It is one where disruption does not trigger a demand collapse, a credit recognition cascade, and a legitimacy crisis.
The minimal plan must target the three accelerants: demand collapse, recognition shocks, and legitimacy failure. The smallest set of interventions that covers all three is four upgrades.
7.1 Upgrade 1: Earnings shock stabilizer
What it is
Automatic temporary support triggered by verified earnings declines, regardless of whether the person is technically unemployed.
Why it works
- Stabilizes demand when high earners experience income impairment
- Reduces precautionary savings among employed households
- Lowers mortgage stress by smoothing income paths
Minimal design
- Trigger based on year over year earnings decline exceeding a threshold
- Benefit formula that partially replaces lost earnings for a limited duration
- Phase out as earnings recover
- Built on payroll and tax data already collected
HAPI impact
Raises emotional adaptability, behavioral adaptability, and social stability quickly.
7.2 Upgrade 2: A small universal dividend rail
What it is
A modest per capita monthly dividend funded by broad bases tied to concentrated value capture.
Why it works
- Restores circular flow so output becomes spendable welfare
- Makes gains visible to households, reducing backlash
- Acts as automatic stabilization without means testing stigma
Minimal design
- Start small and scale gradually
- Fund from broad sources such as ultra normal profits, platform rents, or other broad value proxies
- Distribute universally to preserve legitimacy and simplify administration
HAPI impact
Raises social adaptability and growth potential by protecting legitimacy.
7.3 Upgrade 3: Credit and mortgage underwriting modernization for income volatility
What it is
Update underwriting and capital rules to incorporate income stability risk, not only credit score and historical employment.
Why it works
- Prevents sudden recognition events by pricing risk earlier and more smoothly
- Reduces forced selling dynamics
- Aligns long duration liabilities with realistic income volatility
Minimal design
- Add an income volatility factor into underwriting
- Enhance disclosures and buffers for highly volatile income profiles
- Phase in regulatory standards to avoid sudden tightening
HAPI impact
Raises systemic stability and reduces financial accelerants.
7.4 Upgrade 4: National assurance and liability framework for agentic systems
What it is
Procurement and insurance linked standards for auditability, logging, accountability, and human oversight for high consequence agentic deployments.
Why it works
- Creates an assurance labor market: audit, compliance, incident response, governance
- Improves trust and reduces catastrophic failures that trigger backlash
- Slows reckless substitution without blocking innovation
Minimal design
- Start with a tiered standard: low consequence, medium consequence, high consequence
- Require logging, audit trails, and responsibility assignment in higher tiers
- Tie compliance to procurement eligibility and insurance coverage
HAPI impact
Raises cognitive and social adaptability and supports sustainable adoption.
Together, these four upgrades are small in number and large in leverage. They do not pick winners. They repair the rails that route prosperity and trust.
8. Policy memo to government and regulators
Subject: Stabilize routing, prevent recognition cascades, preserve legitimacy
Problem statement
AI driven labor substitution can weaken household purchasing power faster than institutions can adapt. The macro risk is a reflexive demand collapse, compounded by credit recognition shocks in mortgages and opaque private credit, and amplified by legitimacy loss.
Objectives
- Keep household demand stable during transition
- Prevent sudden recognition events that force deleveraging
- Maintain public trust through visible shared gains
- Enable safe AI adoption with accountability
Recommended actions, in priority order
- Implement an earnings shock stabilizer
- Trigger on verified earnings drops
- Temporary, formula based, automatic
- Goal: damp demand collapse and fear
- Establish a small universal dividend rail
- Start at modest levels
- Fund through broad bases tied to concentrated value capture
- Goal: restore circular flow and legitimacy
- Modernize credit and mortgage standards for income volatility
- Require lenders to incorporate income stability metrics
- Update capital treatment to reflect volatility risk without causing sudden tightening
- Goal: smooth repricing, avoid panic recognition
- Create national assurance standards for agentic systems
- Tiered requirements based on consequence
- Tie to procurement eligibility and insurance
- Goal: trust and accountability markets, reduce backlash risk
Implementation principles
- Automatic triggers reduce political delay
- Universality improves legitimacy and reduces stigma
- Simple formulas reduce capture and administrative drag
- Phased deployment reduces abrupt tightening
Messaging guidance for government
- Frame as “upgrading the rails of prosperity”
- Emphasize stability, fairness, and innovation at the same time
- Avoid framing as punishment of technology
- Make household benefits visible and predictable
9. Policy memo to investors and boards
Subject: Reprice friction rents, underwrite household routing risk, build for agent customers
Problem statement
AI does not need to fully replace systems to compress pricing power. Belief effects shift bargaining. Agentic optimization threatens friction rents across intermediation layers. The largest macro risk is impaired household routing, which hits discretionary demand and credit stability.
Objectives
- Identify exposure to inertia and intermediation rents
- Stress test portfolios for household income routing risk
- Build strategies for an agent mediated demand surface
- Avoid collective behavior that deepens demand collapse
Recommended actions
- Conduct a friction rent exposure audit
Map each revenue stream by dependence on:
- consumer inertia
- information asymmetry
- search and negotiation costs
- switching costs based on habit
Reclassify companies by resilience:
- outcome and reliability moats
- compliance and liability moats
- physical execution moats
- pure attention and friction moats
- Add a household routing factor to macro and credit risk models
- Monitor income volatility in high spending cohorts
- Stress test discretionary categories on high earner spend declines
- Stress test mortgage and private credit holdings under income impairment scenarios
- Reevaluate pricing architectures in SaaS and services exposure
- Seat based pricing is fragile under headcount reduction
- Outcome based pricing with assurances and liability coverage is more resilient
- Integration and compliance bundling becomes a moat
- Build for agent customers
Agents optimize for price, reliability, and terms. They do not feel brand nostalgia.
Winning strategies:
- machine readable terms
- verified quality metrics
- transparent warranties and returns
- fast resolution and clear escalation paths
- auditable reputation systems
- Board level governance: discourage pure margin playbooks in a weak demand regime
When demand weakens, pure labor deletion can accelerate revenue decline. Reward:
- new product creation
- customer trust metrics
- reliability and safety
- hybrid deployment that preserves accountability
Messaging guidance for investors
- Do not treat this as doomerism. Treat it as a routing and bargaining regime change.
- Distinguish productivity growth from household purchasing power.
- Prioritize trust and liability infrastructure as investable themes.
10. Policy memo to AI labs and hyperscalers
Subject: Earn legitimacy through trust, visible shared benefits, and assurance infrastructure
Problem statement
Even if AI improves welfare in the long run, legitimacy can collapse in the short run if gains concentrate and displacement is visible. Legitimacy loss triggers regulatory whiplash and slows adoption, harming everyone. Time and trust are the binding constraints.
Objectives
- Reduce catastrophic incidents and build trust
- Support assurance labor markets and accountability
- Make shared benefits visible
- Avoid becoming the symbolic villain
Recommended actions
- Voluntary assurance commitments ahead of regulation
- Tiered safety and audit standards for agentic systems
- Logging, audit trails, incident reporting
- Clear responsibility assignments for high consequence deployments
- Support an assurance ecosystem
- Fund training and certification for AI auditors and incident responders
- Open tooling for auditability and monitoring
- Partner with insurers and standards bodies
- Build shared benefit mechanisms that are concrete
- Workforce transition funds that pay for modular training and job placement
- Community benefit agreements in data center regions
- Transparent reporting on productivity gains and reinvestment
- Adopt legitimacy first communications
- Avoid triumphalist rhetoric about replacing humans
- Emphasize augmentation, safety, and shared prosperity
- Acknowledge disruption without minimizing it
- Highlight specific household benefits enabled by AI
- Reduce incentives for reckless deployment by customers
- Provide deployment playbooks that include governance and oversight
- Make high consequence automation require stronger guardrails
- Encourage human in the loop for critical decisions where liability is unclear
Messaging guidance for labs
- The message is not “trust us.” It is “verify us.”
- Make safety and auditability part of the product, not a press release.
11. Legitimacy first communications strategy
This section treats communications as a macro stabilizer. If legitimacy collapses, policy whiplash and social instability follow.
11.1 The narrative frame
Use a single consistent frame across stakeholders:
“We are upgrading the rails that route prosperity in an era of abundant intelligence.”
That frame avoids war metaphors. It avoids scapegoating. It identifies the real problem: routing.
11.2 Three messages that must be true and visible
- Households will not be abandoned during transition
This is communicated by automatic stabilizers and visible dividends. - AI deployment is accountable and auditable
This is communicated by assurance standards and incident transparency. - Gains are shared in a predictable, boring way
This is communicated by dividend rails and community investments.
11.3 Message architecture by audience
For households
- Predictability: “If your income drops, support turns on automatically.”
- Dignity: “This is a dividend and a transition contract, not charity.”
- Agency: “Clear pathways to new roles with real job placement.”
For businesses
- Stability: “Demand stabilization prevents a race to the bottom.”
- Clarity: “Assurance standards reduce liability and uncertainty.”
- Incentives: “Outcome based competition replaces friction based rents.”
For investors
- Risk framing: “This is a routing and recognition risk, not only a tech cycle.”
- Opportunity framing: “Assurance, reliability, and infrastructure are the new moats.”
- Governance framing: “Avoid collective playbooks that deepen demand collapse.”
For AI labs
- Trust: “Verify us. Audit us.”
- Responsibility: “We build with accountability.”
- Shared gains: “We contribute to the transition capacity.”
11.4 Tactical communications plan
- Publish a “Transition Scoreboard” monthly
Not GDP. Not stock indices. A routing dashboard:
- earnings volatility measures
- discretionary spend proxies
- early delinquencies in high income metros
- wage based receipts trends
- uptake of assurance standards
- number of people transitioned into assurance and physical build roles
- Prebunk common polarizing narratives
Examples:
- “This is socialism” versus “this is corporate capture”
Prebunk by emphasizing automatic, simple, broad based, and innovation compatible design.
- Use credible messengers
- local leaders in affected metros
- small business owners
- nonpartisan economists
- workforce trainers and unions
- insurers and risk leaders
Credibility is more important than polish.
- Make household benefits arrive first
If benefits arrive after disruption, legitimacy is already lost. Order matters.
12. Appendix: scenario stress tests and implementation notes
12.1 Stress test: what if AI adoption is slower than Citrini assumes?
Even slower adoption can still compress bargaining and pricing via belief effects. But slower adoption buys time for institutions to upgrade routing. In that world, the minimal plan still helps. It reduces fear and builds legitimacy before shocks.
12.2 Stress test: what if AI adoption is faster and more chaotic?
Then assurance standards become even more critical. Without them, accidents and scandals trigger backlash and policy whiplash. The dividend rail and earnings stabilizer become legitimacy insurance.
12.3 Stress test: what if the biggest damage is not unemployment but wage compression?
Then unemployment based supports miss the problem. Earnings shock stabilization becomes the correct tool. Credit models must price wage volatility, not only job loss.
12.4 Implementation notes on keeping the plan minimal
The enemy of minimalism is capture and complexity. To keep this plan minimal:
- Use automatic triggers
- Use universal distribution when possible
- Use tiered standards instead of bespoke rules
- Phase in underwriting changes to avoid sudden tightening
- Publish simple dashboards to keep accountability public
Final synthesis
Citrini’s scenario is valuable because it highlights a plausible macro failure mode: abundant intelligence can weaken wage routed circulation, creating “productive” output without stable household purchasing power. The reflexive loop is credible. The most brittle points are not technical capability. They are emotional and social adaptability, meaning fear, legitimacy, and coordination speed.
The path to 10 out of 10 is not grand and not ideological. It is four boring upgrades that directly attack the spiral’s accelerants: stabilize earnings shocks, establish a small universal dividend rail, modernize credit underwriting for income volatility, and create assurance and liability standards for agentic systems.
Those upgrades buy time, preserve demand, prevent recognition cascades, and keep trust intact. That is how a society converts abundant intelligence into broad welfare rather than a routing crisis.
