Comparative Analysis: The DeepMind AGI Study and the UBI–EUBI–UOR Proposal.

Overview

Both proposals begin with the same diagnosis: artificial intelligence may enormously increase productivity, but it may also separate income from employment and economic participation, concentrating the gains among the owners of technological capital.

They differ mainly in their starting question:

• The DeepMind study asks: Which policy is most efficient and politically feasible at each level of labor displacement?

• The UBI–EUBI–UOR proposal asks: Which rights and human capabilities must be preserved, regardless of the course taken by automation?

The study therefore recommends a conditional sequence—unemployment insurance and tax credits, followed by a negative income tax and, only in the event of structural disruption, Universal Basic Capital. The UBI–EUBI–UOR proposal, in contrast, immediately establishes a universal, permanent, and non-withdrawable floor, upon which it builds opportunities for remunerated learning and human development.

My conclusion is that the two proposals are more complementary than mutually exclusive. The DeepMind study offers the UBI–EUBI–UOR proposal a strategy for implementation, financing, and adaptation based on measurable indicators. The UBI–EUBI–UOR proposal, in turn, addresses the study’s greatest omission: What will provide purpose, autonomy, continuous education, and human participation when employment ceases to be the central organizing principle of social life?

---

1. What the DeepMind study proposes

The essay Economic Policy for AGI, by Julian Jacobs and Alex Imas, evaluates eleven policies according to four dimensions:

1. welfare and resilience;

2. agency and voice;

3. feasibility and efficiency;

4. durability across different economic futures.

The analysis combines literature reviews, public-opinion surveys, and evaluations produced by 51 AI agents configured to represent the profiles of 51 real economists. It was not, therefore, a panel in which 51 economists directly assessed each policy. The scores were produced by agents simulating economists and should be interpreted as an exploratory exercise, not as scientific consensus. The authors themselves acknowledge the need for further empirical validation. DeepMind Institute — Economic Policy for AGI

The study divides the future into three scenarios:

Scenario	Characteristics	Recommended response
1. Mild disruption	Task automation and wage polarization, but aggregate demand for labor remains intact	Expanded unemployment insurance, EITC, and employer-led retraining
2. Broad displacement	Prolonged unemployment and wage compression; job creation proceeds more slowly than automation	Transformation of the EITC into a Negative Income Tax—NIT
3. Labor–capital decoupling	Structural decline in labor’s share of income and predominance of capital returns	Universal Basic Capital—UBC—and funds for sharing in AI-generated gains

This is a “least-regret” strategy: rather than immediately implementing the most radical policy, institutions would prepare the necessary instruments in advance and activate them when specified indicators reach predetermined thresholds.

---

2. What the UBI–EUBI–UOR framework proposes

The proposal published on PoutPourri consists of three layers.

UBI: security as a right

Universal Basic Income would be:

• individual;

• paid from birth;

• unconditional;

• permanent;

• never withdrawable;

• independent of employment, educational performance, or behavioral assessment.

Its foundation is not merely poverty relief. It recognizes the right of every citizen to participate in the wealth and productive capacity accumulated by society.

EUBI: learning recognized as work

Educational Universal Basic Income would add differentiated payments for activities involving:

• learning;

• research;

• creation;

• care;

• professional practice;

• community contribution.

Participation would be voluntary, but the additional payment would depend on verifiable effort, progress, and accomplishment. Failure or withdrawal from a particular learning pathway would never result in the loss of UBI.

UOR: preservation of the human capacity to act

Universal Operational Readiness would be neither a monetary benefit nor a final diploma. It would represent the continuous pursuit and development of:

• cognitive;

• technical;

• social;

• emotional;

• ethical;

• creative;

• operational;

• civic capabilities.

The proposal therefore does more than redistribute income. It seeks to prevent automation from creating a population that is materially supported but cognitively dependent, politically powerless, or socially disposable.

---

3. Direct comparison

Dimension	DeepMind study	UBI–EUBI–UOR
Nature	Framework for evaluating policies	Normative social-contract proposal
Starting point	Economic uncertainty and fiscal caution	Dignity and security as rights
Central unit	Worker, income, and participation in capital	Person, learning, and capacity to act
Universality	Depends on the scenario and policy instrument	Present from the beginning
Relationship with employment	Seeks to preserve incentives and attachment to the labor market	Accepts that employment and income may become separated
UBI	Expensive, insufficiently targeted, and unable by itself to create economic participation	Permanent floor of freedom and dignity
Education	Retraining primarily directed toward reemployment	Lifelong learning, including when no market position is available
Meaning and purpose	Derived primarily from work and economic participation	Also derived from learning, care, creation, investigation, and participation
Ownership of AI-generated gains	UBC, sovereign funds, and equity ownership	Public participation in AI returns, although not yet fully specified
Implementation criterion	Economic indicators and triggers	Permanent right, with gradual expansion of its monetary value
Principal strength	Feasibility, sequencing, and adaptability	Integrated human vision and prevention of human irrelevance
Principal weakness	Reduction of the human problem to economic policy	Insufficient fiscal, legal, and administrative detail

---

4. Why does UBI receive relatively low scores in the study?

In the simulated-economist panel, UBI received:

• 51.2 for welfare and resilience;

• 53.1 for agency and democratic participation;

• 57.7 for durability across scenarios;

• only 35.5 for feasibility and implementation.

These results appear unfavorable, but they require careful interpretation.

UBI was evaluated in isolation

The study evaluates UBI as a uniform cash transfer. It does not evaluate the integrated system:

$$$$UBI+EUBI+UOR+capital participation+universal services.\text{UBI}+\text{EUBI}+\text{UOR}+\text{capital participation}+\text{universal services}.

UBI’s low score for “meaning and human value”—41.4—probably reflects the conclusion that money alone cannot replace the social meaning of work.

The UBI–EUBI–UOR proposal explicitly agrees with this. UBI provides time and security, but it does not guarantee learning, community, guidance, or purpose. EUBI and UOR were conceived precisely to fill this gap.

The score attributed to a stand-alone UBI cannot, therefore, be automatically transferred to the complete UBI–EUBI–UOR system.

The study’s conception of agency is strongly economic

The study measures agency primarily through:

• economic participation;

• ownership of AI-generated gains;

• democratic and workplace influence.

This definition favors Universal Basic Capital: holding shares or units in a fund is regarded as direct participation in the automated economy.

Agency, however, also involves the real capacity to:

• understand algorithmic decisions;

• acquire new knowledge;

• contest evaluations;

• participate in civic life;

• create projects;

• refuse abusive employment;

• live without absolute dependence on employers or bureaucracies.

Under this broader conception, the UBI–EUBI–UOR combination may generate more agency than the study’s scores suggest.

---

5. Is the study’s criticism of UBI valid?

It is partially valid.

“UBI is expensive”

Its gross cost would indeed be high:

Gross cost, however, is not the same as net fiscal cost. Part of the expenditure can return through progressive taxation. For high-income recipients, the UBI received could be more than offset by additional taxes.

Universality can also reduce:

• poverty tests;

• exclusion errors;

• stigmatization;

• intrusive monitoring;

• abrupt benefit cliffs;

• incentives to conceal income.

The study is therefore correct to identify the cost, but it does not demonstrate that targeting is always superior once all administrative, political, and human costs are considered.

“UBI can cause inflation”

This depends on:

• the benefit’s value;

• how it is financed;

• idle productive capacity;

• supply elasticity;

• the availability of housing, energy, food, healthcare, and transportation;

• the degree of market concentration.

A UBI financed through taxation of very high incomes redistributes purchasing power rather than merely creating additional net demand. A UBI financed through monetary expansion in an economy already operating near its real limits would present a greater inflationary risk.

The UBI–EUBI–UOR text already incorporates an important caution: creating money does not automatically create houses, electricity, food, or healthcare professionals. Income must grow alongside real productive capacity.

“A Negative Income Tax is more efficient”

An NIT concentrates resources among people with low incomes and costs less than a UBI of equivalent value. This is a genuine advantage.

Nevertheless, it retains several problems:

• the need to verify income;

• delays between income loss and payment;

• administrative errors;

• reduced marginal incentives while the benefit is being withdrawn;

• dependence on tax declarations;

• a permanent distinction between net contributors and beneficiaries.

Under moderate displacement, an NIT may be efficient. Under widespread structural unemployment, however, a large share of the population would become eligible. At that point, NIT and UBI would begin to converge operationally, although UBI would preserve greater legal clarity: the benefit would not depend on periodically demonstrating insufficient income.

---

6. The DeepMind study’s greatest contribution to UBI–EUBI–UOR

The blog proposal correctly argues that the system must begin before the disruption. It must still answer, however: How should it begin, on what scale, and according to which indicators should each component be expanded?

DeepMind’s scenario model provides a possible answer.

Elements that could exist from the beginning

• a modest and constitutionally protected UBI;

• universal basic services;

• an initial EUBI structure;

• digital and local learning centers;

• UOR as an indicator of social opportunities and capabilities;

• a public fund for participation in technological capital.

Elements that could be adjusted through triggers

Observed indicator	Possible adjustment
Rising temporary unemployment	Expansion of unemployment insurance
Persistent wage compression	Increase in UBI and an NIT-like supplement
Automation proceeding faster than reemployment	Major expansion of EUBI
Decline in labor’s share of GDP	Greater taxation of, or public participation in, capital gains
Simultaneous growth in productivity and structural unemployment	Expansion of UBI, UBS, and social dividends
Accelerating concentration of technological ownership	Expansion of UBC or an AI sovereign wealth fund
Declining cognitive and civic autonomy	Strengthening of programs associated with UOR

The important distinction would be this: indicators could alter the value and scale of UBI, but never eliminate its existence as a right.

---

7. What UBI–EUBI–UOR adds to the study

A more complete response to the problem of purpose

The study recognizes that work possesses psychological and social value, but it remains strongly attached to traditional employment. The EUBI–UOR system expands the concept of contribution to include:

• study;

• care;

• cultural production;

• independent research;

• teaching;

• community participation;

• environmental preservation;

• democratic supervision of AI.

This becomes crucial if the market productivity of human labor falls while the human need for belonging, recognition, and development remains.

Separation between dignity and merit

The proposal’s strongest principle is:

Excellence deserves additional recognition; dignity does not depend on excellence.

This separation prevents the educational system from becoming a form of workfare in which citizens must prove their productivity to preserve the means of survival.

Protection against the algorithmic classification of the population

In a system based exclusively on retraining and “employability,” algorithms could decide who deserves investment according to predicted chances of success. UBI creates a barrier against this utilitarianism: even those who do not present a high “expected return” retain material security.

A richer conception of freedom

Owning capital is important, but insufficient. A person may receive dividends while remaining unable to understand or contest the systems governing their life. UOR adds cognitive, ethical, operational, and civic capability to economic freedom.

---

8. What needs to be strengthened in the UBI–EUBI–UOR proposal

The comparison reveals five areas requiring further development.

1. Ownership of automation-generated gains

The expression “public participation in AI returns” must be transformed into concrete institutions:

• an AI sovereign wealth fund;

• public equity ownership;

• royalties on digital infrastructure;

• progressive taxation of extraordinary profits;

• data dividends;

• individual, collective, or hybrid ownership units.

Here, DeepMind’s UBC should be incorporated as a patrimonial component of the proposal, not as a substitute for UBI.

2. Country-specific financing

Feasibility depends heavily on fiscal institutions, informality, currency arrangements, demographic structure, and productive capacity. A solution designed for the United States cannot be mechanically transferred to Brazil.

3. Governance of EUBI

It will be necessary to specify:

• who accredits activities;

• how progress is measured;

• how fraud and credential inflation are prevented;

• how the right of appeal is guaranteed;

• how privacy is protected;

• how political favoritism is avoided;

• how different capabilities and living conditions are recognized.

4. Limits on UOR measurement

UOR must not become a “social credit” system. Its indicators should evaluate collective conditions and barriers, not classify the moral worth of individuals. A low score should lead to additional support, never the withdrawal of rights.

5. The relationship between UBI and universal services

Money cannot adequately replace healthcare, education, infrastructure, connectivity, housing, and energy when their supply is limited or dominated by concentrated markets. The architecture should explicitly state:

$$$$Real security=UBI+UBS+capital participation+EUBI–UOR.\text{Real security} = \text{UBI} + \text{UBS} + \text{capital participation} + \text{EUBI–UOR}.

---

9. A synthesis superior to either proposal in isolation

The most robust architecture would consist of five pillars:

1. Permanent UBI: a modest, universal, and never-withdrawable monetary floor.

2. Universal Basic Services: healthcare, education, connectivity, transportation, and essential infrastructure.

3. EUBI–UOR: voluntary and remunerated pathways for learning, creation, care, and participation.

4. UBC or a social AI fund: collective and individual participation in the returns from automated capital.

5. Adaptive stabilizers: unemployment insurance, income supplements, and sector-specific policies activated by measurable indicators.

This synthesis preserves the best aspects of both models:

• from DeepMind: prudence, measurement, scenarios, triggers, and capital ownership;

• from UBI–EUBI–UOR: universal rights, advance prevention, lifelong learning, and a comprehensive conception of human agency.

Conclusion

The DeepMind study is stronger as an economic transition strategy. The UBI–EUBI–UOR proposal is stronger as a civilizational project.

DeepMind asks when each instrument should be activated. UBI–EUBI–UOR asks what kind of society we want to preserve. We do not need to choose between them.

The decisive improvement would be to reinterpret the proposal as follows:

UBI constitutes the permanent right; EUBI–UOR preserves learning, purpose, and agency; universal services guarantee access to real goods; and Universal Basic Capital distributes ownership of the gains from automation. Economic indicators regulate the speed and scale of expansion, but not the existence of fundamental rights.

Formulated in this way, the UBI–EUBI–UOR proposal directly addresses the study’s principal criticisms while also going beyond it. It seeks not only to prevent poverty during automation, but also to prevent a technologically abundant society from producing human beings who are economically maintained yet politically irrelevant.

ChatGPT/PoutPourri.