Public AI and the Relationship Economy: Public Economics for the AI Era Connecting Government, Healthcare, and Education

🎥 The YouTube version is also available:

The Relationship Economy: How AI Is Transforming Value, Trust, and Business

Books: Relationship Economy Designing the Economy of the Al Era Chinoba Economics: A Practical Guide: A New Economics Where Knowledge, Decisions, and Trust Create Value

Introduction

The use of AI is not limited to improving corporate efficiency or marketing.

Government procedures.

Healthcare.

Long-term care.

Education.

Transportation.

Disaster preparedness.

Social welfare.

AI is rapidly beginning to extend into these public services as well.

We call the approach of applying AI in the public sphere Public AI.

Yet if Public AI is understood merely as the automation of administrative work, its full potential cannot be realized.

For residents, government, healthcare, education, and welfare are not separate services. They are connected within one life.

A child is born.

A family looks for childcare.

The child attends school.

Someone becomes ill and visits a hospital.

A family member requires care.

During a disaster, the person receives evacuation information.

Later in life, the person uses welfare services.

Administratively, these are managed by different departments and systems. From the resident’s perspective, however, they are all part of one life.

Public AI must do more than improve the efficiency of individual tasks. It must understand the relationships among government, healthcare, education, and other domains, and connect people to the support they need according to their circumstances.

From this perspective, Public AI can be understood as a way to implement the Relationship Economy in the public sphere.

Public Services Are Fragmented

Today, public services are often divided by organization and system.

In government, separate departments may handle:

• Resident registration
• Taxation
• Child-rearing support
• Social welfare
• Disaster preparedness
• Housing
• Employment support

In healthcare, different information is held by:

• Hospitals
• Clinics
• Pharmacies
• Care facilities
• Health insurers
• Municipal health programs

In education, the following are typically run separately:

• Schools
• Teachers
• Boards of education
• Supplementary schools
• Libraries
• Welfare organizations
• Community activities

Each system is necessary. Separating organizations has value in preserving expertise and clarifying responsibility.

But when fragmentation becomes excessive, residents must research systems on their own, find the appropriate office, and repeatedly explain the same information. Those who need support most may be the least able to navigate complex systems.

The role of Public AI is not simply to merge existing systems. It is to connect the support a resident needs as a relationship, while preserving institutional and organizational boundaries.

The Limits of Transaction-Based Public Services

Traditional public services have been designed around individual applications and procedures.

A resident applies. The government reviews the application. The result is communicated.

This is a form of transaction.

Examples include:

• Applying for childcare
• Applying for a subsidy
• Receiving medical treatment
• Enrolling in school
• Applying for long-term care certification

But what people truly need is not only the completion of one procedure.

Consider a family after childbirth. They may need:

• Birth registration
• Health insurance
• Child allowances
• Infant health checkups
• Vaccinations
• Childcare
• Parenting consultations
• Maternal health support
• Support for returning to work

Under conventional systems, residents investigate each program and complete each procedure separately. From the perspective of the Relationship Economy, however, this is understood as a single state: “A child has been born, and the family’s living circumstances have changed.”

Based on that state, the relevant public services can be offered in the appropriate sequence.

This is the shift Public AI requires: from transaction-based public services to continuous support aligned with people’s life circumstances.

What Is Public AI?

Public AI does not simply mean that government agencies adopt generative AI. It is an approach to using AI for public services, social infrastructure, and local communities in order to enhance the well-being of society as a whole.

Where corporate AI primarily emphasizes:

• Revenue
• Profit
• Productivity
• Customer lifetime value

Public AI must emphasize:

• Public benefit
• Fairness
• Safety
• Accountability
• Inclusion
• Sustainability
• Social trust

Even if an administrative AI increases the number of cases processed, public value has not increased if people unfamiliar with digital tools can no longer access services. Even if medical AI improves diagnostic efficiency, patients cannot use it with confidence if they do not understand the basis of its recommendations. Even if educational AI improves learning outcomes, it may widen inequality if access differs according to family circumstances.

Public AI must consider not only efficiency, but also who benefits and who may be left behind.

Public AI as a Relationship Economy

In the Relationship Economy, value arises not from a one-time transaction, but from continuing relationships. The same is true in the public sphere.

Government and residents.

Physicians and patients.

Schools and learners.

Teachers and families.

Municipalities and local communities.

People and AI.

Public value emerges when these relationships are formed and sustained.

For example, a resident receives the right information from government at the right time. Necessary support continues even when a patient changes medical providers. Schools, families, and communities work together to support a child’s learning. During a disaster, municipalities, transportation providers, healthcare organizations, and local residents coordinate.

All of these are forms of value created through relationships.

Public AI is not only a way to make individual public services more efficient. It can become the foundation for understanding relationships in the public sphere and creating the cooperation that is needed.

Government and the Relationship Economy

Government is one of the organizations most deeply involved in residents’ lives. Yet residents often interact with it only when they file an application or encounter a problem.

Moving home. Childbirth. Unemployment. Illness. Long-term care. Disaster.

These are moments when a person’s life circumstances change substantially.

In conventional government services, residents search for programs, visit offices, and submit the required documents. If Public AI has State Understanding, it can change the way government services are conceived.

For example, suppose a resident becomes unemployed. That person may need more than an explanation of unemployment benefits. Relevant support may include:

• Job-search assistance
• Vocational training
• Changes to health insurance
• Pension procedures
• Housing assistance
• Household financial counseling
• Educational support for children

If AI can understand the person’s current life circumstances, family composition, available programs, local support organizations, goals, urgency, and application deadlines, it can guide the person through the necessary support as one coherent journey.

Administrative AI then becomes more than a chatbot that answers questions. It becomes a navigator that connects a resident’s life circumstances with public services.

Knowledge Flow in Government

Government contains an enormous body of knowledge:

• Laws
• Ordinances
• Regulations
• Administrative guidelines
• Eligibility requirements
• Operating manuals
• Past cases
• Official notices
• Statistics
• Regional plans
• Disaster-response procedures

Yet these materials are distributed across departments and systems. Old information can remain after a policy change, and explanations of the same program may differ from one staff member to another.

Through Knowledge Flow, administrative documents can be used to extract and organize the following into a Knowledge Graph:

• Eligible people
• Conditions
• Deadlines
• Required documents
• Exceptions
• Responsible departments
• Related programs
• Decision criteria

AI can then do more than search documents. It can understand which programs are relevant to a particular resident’s current state.

This does not mean that AI should automatically make final legal judgments. When uncertainty exists, or when a resident’s rights may be affected, the case must return to a responsible professional through a Human Gate.

Healthcare and the Relationship Economy

In healthcare, too, a patient is not defined by a single consultation.

Daily life. Prevention. Diagnosis. Treatment. Medication. Rehabilitation. Long-term care. Family support.

All of these are connected over time.

Healthcare today, however, is prone to fragmentation. Hospitals hold clinical information. Pharmacies hold medication information. Care facilities understand daily living conditions. Families observe day-to-day changes. Municipalities hold health-check and welfare information.

Even when each party has important information, the patient’s overall condition cannot be understood unless those parties are connected appropriately.

From the perspective of the Relationship Economy, medical value is not measured only by the number of consultations or prescriptions. It lies in helping patients live with confidence, preventing illness, ensuring that necessary support does not stop, and enabling healthcare providers, care providers, and families to cooperate.

State Understanding in Healthcare

For healthcare AI, the diagnosis alone is not enough. What matters is the patient’s current state.

Even among patients with the same illness, the support required differs depending on whether they:

• Live alone
• Have family support
• Can attend appointments
• Can continue taking medication
• Remain in employment
• Face financial concerns
• Have other health conditions
• Face an urgent situation

State Understanding integrates:

• Clinical Context
• Life Context
• Goal
• Intent
• Risk
• Constraint
• Support Network

For example, the same medication may be clinically appropriate, but a patient may find it difficult to take it at the same time every day. In that case, prescribing the medication alone is not enough. The patient may need an explanation for family members, medication support, home nursing, digital reminders, or coordination with a pharmacy.

By understanding not only illness but also living circumstances and support relationships, AI can help make healthcare more continuous.

Healthcare Requires Strong Trust Infrastructure

Healthcare is a domain of Public AI in which exceptionally high trust is required. Health information is highly sensitive, and AI recommendations can have major consequences for life and well-being.

Medical AI therefore needs clear Boundaries, including:

• Information that AI may access
• The scope limited to diagnostic support
• Content AI may propose automatically
• Decisions requiring physician confirmation
• The patient’s consent
• Escalation in emergencies
• Records of the basis for decisions
• Data-retention periods

Presenting diagnostic candidates is not the same as making a final diagnosis. Presenting treatment options is not the same as applying them automatically to a patient.

Clear Human Gates are needed so that physicians and patients can participate in decisions.

Decision Trace must also record:

• Which symptoms were considered
• Which test results were used
• Which guidelines were consulted
• Which options were compared
• Who made the final confirmation

Trust in healthcare is formed not by AI accuracy alone, but also by explainability, control, and clear responsibility.

Education and the Relationship Economy

The value of education is not determined by a single lesson or examination. Children and students learn over long periods of time, growing through relationships with teachers, families, friends, learning materials, communities, future schools, and companies.

Traditional education systems have mainly recorded:

• Attendance
• Grades
• Test results
• Assignment submissions
• Grade progression

But a learner’s true state cannot be expressed through numbers alone.

The learner may not understand the material. They may have lost confidence, developed a strong interest in a particular field, face difficulty at home, struggle with peer relationships, or have not yet found a future goal. Such context has a major impact on learning.

Educational AI must do more than give correct answers. It must understand the learner’s state and connect the learner to the appropriate people and opportunities.

From AI Tutor to Learning Relationship

When generative AI is introduced into education, it can answer questions and create problems as an AI Tutor. This is useful, but it alone does not make educational AI part of the Relationship Economy.

What matters is creating a continuing Learning Relationship between AI and the learner.

If AI can understand:

• What the learner understands
• Where the learner is struggling
• Which explanatory methods work best
• What interests the learner
• Which goals the learner has
• When support is needed

it can provide learning support suited to each individual.

It can also go beyond AI alone by:

• Sharing relevant information with teachers
• Proposing support methods to families
• Introducing libraries and community activities
• Connecting learners to specialists
• Guiding them to future learning opportunities

This is an evolution from an AI Tutor to a Learning Coordinator that connects learners, teachers, families, and communities.

Community Graph in Education

Many relationships influence learning outcomes in education, including:

• Learners and teachers
• Learners and learning materials
• Learners and peers
• Schools and families
• Schools and communities
• Schools and companies
• Learning content and future occupations
• AI and learners
• AI and teachers

Using a Community Graph makes it possible to represent relationships such as:

• Which materials helped understanding
• Which teacher support was effective
• Which activities increased motivation to learn
• Which fields a learner’s interests are expanding into
• Which supporters the learner should be connected with

However, relationships among children and students must not be evaluated or monitored excessively. It is highly risky for AI to classify friendships or personalities unilaterally.

Inferences must be distinguished from facts, and learners, guardians, and teachers must have a means to correct them. Educational AI should not be used to assign fixed scores to learners; it should be used to support change and possibility.

Government, Healthcare, and Education Are Connected

Government, healthcare, and education may appear to be separate domains. In residents’ lives, however, they are deeply interconnected.

For example, imagine that a child’s school absences begin to increase. The cause may not be motivation to learn alone. It could involve health problems, household finances, care or housework responsibilities, peer relationships, developmental challenges, transportation issues, or many other factors.

When a school cannot address the problem alone, healthcare, welfare, government, and community support need to coordinate.

This does not mean that Public AI should share information across domains without limits. On the contrary, under informed consent and clear Boundaries, the essential task is to identify necessary support relationships and connect the person with the appropriate organization.

Rather than integrating government, healthcare, and education into one giant database, we should build a relationship infrastructure that preserves each domain’s responsibility and authority while enabling coordination when necessary.

The Runtime Society That Supports Public AI

Public services involve many specialized actors:

• Government Agents
• Healthcare Agents
• Education Agents
• Welfare Agents
• Disaster-Response Agents
• Transportation Agents
• Community Agents
• Human Professionals

The Runtime Society is a framework in which these actors have defined roles and coordinate when necessary.

For example, when a disaster occurs:

The disaster-response Agent detects hazardous areas.

The transportation Agent checks available means of travel.

The healthcare Agent identifies people who need assistance.

The welfare Agent coordinates support for older adults and people with disabilities.

The government Agent provides shelter information.

Through a Human Gate, responsible personnel confirm important decisions.

When multiple AIs and people coordinate in this way, responses can consider the safety of the whole community rather than optimizing individual functions separately.

In the public sphere, however, rights protection sometimes takes priority over efficiency. Even when AI determines an action to be optimal, it must not disregard a person’s wishes or legal rights.

Runtime Society therefore requires not only technical coordination, but also the design of public values and responsibility.

Next Best Action in Public AI

In corporate AI, Next Best Action is sometimes used to identify actions that increase revenue or customer lifetime value. In Public AI, its purpose is different.

In government, it may mean:

• Guiding someone to the appropriate program
• Notifying them of an application deadline
• Connecting them to a specialist office
• Prioritizing emergency support
• Taking no action and waiting for the person’s own decision

In healthcare, it may mean:

• Recommending a consultation
• Escalating to a physician
• Providing medication-support guidance
• Coordinating with family members or caregivers
• Prompting an emergency response

In education, it may mean:

• Suggesting different learning materials
• Consulting a teacher
• Recommending rest
• Connecting a learner to an activity of interest
• Proposing specialized support

Next Best Action in the public sphere is not an action that maximizes efficiency. It must be an action that supports the person’s rights, safety, growth, and stability in daily life.

How Do We Measure Public Value?

The outcomes of Public AI cannot be measured only by case volumes or cost reduction. Possible indicators include the following.

Government

• Procedure completion rate
• Rate of successful access to support programs
• Time from consultation to resolution
• Rate of missed support opportunities
• Resident satisfaction
• Reduction of the digital divide

Healthcare

• Prevention rate
• Treatment-continuation rate
• Readmission rate
• Medication-adherence rate
• Patient quality of life
• Healthcare and long-term care coordination rate

Education

• Learning-continuation rate
• Level of understanding
• Motivation to learn
• Rate of connection to support
• Prevention of school refusal and disengagement
• Equity of learning opportunities

From the perspective of the Relationship Economy, public value is not measured by how many times someone uses a service. It is measured by whether people who need support are connected to it, and whether that relationship continues.

Trust Infrastructure Required for Public AI

In the public sphere, the use of AI may directly affect citizens’ rights and lives. Trust Infrastructure is therefore essential.

At a minimum, it requires:

• Clear purposes of use
• The individual’s consent
• Data use limited to what is necessary
• Identity and Role management
• Access control across organizations
• Separation of inferences from facts
• Testing for discrimination and bias
• Decision Trace
• Human Gate
• Mechanisms for appeal
• Correction and reconsideration
• Audit

For example, if AI infers that a household “needs support,” that inference must not be treated as a definitive assessment. An incorrect inference can cause serious harm if it affects education, welfare, employment, or other services.

AI can support decisions. It must not restrict people’s rights through classifications that cannot be explained.

Trust in Public AI is not a demand that people simply “trust the AI.” It is created through a system in which citizens can review, correct, and challenge AI-supported decisions.

From Public AI to Public Relationship

The true purpose of public services is not merely to operate programs. It is to create a society in which residents can live with security and confidence.

That requires relationships in which:

Government understands residents.

Healthcare understands patients’ lives.

Education supports learners’ possibilities.

Communities prevent isolation.

People and AI coordinate through appropriate roles.

The ultimate value of Public AI is not AI performance. It is whether it can create better relationships between public institutions and citizens.

Trustworthy government.

Continuous healthcare.

Education that stays close to each individual.

Communities that naturally connect people to the support they need.

The relationship infrastructure that supports these outcomes is Public Relationship.

Implementing Public AI with the Chinoba Platform

When Public AI is implemented as a Relationship Economy, the architecture of the Chinoba Platform can be applied.

Public Data and Documents
↓
Laws, systems, healthcare and educational knowledge, and regional information

Knowledge Flow
↓
Structures necessary conditions, support content, and related programs

Knowledge Graph / Community Graph
↓
Represents relationships among residents, government, healthcare, education, welfare, and communities

State Understanding
↓
Understands current living, health, learning, and community conditions

Decision AI
↓
Evaluates candidate forms of necessary support

Trust Infrastructure
↓
Confirms consent, authority, Boundaries, and Human Gates

Operational AI
↓
Provides guidance, reservations, application support, and connections to specialists

Decision Trace
↓
Records why support was offered and who confirmed it

Feedback and Learning
↓
Reflects support outcomes in the improvement of future programs

Through this cycle, public services can move from one-time procedures to continuous relationship-based support.

Conclusion

Public AI is neither an AI that summarizes administrative documents nor an AI that replaces physicians or teachers. It is an infrastructure for crossing the divisions among government, healthcare, education, welfare, and communities to create the support relationships that people need.

Residents are not people divided by program. They live one life while holding multiple roles: patient, guardian, taxpayer, learner, and community member.

That is why Public AI needs State Understanding.

Knowledge Flow delivers public knowledge.

Community Graph understands relationships.

Trust Infrastructure protects rights and safety.

Runtime Society coordinates government, healthcare, education, and human professionals.

With this framework, public services can evolve from “a system for processing applications” into “a relationship infrastructure that continuously supports people’s lives.”

The Relationship Economy is not an idea solely for increasing corporate revenue. It can also be used to improve social trust, inclusion, and quality of life across society.

We see the implementation of the Relationship Economy in the public sphere as an important direction for Public AI.

In the next article, we will integrate the Knowledge Flow, State Understanding, Community Graph, Trust Infrastructure, Runtime Society, and Public AI concepts introduced so far, and explore what Chinoba Economics is: a new economic design for the AI era.