Exploring whether broadly accessible AI could become a new form of public infrastructure—expanding human capability while protecting freedom and encouraging global cooperation.
For much of modern history, national progress has been built through physical infrastructure. Governments invested in roads, bridges, power systems, water networks, schools, hospitals, and communications because these systems expanded what people and economies could do.
National security followed another logic. Countries invested heavily in military capability and deterrence because governments considered them necessary to protect their populations and national interests.
Both remain important. Physical infrastructure will continue to support everyday life, economic activity, and technological development, while countries will continue to consider legitimate security needs.
But the emergence of artificial intelligence raises a new question:
Should broadly accessible AI become part of the basic public infrastructure available to people?
AI differs from many traditional forms of infrastructure because it has the potential to assist individuals directly—helping people learn, understand information, solve problems, create, communicate, and navigate increasingly complex systems.
This raises a broader possibility worth exploring: whether some future investment—including resources devoted to continued military escalation—could instead expand access to technologies that increase human capability.
The idea is not to replace physical infrastructure or eliminate legitimate national defense. It is to broaden the conversation about what infrastructure means in an increasingly technological world.
Technology Should Serve People
Many forms of infrastructure improve human life indirectly. Roads improve mobility. Power systems provide energy. Communications networks connect people.
AI has the potential to operate closer to the individual.
A capable and broadly accessible AI system could provide forms of assistance that historically depended heavily on income, location, education, or institutional access.
A student could receive individualized educational support in a local language. A small-business owner could obtain help with planning, analysis, writing, or understanding regulations. Someone confronting an unfamiliar government process could receive plain-language assistance understanding forms and requirements.
AI could also help people understand complex information, prepare better questions for professionals, develop ideas, learn new skills, and reduce routine administrative burdens.
The larger principle is straightforward:
Technology should progressively reduce unnecessary human burdens and expand people’s ability to learn, create, participate, and live meaningful lives.
The question is therefore not only what technology can produce, but also:
What can technology enable people to do for themselves—and what would it take to keep that opportunity broadly accessible at scale?
AI as an Equalization Infrastructure
New technologies frequently arrive unevenly.
Those with greater wealth, education, infrastructure, and institutional access often benefit first. Without broader access, AI could follow the same pattern.
A wealthy student could have continuous access to advanced tutoring while another student has none. Large companies could deploy powerful analytical systems while small businesses cannot afford comparable capability. Countries with extensive technological resources could accelerate while others become increasingly dependent on systems developed elsewhere.
Broad public access offers another possibility.
If capable AI becomes accessible regardless of income or geography, it could help raise the capability floor rather than simply raise the technological ceiling.
This does not mean everyone would achieve the same outcomes. People have different circumstances, interests, abilities, and opportunities.
The more fundamental question is:
Should a person’s economic circumstances determine whether they can access one of the most powerful knowledge and reasoning technologies available—or only how capable a version of it they get?
Local Capability Without Technological Isolation
Broad access does not mean that every country would need to develop its own advanced AI system from the ground up.
Countries differ greatly in population, resources, technical capability, energy availability, languages, and development priorities. Different approaches could therefore emerge—domestically developed systems, adapted open models, regional infrastructure, public-private arrangements, or international collaboration.
At the same time, some level of local technological capability may become increasingly valuable as AI becomes more deeply involved in education, commerce, government services, communication, and everyday life.
Local capability can also help preserve languages, knowledge, history, and cultural perspectives that might otherwise receive limited representation in globally dominant systems.
This need not mean technological isolation.
Countries could build or adapt technology for their own populations while contributing knowledge, languages, research, experience, and ideas to broader global development.
The underlying possibility is:
Build capability locally where it benefits people, while cooperating globally where shared knowledge and collaboration benefit humanity.
But this raises another question:
At what point does adapting technology developed elsewhere become technological dependence rather than meaningful local capability?
From Arms Competition to Human-Capability Competition
National security is a legitimate concern, and geopolitical risks cannot simply be ignored.
But continued military competition raises another question:
At what point does additional investment in increasingly destructive capability provide less human benefit than investment in increasing human capability?
Military competition can become self-reinforcing. One country expands its capability because another has expanded its own, which can lead to further responses on both sides.
AI and other emerging technologies suggest that nations could increasingly compete in other ways as well—in education, scientific discovery, public health, productivity, environmental solutions, technological accessibility, and the ability of ordinary people to solve problems and create new opportunities.
Competition itself does not necessarily have to disappear.
Perhaps more of it could gradually become a different kind of competition:
Who can empower people most effectively—rather than only who can build the greatest capacity to deter or threaten others?
That is a direction worth examining. But a difficult question remains:
Does expanding human capability actually reduce distrust between rivals, or could it simply develop alongside military deterrence without replacing any part of it?
Build Locally, Cooperate Globally
AI creates opportunities and risks that cross national borders.
Systems developed in one country may eventually affect education, economies, information, cybersecurity, scientific research, and people elsewhere.
This makes international cooperation increasingly relevant.
Countries could exchange research, develop common approaches to safety, share useful innovations, support languages and communities that might otherwise be overlooked, and learn from one another’s experience.
At the same time, cooperation does not require every society to adopt identical technologies, values, or institutional arrangements.
A useful principle to explore may be:
Local capability where independence and cultural context matter; global cooperation where humanity shares the risks and benefits.
Public AI and Human Freedom
Broadly accessible AI would bring risks alongside its potential benefits.
A system capable of assisting millions of people could also become an instrument of surveillance, censorship, manipulation, or excessive control.
Any discussion of public AI infrastructure should therefore also consider protections for privacy, individual freedom, transparency, independent oversight, and protection from political or commercial manipulation.
The principle is important:
Technology intended to empower people should not become technology used to control them.
Human-centered technology should expand human agency rather than quietly diminish it—whether the influence comes from government or private institutions.
But control is not the only concern. A widely accessible AI system could also confidently provide information that is incomplete, biased, or simply wrong.
AI can assist people, but it should not replace individual judgment or responsibility for important decisions. At the same time, those who develop and operate AI systems have a responsibility to improve reliability, address known problems, and communicate important limitations.
As people increasingly rely on AI for information, learning, analysis, and decision support, determining where those responsibilities begin and end will become increasingly important.
The question therefore becomes:
How should responsibility be shared between those who develop and operate AI systems and the people who choose to use and rely on them?
Making the Idea Practical
AI may appear digital, but the infrastructure supporting it is physical. It depends on energy, computing capacity, communications networks, technical expertise, cybersecurity, and continuing investment.
Different approaches could be explored depending on a country’s circumstances—including domestic systems, adapted open models, public-private arrangements, regional infrastructure, or international cooperation.
Questions of cost, energy use, accessibility, privacy, safety, governance, and long-term technological dependence are all important to examine.
These questions do not weaken the idea. They help distinguish a meaningful possibility from a slogan.
A Different Measure of Infrastructure
Infrastructure has always reflected what societies consider important.
Roads expand movement.
Power systems expand access to energy.
Schools expand knowledge.
Hospitals protect health.
Communications networks expand connection.
Defense systems seek to provide security.
AI introduces another possibility:
Infrastructure that directly expands human capability.
As AI becomes more capable, perhaps national progress can increasingly be examined not only through economic production, physical infrastructure, or military strength, but also through another question:
How much knowledge, reasoning capability, creative power, and practical assistance is available to ordinary people—regardless of wealth, geography, language, or social position?
A Vision 2060 Question
Vision 2060 does not presume that a particular model of public AI infrastructure is the answer.
Instead, the idea raises questions worth examining:
If artificial intelligence becomes one of humanity’s most powerful tools, how can its benefits become broadly available rather than increasingly concentrated among a small number of corporations, governments, countries, or wealthy individuals?
And perhaps an even broader question:
Could humanity progressively direct more of its technological and financial capacity toward expanding people’s ability to learn, create, solve problems, and live meaningful lives—while reducing the resources devoted to increasingly destructive capabilities and cycles of escalating threats?
The purpose is not to prescribe a single policy, but to broaden the conversation about what technological progress is ultimately for.
As artificial intelligence becomes increasingly capable and influential in everyday life, the opportunity is not simply to make our systems more powerful but to consider how technology might expand human capability, protect human freedom, reduce unnecessary burdens, and contribute to a more meaningful, fulfilling, and sustainable human future.
Ultimately, the promise of artificial intelligence may depend not only on how powerful it becomes, but on how broadly that power can help people build their own capabilities and improve their lives.