
The Great Conversation: What Happens When Private Technology Becomes National Power?

Hello everyone, and welcome to The Great Conversation.
For much of history, the symbols of national power were relatively easy to recognize.
Armies.
Warships.
Factories.
Oil.
Nuclear weapons.
Industrial capacity.
Economic strength.
Control enough of these things, and a nation could project its influence far beyond its borders.
But a new form of power may now be joining that list.
Artificial intelligence.
A recent Telegraph article examines the growing role of AI in the competition between the United States and China.
Its argument begins with an interesting comparison.
When nations want to demonstrate military strength, they display weapons.
Aircraft.
Missiles.
Ships.
Soldiers.
But increasingly, technological leaders themselves are becoming part of the demonstration.
Executives.
Chip manufacturers.
AI companies.
Data centers.
Computing capacity.
Models.
Algorithms.
The message is no longer simply:
Look at our military.
It is also:
Look at our technology.
Look at what our companies can build.
Look at how much computing power we possess.
Look at the intelligence our machines can produce.
And perhaps most importantly:
Look at what we may be capable of doing next.
This represents an interesting evolution in the meaning of geopolitical power.
The Telegraph article compares today's competition over artificial intelligence with the nuclear competition of the twentieth century.
There is an important difference, however.
Nuclear weapons were largely developed, owned and controlled by governments.
The most powerful AI systems today are largely being developed by companies.
Companies such as OpenAI, Anthropic, Google, Microsoft, Meta and others are building systems that governments increasingly view as strategically important.
Nvidia became extraordinarily valuable because its chips became essential infrastructure for training and operating many advanced AI systems.
This means some of the technologies that governments increasingly consider vital to national power are not actually owned by governments.
They are owned by corporations.
And sometimes enormous influence over those corporations can be concentrated among a relatively small number of executives, founders and investors.
That creates an unusual relationship.
Governments regulate technology companies.
Technology companies depend on governments.
Governments depend on technology companies.
Companies compete globally.
Governments compete geopolitically.
And sometimes the interests of a company and the interests of the country in which it operates may overlap.
Sometimes they may not.
This is not entirely new.
Private companies have influenced political power for centuries.
Railroads once shaped national economies.
Oil companies became geopolitical actors.
Banks influence governments.
Defense contractors build military systems.
Telecommunications companies operate infrastructure societies cannot function without.
But artificial intelligence introduces something different.
AI is not merely one industry.
It may become a layer running through many industries simultaneously.
Medicine.
Finance.
Education.
Manufacturing.
Defense.
Scientific research.
Media.
Software.
Transportation.
Government.
Communication.
If AI becomes deeply embedded across all of these areas, then controlling advanced AI may mean controlling something closer to an infrastructure of intelligence.
And that raises a difficult question.
At what point does a private technology become a public strategic concern?
We have encountered versions of this question before.
Electricity began as a technological innovation.
Eventually, reliable access to electricity became essential infrastructure.
Telecommunications followed a similar path.
The internet began as a network.
Today, entire economies depend upon it.
Artificial intelligence may be following another version of that journey.
It begins as a product.
Then it becomes a platform.
Then infrastructure.
And eventually, perhaps, power.
But even the idea that nations are participating in an "AI race" deserves examination.
The United States has increasingly described artificial intelligence as a strategic competition, particularly with China.
The assumption is intuitive.
Whoever develops the most advanced AI first gains an enormous advantage.
But the Telegraph article introduces another possibility.
What if the competition is not simply about who builds the most powerful model?
What if it is also about who makes AI cheapest?
Who distributes it most widely?
Who integrates it most effectively into the economy?
Who allows businesses to adapt it?
Who controls the hardware?
Who controls the energy required to run it?
And perhaps even who decides not to control it?
China has increasingly supported open AI models that can be downloaded, modified and adapted by others.
That strategy changes the economics.
Instead of capturing all of the value inside the AI model itself, a country might make AI inexpensive and capture value elsewhere.
Manufacturing.
Robotics.
Vehicles.
Industrial systems.
Consumer products.
Infrastructure.
In that world, the most strategically successful AI may not necessarily be the most expensive or the most powerful.
It may be the one that becomes ordinary.
Cheap.
Accessible.
Embedded everywhere.
This is where the analogy with previous technological revolutions becomes interesting.
The most important technology is not always the technology with the highest price.
Sometimes it is the technology that becomes so inexpensive and widespread that everyone begins building upon it.
Electricity transformed civilization not because a few organizations possessed extraordinary amounts of electricity.
It transformed civilization because electricity became available almost everywhere.
The internet became transformative because billions of people gained access to networks.
Computing changed society as computers became cheaper and smaller.
Perhaps intelligence itself could follow a similar path.
If so, the geopolitical question changes.
Instead of asking:
Who has the most powerful AI?
We might eventually ask:
Whose AI became the foundation upon which everyone else built?
But this introduces another problem.
Concentration.
Microsoft chief executive Satya Nadella has warned against a future in which a handful of AI models capture the value created across entire industries.
That concern reaches far beyond competition between technology companies.
Imagine a world in which millions of businesses depend on only two or three AI systems.
Those systems might influence how companies operate.
How information is organized.
How software is created.
How workers perform their jobs.
How research is conducted.
How people access knowledge.
Perhaps even how governments make decisions.
At that point, the companies operating those systems would possess something unusual.
Not political authority in the traditional sense.
They would not necessarily make laws.
They would not command armies.
They would not collect taxes.
But they might control infrastructure through which enormous portions of society think, communicate and operate.
Power does not always require a government office.
Sometimes power comes from controlling what everyone else depends upon.
This is why debates over AI regulation are becoming so complicated.
One side worries that excessive regulation could slow innovation and allow rival countries to gain an advantage.
Another worries that insufficient regulation could allow extraordinarily powerful systems and companies to develop without adequate oversight.
Both concerns can be true at the same time.
Move too slowly, and another country may gain strategic advantage.
Move too quickly, and society may deploy technologies whose consequences it does not fully understand.
Regulate too aggressively, and established companies may become even stronger because smaller competitors cannot afford compliance.
Regulate too lightly, and the most powerful companies may become stronger anyway.
There is no obvious equilibrium.
And underneath the regulatory debate lies an even older political question.
Who governs whom?
Governments traditionally claim authority over corporations operating within their borders.
But what happens when governments themselves become dependent upon technologies produced by those corporations?
What happens when military systems require their chips?
When government agencies use their AI?
When national economic competitiveness depends upon their innovation?
When political leaders bring technology executives to international summits as symbols of national strength?
The relationship begins to change.
A corporation may still exist under the authority of the state.
But the state may simultaneously depend upon the corporation.
Power becomes reciprocal.
And history contains many examples of private institutions becoming powerful enough to influence the governments that supposedly governed them.
Artificial intelligence may simply be creating a twenty-first-century version of an old political problem.
But there is another possibility.
Perhaps focusing only on corporations and governments misses the most important transformation.
The deepest political power of AI may eventually come not from who owns it, but from how deeply society relies upon it.
If AI systems increasingly help determine what information people see, how information is summarized, what recommendations people receive, what opportunities are presented, what risks are detected and how decisions are made, then AI may influence society without ever issuing a command.
Not by governing people directly.
But by shaping the environment in which people make choices.
That is a subtler form of power.
And potentially a more pervasive one.
Because political power traditionally tells people what they may or may not do.
Informational power influences what people believe is possible, important or true.
Artificial intelligence could eventually participate in both.
Which makes the question of control much larger than who wins an international technology race.
It becomes a question about the architecture of society itself.
Who builds these systems?
Who owns them?
Who regulates them?
Who audits them?
Who can compete with them?
Who can access them?
Who decides what they are permitted to do?
And what happens when governments, corporations and citizens all become dependent upon the same systems?
Perhaps the future of political power will not belong entirely to governments.
Nor entirely to corporations.
Nor entirely to artificial intelligence.
It may emerge from the relationship between all three.
Governments possessing legal authority.
Companies possessing technological infrastructure.
AI systems possessing increasingly powerful capabilities.
And citizens trying to determine how much authority they are willing to surrender to any of them.
Which raises a deeper question...
If artificial intelligence becomes infrastructure that governments, businesses and citizens all depend upon, who should ultimately have the power to decide how that intelligence is built, controlled and used?
Take a moment to reflect.
And click the link below to see more on this subject.

