Can Everyone Become an Entrepreneur?

The scale problem in Naval Ravikant’s future of work—and why an AI-enabled one-person billion-dollar company would not make entrepreneurship a universal economic solution.

By Philos Kim
October 4, 2026

An opportunity that can enrich a small number of early adopters does not automatically become an economic system that can support everyone.

Naval Ravikant has offered an arresting vision of the future.

“In the future, whether it’s 50 or 100 years from now, virtually everyone will be working for themselves,” he said during a 2019 appearance on The Joe Rogan Experience.¹

Satish Kanwar made a related argument in a 2017 TEDxToronto talk titled Why Entrepreneurship Is the Future of Work. His formulation was broader and more careful: entrepreneurship is not merely a job title or the act of creating a conventional company. It is a mindset—initiative, experimentation and the willingness to solve problems—that can exist inside employees as well as founders.²

Artificial intelligence makes both visions more plausible.

A single person can now research markets, write software, develop advertising, create images, translate content, answer customers and analyze data with capabilities that once required a team. The cost of starting certain businesses is falling. People who recognize the opportunity early may create millions—or billions—of dollars in value.

That possibility should be taken seriously.

But it does not answer the question society must ask.

The One-Person Billion-Dollar Company May Be Real

Sam Altman has said that he and a group of technology executives maintained a betting pool over when the first one-person billion-dollar company would appear.

“Which would have been unimaginable without AI and now will happen,” he told Reddit cofounder Alexis Ohanian.³

He may be right.

A person could create a valuable software platform, scientific breakthrough, financial system, media property or intellectual asset and use AI agents to perform much of the work previously divided among hundreds of employees.

But a one-person billion-dollar company would be an extraordinary outcome precisely because it would be rare.

It is better understood as a new kind of attainable lottery win.

Unlike a random lottery ticket, the entrepreneur may need unusual insight, judgment, timing, persistence, capital and skill. The achievement would be earned. It could also become more attainable than building an earlier billion-dollar company that required thousands of employees.

But making an exceptional outcome more attainable does not make it typical.

The existence of one AI-enabled billionaire would prove that one person can build extraordinary value. It would not prove that billions of people can each build a viable market.

In fact, the very success of a one-person billion-dollar company could demonstrate the opposite.

If one individual can serve a global market with almost no employees, that company may eliminate opportunities that would once have supported thousands of workers and hundreds of smaller suppliers.

The company’s value may rise because it captures an enormous amount of demand with an extraordinarily small payroll.

That is remarkable productivity.

It is not broad participation.

The Fallacy of Composition

Economists have a name for the mistake at the center of the universal-entrepreneur argument: the fallacy of composition.

What is true for one participant—or even one million participants—may become false when everyone attempts it simultaneously.

A restaurant can attract customers by staying open later than its competitors. Every restaurant cannot gain the same advantage by staying open later.

An early AI consultant can earn premium fees because few competitors understand the technology. If millions of people offer comparable AI-generated services, scarcity disappears. Prices fall. Customers expect faster work for less money.

A one-person company can grow rapidly by selling into a market filled with wage earners. It cannot assume that those wages will remain available if one-person companies and automated corporations eliminate employment throughout the economy.

The problem is not that too many people would become creative or self-directed.

The problem is that a business needs more than the ability to produce.

It needs customers.

AI Can Create Producers. It Cannot Automatically Create Buyers.

The entrepreneurial vision focuses on the supply side of the economy.

AI gives people tools to produce software, writing, designs, analysis, entertainment, education and advice. It may eventually help small operators coordinate manufacturing, logistics and physical services as well.

But lower production costs do not create unlimited demand.

People have limited time and attention. Households have limited income. Organizations have limited budgets. Physical products require energy, materials, factories, transportation and storage. A customer does not need seven billion accountants, seven billion marketing agencies, seven billion software applications or seven billion personalized brands.

Even bespoke physical production does not escape the constraint.

AI may help a person design a unique product and place an order for ten units. Someone must still obtain the raw materials, operate the machinery, maintain the factory, package the order and deliver it. Small production runs can become less expensive. They do not become costless.

AI can democratize the ability to offer something for sale. It cannot guarantee that anyone will buy it.

This distinction becomes more important as AI spreads.

The first person using AI to perform ten people’s work possesses a competitive advantage.

When every competitor possesses the same capability, the advantage shifts elsewhere—to reputation, capital, proprietary data, distribution, physical infrastructure, customer relationships and ownership of the platforms through which business is conducted.

The technology may become widely available while economic power becomes more concentrated.

We Already Have Millions of One-Person Businesses

The one-person business is not a futuristic invention.

The United States already has tens of millions of them.

The Census Bureau counted 29.8 million nonemployer businesses—businesses without paid employees—in 2022. Together, they reported $1.7 trillion in receipts.⁴

Those figures demonstrate the importance and reach of entrepreneurship.

They also reveal its limits as a universal substitute for employment.

The same Census release counted 35.7 million employer and nonemployer businesses with $51.7 trillion in total receipts. Nonemployer businesses therefore represented roughly 83 percent of the business count but only about 3 percent of reported receipts.⁴

That comparison is not an indictment of small enterprise. Many nonemployer firms provide their owners with independence, useful supplemental income or a satisfying livelihood. Some will grow into substantial employers. Others are intentionally small.

But registering a business is not the same as creating a stable living.

A 2026 report from the Federal Reserve Banks found that nonemployer firms were less likely to be profitable than businesses with employees. Sixty-four percent relied on their owners’ personal funds to address financial challenges, compared with 54 percent of employer firms.⁵

The Bureau of Labor Statistics projected self-employed workers to remain about 5.8 percent of total employment in 2034—almost unchanged from 2024.⁶

AI may change that trajectory. But the existing numbers warn us against treating the legal existence of a business as evidence that entrepreneurship can reliably replace wage income, benefits and economic security.

Entrepreneurial Freedom Can Conceal Transferred Risk

Working for oneself can be genuinely liberating.

It can also be a label placed on economic insecurity.

A company can replace employees with contractors and describe the result as entrepreneurial opportunity. The workers may gain flexibility, but they may also inherit the risk of inconsistent demand, unpaid administrative work, equipment costs, insurance, retirement saving, downtime and illness.

AI platforms could intensify this transfer.

A worker may be told that losing a job is no longer a serious problem because AI makes it possible to start a business. The worker then enters a crowded marketplace controlled by a few platforms, pays subscription and transaction fees, competes with thousands of similar providers and bears the cost when customer demand does not materialize.

The institution has removed a salary.

The individual has received an opportunity to bid for one.

A society cannot convert every lost paycheck into an invoice and call the transition complete.

Entrepreneurship should be a meaningful choice.

It should not become the explanation offered to everyone for why stable employment and employer-supported benefits are disappearing.

The Middle Class Was Not Built Entirely From Founders

The United States prospered during the great expansion of its middle class not because every household founded a company, but because millions of people received dependable incomes and could participate in mass consumption.

Workers bought homes, cars, appliances, education, entertainment and services. Those purchases supported businesses. The businesses created additional jobs and investment. A broad tax base supported infrastructure and public institutions.

Wealthy people did not need the middle class to remain poor in order to become wealthy.

They benefited from owning companies that served a vast population with purchasing power.

AI could create greater total wealth while weakening that reinforcing cycle.

If income moves from millions of wage earners toward a much smaller number of owners of models, platforms, capital, data and infrastructure, production may rise while widespread purchasing power deteriorates.

Millions of nominal businesses do not solve that problem if most of them are competing for a shrinking pool of customers.

Kanwar’s Stronger Argument Should Survive

The answer is not to dismiss entrepreneurship.

Kanwar’s deeper argument—that initiative and entrepreneurial thinking will become more important—is persuasive. Workers inside companies will increasingly need to identify problems, combine tools, exercise judgment and create new value. AI may allow more people to test ideas without first obtaining permission or raising substantial capital.

That is a real expansion of human agency.

But entrepreneurial capability and business ownership are not identical.

A nurse can think entrepreneurially while working in a hospital. A machinist can improve a process inside a factory. A government employee can redesign a public service. A technician can identify a customer problem without accepting all the financial risk of becoming an independent contractor.

Society should cultivate initiative without pretending that every person must become a founder, salesperson, accountant, benefits administrator and collections department simply to earn a living.

A Better AI-Entrepreneurship Agenda

AI-enabled entrepreneurship can be part of a healthier future of work.

It cannot carry the entire future by itself.

A serious policy agenda would:

• Teach entrepreneurial problem-solving without presenting business ownership as the only respectable response to displacement.
• Preserve genuine employment, apprenticeship and entry-level pathways for people who want or need stable work.
• Create portable healthcare, retirement and training systems that do not make independence financially reckless.
• Prevent dominant platforms from capturing most of the value created by millions of small operators.
• Expand access to AI tools, computing capacity, capital and distribution rather than allowing those resources to concentrate among incumbents.
• Support cooperatives, employee ownership and other models that let more people share in the productive assets of automation.
• Measure whether AI businesses are creating sustainable household income—not merely counting business registrations or celebrating exceptional valuations.
• Protect aggregate demand by ensuring that the gains from automation circulate broadly enough to support customers as well as producers.

The goal should not be to prevent an individual from building a billion-dollar company.

We should celebrate that achievement when it creates genuine value.

The goal should be to avoid mistaking that exceptional success for a plan capable of supporting an entire population.

The Question Is Not Whether Someone Can Win

Naval Ravikant may be right that many more people will eventually work for themselves.

Satish Kanwar may be right that entrepreneurship—or at least an entrepreneurial mindset—will become central to the future of work.

Sam Altman may be right that AI will help create a one-person billion-dollar company.

All three possibilities can be true.

None establishes that billions of independent businesses can all produce adequate income, nor that entrepreneurship can absorb workers displaced faster than customers and new markets emerge.

The question is not whether someone can use AI to become extraordinarily wealthy.

Someone almost certainly will.

The question is whether the economic model remains viable when the opportunity is offered to everyone, the competitive advantage is no longer scarce and the former employees who were supposed to become entrepreneurs are also the customers every entrepreneur needs.

An economic opportunity is not a social solution unless it remains viable when offered to everyone.

AI may give more people the tools to build.

Society must still build an economy in which the maximum practical number of people can earn, participate and thrive.