VC-optional entrepreneurship

For years, if you wanted to build an ambitious tech startup, there was a path that seemed almost inevitable: quit your job, raise venture capital, hire fast, build fast, grow fast, and chase an outcome large enough to justify the next round.

That was not just ecosystem folklore. It was, in large part, an economic consequence. Building was expensive, testing was expensive, hiring was expensive, and going to market was expensive. Before you knew whether the market wanted what you were building, you already had to finance months or years of product, engineering, design, sales, marketing, legal, operations, and infrastructure.

That is why venture capital became the default operating system of tech entrepreneurship. Not because it was always the best path for every founder, but because for many startups it seemed like the only one possible.

I lived it as a founder. I lived it raising capital in the United States and in Europe. I lived it selling companies. And I have seen it from the other side too, investing in startups and accompanying other founders.

But something has changed. AI does not eliminate the risk of building a company. It does not turn a bad idea into a good company. It does not replace judgment, market knowledge, distribution, trust, sales, or team building. But it changes something deeper.

AI does not eliminate founder risk. It changes the cost of trying.

When the cost of trying to create a company changes, so does who can become a founder, when they need capital, how much they have to risk before validating, and what kind of startup they can build without playing the venture capital game from day one.

The startup game was never symmetrically distributed

The entrepreneurship narrative often talks about risk as if every player took on the same kind of risk. They do not.

The LP, the investor putting money into venture funds, does not usually invest their entire financial life in a single startup. They typically allocate a slice of their wealth to a specific asset class, inside a broader portfolio that may include public markets, fixed income, real estate, private equity, and other assets.

The GP, the fund manager, does not bet on a single company either. They build a portfolio. They know many investments will not work. Venture capital is designed around that reality. Returns follow a power law where a very small minority of companies explains a disproportionate share of the total. a16z, citing Horsley Bridge data, noted that around 6% of investments generated about 60% of total returns.

In the startup ecosystem, that concentration has been given a positive name: focus. The system demands focus on one idea, one company, one narrative. But that focus is not neutral. It concentrates a company, a team, a cap table, years of salary, reputation, emotional energy, personal wealth, and opportunity cost into a single bet.

For capital, a failing startup can be a line in a portfolio. For the founder, it can be a life-changing rupture.

Venture capital is not neutral

When VC comes in, it is not just money. It is a logic. It is an expectation of growth, a speed, a need for size, and a specific definition of what winning means.

A founder can build a profitable, efficient company, deeply valuable to its customers, and sell it for five, ten, or twenty million euros. For many founders, that outcome can change their lives. But for a venture capital fund, the same outcome may not be enough.

VC does not just put in money. It changes the definition of success.

If a fund enters a seed round at a twenty-million valuation, it is not investing so the company can sell for ten. It needs some portfolio companies capable of returning a meaningful share of the entire fund. That pushes the conversation toward outcomes in the hundreds of millions or more.

The founder may want a great company. The VC needs a venture-scale company. There is no necessarily bad faith. They are simply different games.

The real cost of founder failure

Saying a startup fails is too simple. Failure for an investor means a capital loss within a portfolio. Failure for a founder can mean three, five, or seven years of lost salary, lost career progression, seniority, bonuses, stock options, stability, and reputation inside an organization.

It can also mean having worked for years on a low salary only to end up with illiquid equity. It can mean being heavily diluted across several rounds. It can mean having built value for others without capturing enough upside.

It can also mean psychological wear. The pressure of not failing, of holding the team together, of raising the next round, of making payroll, of convincing the market, of justifying the previous valuation, and of not letting down investors, employees, family, and yourself carries a cost that does not appear in any return model.

For an investor, failure can be financial. For a founder, it is often biographical.

You are not just putting in money. You are putting in identity.

Entrepreneurship was not always rational for the best profiles

For years we have said there was a shortage of entrepreneurs. But maybe the problem was not a lack of entrepreneurial talent. Maybe the problem was the design of the risk.

Many of the best potential founders were working inside companies. In consultancies, telcos, banks, scaleups, tech companies, Big Four firms, multinationals, industrial groups, or listed corporations. They were not idle. They had solid salaries, progression, professional brand, network, relative stability, and future optionality.

Academic literature has long pointed out that the average entrepreneur does not necessarily receive a clear financial premium for taking on concentrated risk. Barton Hamilton found that many entrepreneurs had lower initial incomes and lower income growth than people who stayed in salaried employment.

The startup ecosystem did not just select for talent. It selected for risk tolerance.

Capital diversified. Talent concentrated.

This is, for me, the sentence that summarizes the old game.

The LP diversified by asset class, fund, manager, geography, and vintage. The GP diversified across many startups and earned fees while waiting for carry. The founder concentrated salary, career, reputation, energy, wealth, and years of professional life in a single company.

The problem was not that building a company was risky. The problem was that the risk was asymmetrically distributed. VC turned startup risk into an asset class. The founder lived it as a life decision.

For years that asymmetry was hard to avoid because the cost of building seemed to make it inevitable. Until AI showed up.

AI changes the cost of trying

AI does not make entrepreneurship easy. But it reduces the cost of trying. And that is enormous.

Market research is cheaper. Prototyping is faster. Writing code can be accelerated. Creating content costs less. Analyzing competitors is more accessible. A small team can do much more than it could five years ago.

McKinsey estimated that generative AI and other current technologies have the potential to automate activities that absorb between 60% and 70% of employees' time. Bessemer has observed AI-native companies with ARR-per-employee levels far above traditional SaaS benchmarks.

The old question was: how much capital do I need to raise to get started? The new question is different.

The new question is not how much capital you need to start. It is how much you can validate before you need capital.

VC stops being the default operating system

I do not think venture capital will disappear. On the contrary. It will remain essential for many companies. It makes sense when you need to build infrastructure, chips, foundational AI models, biotech, energy, defense, deep tech, or platforms with strong network effects.

But not every digital startup is that kind of company. Many tech opportunities are more vertical, more B2B, closer to deep industry expertise, more revenue-oriented, more capital-efficient, and less dependent on blitzscaling.

For those companies, AI enables a different sequence: build before raising, validate before quitting, reach customers before diluting, operate with small teams, look for early revenue, and decide later whether VC makes sense.

What changes for the founder

This change can open the door to a kind of founder who was previously left out of the game: the expert operator. The person who has spent ten, fifteen, or twenty years solving real problems inside an industry. Someone who knows customers, processes, inefficiencies, budgets, friction points, distribution, regulation, sales, implementation, and operations.

For years, that person may not have started a company because the jump did not pay off. They had too much to lose. Now that can change. Not because they can build a startup without risk, but because they can start building with less initial exposure.

The shift is not entrepreneurship without risk. It is changing the order of decisions: evidence before commitment, validation before capital, customers before scale.

Why I founded Junyo*

This thesis is one of the reasons I founded Junyo*. Not because I believe venture capital does not work. It does. Not because I believe risk can be eliminated. It cannot. Not because I think everyone should be a founder. They should not.

I founded Junyo* because I believe AI lets us redesign a part of the startup game that for years was overly conditioned by the cost of building. Junyo* exists to help expert operators create AI-native companies without having to quit their jobs on day one and without having to play the VC game before they have validated enough.

If someone has deep industry knowledge, if they know where the friction is, if they understand the customer, and if they can work with a small, complementary team, AI can help them turn that knowledge into a company in a way that was not possible before.

Junyo* is born from a simple idea: build before raising, validate before jumping, and decide with more evidence.

This is not about building less ambitious startups. It is about building them with a smarter risk architecture.

The next generation of founders is working for someone else today

For years, the startup ecosystem looked for founders in very specific places: young people, people willing to drop everything, people able to take on extreme risk. That profile will continue to exist. But maybe the next generation of great founders is somewhere else. Maybe they are working for someone else today.

Maybe they are inside a telco, a bank, a consultancy, an industrial company, a Big Four firm, a scaleup, a hospital, an energy company, or a corporation that has spent years accumulating problems no one has yet turned into product. Maybe they did not lack ideas. Maybe they lacked a model that did not force them to leap into the void before having evidence.

The old game said: quit your job, raise capital, grow fast, and chase a venture-scale exit. The new game can say: build before raising, validate before jumping, use AI to lower the cost of trying, and decide whether venture capital makes sense once you have more evidence.

The future of entrepreneurship will not be anti-VC. It will be VC-optional.