AI democratizes entrepreneurship and company creation, but it concentrates capital in fewer, bigger winners. The end state is a venture industry that becomes more exclusive, not less.

Both halves of that sentence sound reasonable on their own. Put together, they make most investors uncomfortable, because the second half undoes the business model that the first half seems to promise.

I think about technology through the cost of experimentation. Every major platform shift lets entrepreneurs achieve more with less. I wrote about “The impact of AI on the cost of starting and running a business” early 2023 here.

In short, the internet dramatically lowered the cost of distribution. Mobile lowered the cost of access, putting a computer and a distribution channel into billions of pockets. AI is now lowering the cost of creation itself.

Software, content, research, design and increasingly entire workflows can be built by small teams at a speed and quality that would have required significantly more people and capital just a few years ago.

As the cost of creating anything in front of a computer approaches zero, we should expect many more people to create. And that is exactly what the data shows.

Nasdaq's Economic Institute found that solo business applications in the US rose more than 20% since early 2025, while applications from companies likely to hire stayed flat. Nearly half of the solo growth came from high AI-adoption sectors like technology, finance and professional services.

Stripe's economics team sees the same pattern on the revenue side. Across solopreneur platforms, the number of one-person businesses earning $5M+ and $10M+ nearly tripled between 2023 and 2025. And among all new businesses on Stripe, roughly 30% more of the 2025 cohort crossed $1M in cumulative revenue within their first year than the 2023 cohort did.

More founders. Smaller teams. Faster to revenue. That part of the thesis is no longer a prediction, it’s a reality.

That’s at least my personal belief.

The common assumption is that a boom in company formation mechanically produces a boom in venture-backable companies. I believe the opposite is happening.

AI makes it easier to build and enter a market. But the same technology also allows the best companies to scale faster than anything we have seen before. Capital, talent and customers increasingly concentrate around a small number of category leaders.

The numbers are already extreme:

Record money. Fewer companies.

Said differently, aggregate private market cap is growing but the number of companies and funds benefiting is shrinking.

Watch how quickly the vocabulary has moved.

A decade ago, everyone was talking about unicorns. Five years ago, decacorns became the benchmark. Two years ago, investors started talking seriously about centacorns. Today, trillion-dollar technology outcomes are part of the conversation with SpaceX, Anthropic, and OpenAI in the lead, and the 30 most valuable private companies now worth a combined $3.9T.

I expect this power law to become even more extreme. More companies will be created, but a smaller share will capture an ever larger portion of the value.

For venture investors, that means more capital chasing fewer true outliers, and ownership in those winners becoming increasingly expensive.

If you can build a great business with very little capital, but you are unlikely to become the global category leader that venture requires, then raising traditional VC makes less and less sense.

The funnel data supports the caution. Of startups that raised a seed round of $1M+, graduation rates to the next round have fallen sharply: 55%+ through 2020, 24% for the 2023 cohort and 16% for the 2024 cohort. The recent cohorts are still young and some will still raise, but the direction is unmistakable.

Taking venture money commits you to a specific outcome distribution. If AI lets you reach $5M or $20M of profitable revenue with a team of ten, that commitment might be a bad trade.

I therefore expect bootstrapping, seedstrapping and other capital-efficient paths to become the default for a growing share of ambitious founders. And one question I keep asking founders for pitch me becomes increasingly important: Are you sure you want to take the venture route and work towards a binary outcome?

So where should capital concentrate over the next years?

Win access to the obvious outliers, or find the hidden gems before anyone else does. Everything in between will become increasingly difficult.

At one end are the companies everyone wants to own. As capital concentrates in fewer, bigger winners, access becomes the scarce resource. Not capital, not information, not even conviction.

This favors firms with a powerful venture flywheel: brand, reputation, track record, networks and expertise create access to the best founders. Backing those founders creates returns, references and reputation, which further improve access to the next generation of outliers.

This flywheel compounds. It is increasingly difficult to compete with, and increasingly difficult to build from scratch.

At the other end are the companies with outlier potential before the market recognizes them as such.

Here, the advantage comes from focus. Narrow the search space by stage, geography and/or sector, then combine unique sourcing, deep expertise and differentiated judgment to see potential before others do.

Smaller, highly focused funds can build a real edge here and earn meaningful ownership before an outlier becomes obvious. This is also the group of investors where I expect the most product innovation in venture capital to come from.

The best investment firms might even find ways and structures to play on both ends of the spectrum.

But if you have neither privileged access to the companies everyone knows are exceptional, nor an edge in finding them before everyone else, it becomes increasingly unclear what you are being paid for.

As technology commoditizes access to information, sourcing and basic diligence, this middle ground gets harder to defend every quarter.

AI is doing two things at once.

It is making company creation radically cheaper, which produces more founders, more experiments and more real businesses than we have ever had.

It is also making the winners bigger and faster, which pulls capital into fewer names and prices ownership in them out of reach for most.

More companies. Fewer venture outcomes. A widening gap between the firms that can reach them and the firms that cannot.

Pick your end of the barbell deliberately, and build the edge that end actually requires.

Stay driven,
Andre