Neil Rimer thinks the AI money is coming back out

We have all spent the last two years watching a select few companies turn into nation-states. Between the hyperscalers providing the compute and the labs training the foundation models, the amount of capital flowing into a tiny sliver of Silicon Valley is unprecedented. But if you talk to the people who have been through multiple cycles, they will tell you this level of concentration usually breaks. Neil Rimer, one of the veterans behind Index Ventures, is starting to say the quiet part out loud: that money is going to have to come back out.

The Gravity of Capital

In every tech boom, there is a period of massive consolidation. We saw it with the early internet and again with mobile. Right now, AI is in its hoarding phase. The giants are sitting on mountains of cash, GPU clusters, and proprietary data. For a founder trying to build something today, it often feels like you are just a tenant on a plot of land owned by Microsoft or Nvidia. You pay your rent in tokens or compute credits and hope they do not decide to build your feature set into their next update.

Rimer’s perspective is not just about fairness; it is about systemic sustainability. When wealth concentrates this heavily in one corner of the economy, it creates a vacuum. The rest of the ecosystem starves for liquidity. He suggests that this capital will eventually be redistributed, whether that happens through voluntary reinvestment, market corrections, or the less pleasant route of regulatory intervention.

What This Means for the Builder Class

If you are building a startup right now, you should not be looking at the billions flowing into foundation models as a sign of where you should be. That is old news. The real opportunity, and where Rimer seems to be pointing, is in the redistribution phase. This is the moment where the infrastructure has been built and the tools are being handed down to the people who actually solve problems for humans.

We are moving away from the era of training and moving into the era of application. For a founder, this means the goal is no longer to be the biggest fish in the pond, but to be the most efficient user of the resources the giants have already paid for. The wealth redistribution Rimer talks about often looks like lower costs for compute, better access to open-source models, and a broader market for AI-driven services that do not require a billion-dollar balance sheet to operate.

The Threat of Involuntary Redistribution

There is a darker side to this prediction. Rimer mentions that redistribution can be involuntary. In the tech world, that usually means the government gets involved. We are already seeing the early rumblings of antitrust cases against the major players in the AI space. The concern is that if these companies do not find a way to share the surplus they have created, regulators will do it for them by breaking up their monopolies or taxing their massive gains.

For those of us on the ground, regulation is a double-edged sword. It can provide a level playing field, but it can also stifle innovation by creating a bureaucratic mess that only the incumbents can afford to navigate. The ideal scenario is a market-driven redistribution where the giants start acquiring smaller, specialized companies again, or where the cost of entry drops so low that the monopoly power of the big labs fades away naturally.

The Skeptic's View on 'Voluntary' Sharing

I have spent enough time around founders and VCs to know that nobody gives up a competitive advantage voluntarily. If the AI giants are going to let wealth flow back into the ecosystem, they will only do it because they have to. Maybe they realized that their models are only as good as the applications built on top of them. Maybe they realize that a stagnant economy cannot afford to buy their products.

The takeaway for builders is simple: stop chasing the hype of the foundation. That ship has sailed and the winners are already sitting on the deck. Your job is to build the tools that make use of the secondary and tertiary effects of this wealth. When the money starts leaking out of the big labs and into the specialized sectors—healthcare, logistics, local governance—that is where the real founders will make their mark.

The Long Game

Rimer’s thesis is a reminder that we are still in the early innings. The current concentration of wealth is a feature of the development stage, not the final state of the industry. The history of technology is a history of cycles. The giants build the roads, and the builders drive the cars.

As the capital starts to move, keep your eyes on the edges. The most interesting things in crypto and AI right now are not happening in the boardroom of a trillion-dollar company. They are happening in small Discord servers and shared offices where builders are figuring out how to take the raw power of these models and turn them into something that actually works for a specific group of people. That is where the redistributed wealth will land.

The question for every founder today is not how to compete with the giants, but how to be ready for when they are forced to let go of the keys.

• Watch the margins: As foundation models become commodities, the value shifts to the interface and the data.
• Ignore the noise: Billion-dollar rounds for infrastructure companies are a lagging indicator, not a leading one.
• Build for humans: The giants are building for the machine; the wealth will flow to those who build for the user.

We are entering a period where honesty about the market matters more than optimism. Rimer is right to be cautious, and we should be too. But in that caution lies the roadmap for the next generation of builders who are ready to catch the spillover from the AI explosion.

Read the original at TechCrunch AI →