A founder choosing what to build today is making a bet that may take years to mature. The financing market can change long before the product is ready. That creates a practical problem: how do you distinguish a durable opportunity from a sector that happens to be attracting the largest checks?

AI makes that problem unusually visible. Funding can be enormous at the top of the market while remaining difficult to access for an individual company. Meanwhile, energy and industrial businesses are raising capital to build products whose economics depend on wells, factories, materials and customers willing to buy physical output.

So, is AI funding drying up? The data says no. It is cooling from an extraordinary peak and is heavily concentrated in a few giant rounds, but AI still takes most of the world’s venture capital. The more useful question is what comes after the boom: what a broader, post-AI venture market could reward, and what founders should build for it.

To answer that second question, we look at two physical-technology companies, Fervo Energy and Form Energy. Their decisions about repeatability, buyers, manufacturing and financing show what investors in the next cycle will want to see.

Crunchbase reported $114 billion of global funding to AI companies in 2024 and $211 billion in 2025, roughly 85% growth in a single year. [1]

The more recent picture is different. Crunchbase recorded $242 billion for AI in Q1 2026, more than in all of 2025, and $102 billion in Q3, 57.9% less. [2][3]

That drop is real, but it is a fall from an exceptional quarter, not a collapse. Even in Q3, AI received 64% of all global venture funding. The $102 billion it raised is close to what the entire global venture market raised a year earlier: $104 billion in Q3 2025. And Q3 2026 set an all-time record for billion-dollar rounds, with 27, up from 14 in Q1. [3]

Activity also held up below the biggest deals. Roughly 6,000 startups raised money in Q1 and close to 6,000 in Q3. [2][3] The dollars fell; the number of companies getting funded did not.

For the venture market as a whole, Q3 was down 25% from Q2 but up 53% from a year earlier. [3] A market can look weak against its recent peak and strong against the prior year at the same time.

That distinction prevents two errors. One is treating every lower quarter as evidence of a structural collapse. The other is using record totals to dismiss the possibility that fundraising is getting harder for some founders. Neither conclusion follows from a single aggregate.

In the first half of 2026, OpenAI and Anthropic alone raised $217 billion of the $510 billion invested in startups worldwide, across every sector: 42.5% of the total. [4]

When a few companies account for so much activity, the calendar of their rounds can dominate the market narrative. A quieter period for those companies can pull down the total without telling us much about seed investors or applied businesses. Conversely, another exceptional round can conceal weakness elsewhere.

Concentration also doesn’t mean these rounds took money away from other founders. A real drought would look different: repeated declines outside the largest rounds, fewer funded companies and harder follow-on rounds. So far, the number of funded startups has held steady, which is not what a drought looks like.

So, is AI funding drying up? No. It has fallen from an exceptional peak driven by a handful of giant rounds, but AI still captures most of the world’s venture capital, billion-dollar rounds hit a record in Q3, and thousands of startups keep raising every quarter. That doesn’t make the current cycle permanent. It means founders have room to think about what the next one will reward, without pretending this one has ended.

If the money isn’t running out, why talk about a post-AI market at all? Because the companies that win the next cycle are being built now, while attention and capital are concentrated elsewhere. “Post-AI” doesn’t mean companies stop using AI. It describes a market in which venture investors back a wider range of technological advantages, and in which money leaving one category doesn’t automatically flow to another: it has to be won. The rest of this piece looks at where that opportunity is forming and what founders should build to capture it.

Artificial intelligence will remain profoundly important. I see it as an extraordinary milestone for humanity—one that will change the course of history by expanding what people can discover, create and build. A slowdown in funding does not diminish that significance. The next venture cycle will take shape in a world already being transformed by AI, with opportunities both in the technology itself and in the physical systems that turn new capabilities into lasting economic value.