At Web Summit Vancouver, Techstars Founder and CEO David Cohen took the stage to talk about how the startup landscape is shifting and what it takes to build a defensible company in the age of AI.

A few themes stood out for founders, and they should shape how you think about where you build, how you raise, and what makes your company hard to copy.

Founded Everywhere, Funded in Clusters

Fifteen years ago, unicorns came out of four countries. Today, the number is closer to 60. For five years running, a smaller share of new companies has been founded in the San Francisco Bay Area, with startup activity spreading to specialized hubs. Space companies are clustering in Los Angeles and Denver. LLM builders are still in the Bay Area. Founders applying AI to a vertical are increasingly everywhere else.

The catch: while founding is decentralizing, funding still concentrates. Roughly 80 percent of capital flows into mega rounds, most of them in the Bay Area. Strip those out, and the rest of the market is moving to more places than ever.

Software Is Close To Free. Defensibility Is the New Bar.

David's line cut through the noise: “What are you doing that a kid with a laptop can’t copy this weekend?” Beautiful software, on its own, no longer wins meetings.

What counts as a real moat?

• Proprietary data loops that get stronger with every customer

• Regulatory hurdles already cleared

• Humans in the loop who matter to the solution

• A physical connection to the world (hardware is hot again)

• Verticals with built-in defensibility, like healthcare, fintech, and space

If you are raising right now, expect every conversation to come back to this. Investors want proof, not claims. Show the structure of your data, how much you collect per customer, how it feeds back into your model, and why no one else can get it.

The Org Isn’t Ready for What the Software Can Do

A second tension David surfaced: founders can ship features faster than their customers, partners, or even their own teams can absorb them. Companies are throttling capability on purpose, waiting for the market to catch up. Speed of building is no longer the constraint. Operationalizing is.

What Techstars Looks for in Founders

With 25,000 applications a year and climbing toward 50,000, Techstars is filtering for a few core traits:

• Founders who do stuff, instead of waiting on capital to act

• An intrinsic why that will outlast hard days

• Team, market, and progress over the specific idea, because the idea will shift

One Last Thought

Forget the one-person billion-dollar company. David is more interested in the zero-person unicorn, and he thinks it is coming. The takeaway for founders is not that humans matter less. It is that the bar for defensibility has moved, and the founders who internalize that early will be the ones building the next wave.

Read More