The Podcast

Adam Link spent more than a decade in tech, most of it in crypto, including years at Coinbase while it grew from a startup into a public company with thousands of employees. He has a bachelor’s in finance and a master’s and doctorate in computer science, and he’s started and sold companies, invested as an angel and as a VC, and led large engineering teams. Today he runs Fireweed Capital, a financial planning firm based in northern Minnesota, and a lot of his clients come from tech.

The idea I keep coming back to is Adam’s argument about execution. Founders have always heard that ideas are worthless until someone builds them, and Adam says AI has now made the building cheap too. He asked Claude to plan a CRM around the features he liked in the leading financial-services CRMs, let it build over a weekend, and was running his firm on it Monday morning. If a customer can rebuild your product in a weekend, the software can’t protect your business anymore. Adam thinks that protection has to come from the work around the software: support, sales, delivery, and the industry relationships that take years to build. His example is logistics software, where the company selling it may have to start running trucks.

That changes what Adam looks for as an investor. Silicon Valley used to turn away any founder without a technical co-founder, and Adam expects the reverse now. Claude becomes the technical co-founder, and the founder with 20 years in an industry, who knows a real problem and can land the first orders, becomes the one worth funding. The gaps worth building in are the ones an AI agent can’t close by connecting two systems through an API, the connection that lets one program talk to another.

Adam has already acted on that view with a marketing platform he pays for. The platform has no public API, so loading his clients meant uploading a spreadsheet by hand. He had Claude open a browser session, watch the calls the platform’s own interface sends to its servers, and map the private API underneath, then built his own tool on top of it. The vendor’s support team told him he isn’t allowed to do that. Adam’s answer is that he’ll use far more of their product this way, and by his estimate, companies that hide their API behind a UI have six to twelve months before customers like him, or a competitor with an API, take that choice away from them.

We also get into why big companies struggle to fund new ideas, since process replaces the relationships that let a startup founder get the benefit of the doubt. Adam explains how Coinbase paired an internal startup program with its venture arm, so employees who left to start companies could still get funded and often came back through acquisitions. He also lays out where he thinks crypto is headed: out of sight, as the way people hold and trade ownership in companies too small to go public.

Takeaways

Here’s a teaser…

On Linkedin.

What You Missed on Sunday

Here’s what we covered in Sunday’s newsletter edition…

In 1985, researchers asked 2,994 people who’d just started businesses what their odds of success were. 81% said 7 in 10 or better, and a third said 10 in 10. Asked about the odds for someone else’s similar business, the same people averaged under 6. The lower number was the realistic one: about half of new businesses close within five years.

Most of us rate our own odds in innovation just as generously. Every change produces a victim, and the people who approve it go ahead because they hope they’ll be on the winning side.

This week, I’m looking at how that hope lets people sign off on someone else’s loss without remorse, and at the question anyone approving an innovation should have to answer: if this succeeds, who gets hurt?