This deep dive expands on the Loop Asia conversation with Chris Birrell, Managing Partner at Func Ventures.

Partway through the conversation, Chris Birrell states a claim almost as an aside: the interface layer of software is commoditizing, and the real value is moving into what's underneath — the workflow, the transactions, the physical actions triggered in the real world. It's the load-bearing idea behind his entire investment thesis.

The argument, borrowed loosely from Clayton Christensen's work on disruptive innovation, is that in a maturing technology stack, advantage and margin migrate away from the interface a customer sees and toward the subsystem doing the harder, less visible work. Birrell's reference point is the PC industry, which he watched from inside as a CIO with real purchasing power: Intel, AMD, and Microsoft captured the overwhelming share of value created by the PC boom, while the physical box — the thing with a brand name on the front — became a commodity assembled from parts anyone could source. The chip and the operating system, the parts nobody outside the industry thought much about, were where the money actually was.

Birrell's bet is that SaaS is going through the same transition, one layer up: the dashboard and workflow builder — the visible, brandable part of enterprise software for two decades — is commoditizing fast, as AI copilots and low-code tooling make a competent front end dramatically cheaper to build. What doesn't commoditize as easily is the layer beneath it: the API that moves the money, verifies the identity, routes the message, or triggers the action in the physical world. It's also why he's built a fund around never wanting to be "another AI VC" — everyone chasing the interface layer of the AI wave is competing in the layer he thinks will commoditize fastest, because the barrier to building a plausible interface has collapsed. He wants to be underneath that, in what he calls the orchestration and API layer, where whoever owns the actual capability keeps the pricing power as the interface calling it becomes interchangeable.

There's a test embedded in this thesis: could a consumer, prompting their own chatbot, eventually call your API directly, with no interface in the loop? He raises it half as a joke about someday prompting a refrigerator, but the serious version is worth sitting with. If the honest answer is "yes, eventually," the interface was never the moat — the API underneath it was, whether the company realised it or not.

It also explains why Birrell is a committed B2B investor, and treats that as a choice rather than a default: consumer products live and die by interface — brand, design, habit formation, exactly what Christensen's theory says gets commoditized first. B2B infrastructure, especially infrastructure plugging into other systems rather than facing an end user, is less exposed — the buyer isn't evaluating how it looks, but whether it survives the next budget review.

The thesis has an obvious limit, and Birrell doesn't pretend otherwise: subsystems hold value only as long as they stay genuinely hard to replicate. Christensen's later work is as much about the eventual commoditization of a stack's low end as about value migrating there first — the chip and the operating system held PC-era value for decades, not forever; ARM and cloud computing have been redrawing that map for years. The API layer Birrell is betting on will face the same pressure. His answer, embedded in the "solving a difficult problem" criterion atop SUPERIOR, is that the only durable position in a stack that keeps sliding downward is to keep solving the problem nobody else has fully solved — and re-earn that position as the stack moves again.