The Hidden Cost of Easy

AI has dramatically lowered the barriers to creating new companies. A couple of decades ago, building a startup required substantial engineering, a sort of “proof of work” baked into the system. The advent of cloud infrastructure reduced initial investment needs but still depended on strong technical execution. Today, anyone can select a vertical and build a thin layer of software with less overhead than ever before. And because the target customer is likely to be less savvy, they might misperceive the offering as revolutionary. Essentially, founders often get credit for building something primarily constructed by OpenAI.

As a result, many of these companies show early revenue growth that might have secured investment just a few years ago. Yet, as an investor, I remain cautious. I believe we are witnessing market fragmentation at unprecedented rates. Every hustler in America is looking for opportunities to monetize AI, quickly capitalizing by selling into their local networks. However, for every AI-entrepreneur in Chicago, there’s another in Boston, New Orleans, and countless other cities, each rapidly penetrating their respective local markets. Eventually, these local networks will saturate, revenue growth will stall, and private equity firms will have a field day rolling up hundreds of similar businesses.

I believe the solution lies in embracing difficulty. Of course, there is already an investing category known as “hard tech” (e.g., aerospace). But I’m referring to something broader, inclusive of traditional software—solving genuinely hard problems involving deep integrations, complex regulations, acute precision, and high scalability. These endeavors hold value precisely because they’re difficult to replicate. The true prize here is hard-won technological depth.