When Richard Raizes and I began investing in Web3 technologies, we spent years refining a universal framework for evaluating opportunities. Richard always said, "Adoption is everything," and that became the cornerstone of our approach.
Instead of being Bitcoin or Ethereum maximalists, we focused on being adoption maximalists, which helped us uncover and invest in misunderstood and underappreciated opportunities.
Nearly a decade later, we are now hard at work figuring out how our framework must evolve to account for the unique opportunities and challenges of investing in intelligent technologies.
We believe that the two key additions to our investing framework should be data and distribution (more on this later), especially as the goal is to identify under-the-radar opportunities in intelligent technologies.
To demonstrate why understanding an organization’s proprietary data is so important, let’s consider a thought experiment:
If I could hire the entire Tesla team and give them unlimited resources, could they recreate Tesla’s Full Self-Driving (FSD) software for my company?
Most people would assume the answer is yes. However, I’m convinced it’s no.
While the neural network and hardware might be replicable, the data used to train Tesla’s models is proprietary and lies at the core of what differentiates their FSD. Tesla likely already has an insurmountable lead, with billions of miles of clean, proprietary video and sensor data feeding its models.
A lot of people debate whether Tesla will win the autonomous vehicle race. But what if it has already won, thanks to the deep and likely insurmountable data moat it has built and continues to expand?
What many investors overlook is that data becomes more valuable the more specialized it is. Proprietary data, gathered over time and in the right context, creates a compounding advantage that is incredibly hard for newcomers to replicate.
Many companies sitting on untapped goldmines of proprietary data aren’t even recognized as AI or tech companies today. Across industries like manufacturing, healthcare, logistics, and finance, these companies have accumulated years or even decades of domain-specific data.
These under-the-radar data giants may possess some of the most valuable raw materials for the intelligent age. The future of intelligent technologies lies not just with companies creating new AI algorithms, but with those that own the data that makes those algorithms truly intelligent.
But data alone isn’t enough. Organziations that can scale their AI-powered products globally through established distribution channels will have an edge over those that merely possess the best technology or datasets.