Back to the Future

The number of startups doing really hard stuff has skyrocketed recently. Bitcoin, drones, food substitutes, 3D printing, bioinformatics, deep learning. We used to see these “science fiction” or “moonshot” startups only occasionally. Now it feels like every other startup (or at least 30%) is working on something pretty wacky.

These ideas are like old school Silicon Valley startups: Fairchild Semiconductor, Intel, Sun and Silicon Graphics. In order to work at those places (let alone start one), you needed to have one big brain – and back then, that was usually validated by an advanced technical degree. Deep technical expertise and know-how (i.e., being a smart motherf*cker) created a barrier to entry.

Similarly, the startups in today’s “moonshot” areas require unique intelligence, dogged patience and deep domain expertise (or at least, the capacity and patience to acquire that expertise). And maybe it’s not raw intelligence but also a lifelong learning and obsession with a particular area that can’t be mastered in a few months: biology, mechanical engineering, advanced mathematics. The “patience” part is particularly true because it’s more likely that these technologies are overhyped in the short-term but under-hyped in the long-term.

As an investor, these startups are both easier and harder to evaluate. On the one hand, the deep domain knowledge can be used as a filter to screen startup opportunities and it’s easier to identify. On the other, it’s hard to understand the technologies/opportunity and easy to forget that “hard” doesn’t always mean “valuable.”