Year after year, vertical-specific investors become increasingly central to the venture world.

Andreessen Horowitz’s American Dynamism practice has carved out a remarkable brand in the “world of atoms.” Paradigm first made a name for itself with bold crypto bets. Numerous funds are currently spinning up to take advantage of the AI gold rush specifically.

It’s understandable: As the venture world becomes increasingly competitive and rich, investors need to build their teams (and brands) to be as targeted and high-impact for entrepreneurs as possible. Choosing a vertical makes commanding capital from LPs more straightforward as well.

And this strategy has worked. The investors mentioned above are often at the top of founders’ fundraising lists. It’s no surprise that other firms feel increasingly pressured to show off their expertise in various spaces by publishing market maps and investment theses.Is this trend the future? Will every possible startup soon have an “expert” VC to seek out in their space? I think not; generalist investing has rightfully been (and will always be) the primary mode of VC.

Let’s get a few assumptions out of the way. “Generalist” investing does not mean lack of technical knowledge. It does not mean a lack of preference of some verticals over others. And it certainly does not mean unsophistication in their network.

The eternal relevance of generalism in venture comes down to two simple and easy-to-prove facts: (1) Revolutionary tech companies are thematically unpredictable, and (2) transcendent founder talent is still needed even in the most fruitful spaces.

Looking empirically at some of the greatest investments of all time, it’s difficult to see how concentrating on one theme would have converted to a winning deal. Google was following several preexisting search engines. Facebook had already lost to MySpace as far as anyone could tell. UberCab was a small market, and ride-share wasn’t envisioned from the start. Clearly, the historical trends around consumer internet, social media, and the gig economy were unseeable at the time, even if hindsight feels obvious.I don’t buy that any forward-looking trend today is more than directionally right. The green tech theses of the 2000s yielded almost nothing at all. The “sharing economy” themes of the 2010s never quite replicated after Uber/Lyft/Airbnb created the concept. This sets a high burden of proof for any conception of vertical-specific investing in the 2020s.

Even today, it’s unclear which vertical-specific firms have capitalized on the emerging winners in the AI space — has OpenAI or Scale led to a firm-returning (not just fund-returning) outcome for an AI-focused investor as much as it has for generalist investors?

But of course it’s easy to paint an extreme picture. In reality there are plenty of excellent firms that nail their specialized vertical. In any case, the problem of “founder picking” persists. And it’s quite the challenging problem.While some spaces like healthcare or defense may have returns 2x or even 10x greater than other verticals, the best startup within those spaces could have returns 100x greater than its more typical counterparts. The talent of founders and outcomes of their startups have unimaginably more variance than any other dimension. Thus emerges the paradox: The specialist investor must have the same eagle eye for outliers as the generalist investor, but with narrower aperture.

And if, as a firm, you are able to cultivate that eye for outlier founders, why not apply that rare skill across the entire opportunity set? The next Google might be somewhere you don’t expect.

Now, readers may wonder about adding value to founders and winning deals while lacking any sort of specific expertise. How can a generalist compete with a specialist when it comes to customer introductions, or tactical advice?My answer is simple: We can’t, but we don’t need to. The specific product and marketing challenges faced by entrepreneurs tend to be far too specific for an outsider to properly help. Introductions in “supply chain” are insufficient compared to introductions to “decision makers at roofing supply distributors with $100 million to $500 million revenue who use Epicor as their system of record” as one portfolio company recently requested. Which VC has a roster like that lined up and ready to go?

Ultimately my job as a VC is selling a few simple products: cash, trusted reputation, and access to as much downstream capital as possible. And I can do that just as well as anyone who invests in only AI.

To be clear, the venture world is evolving fast, and specialist investors play a valuable role. But I urge my generalist friends to be proud. With a radical focus on outliers wherever they may be, history is on our side.