Elad Gil didn’t grow up in Silicon Valley. The son of a nurse and an academic studying education psychology, he grew up akin to a military brat, moving between Michigan, San Diego, Israel, and Arizona in his early years.

But he seems to have a golden touch when picking startups that turn into unicorns or decacorns. Gil, who typically makes bets in seed-stage companies, has cut checks to some 25 startups that have since reached the vaunted unicorn status, including Instacart, Stripe, and Coinbase.

And there's more: Other companies in his portfolio such as PagerDuty, Square, and Pinterest are already trading in public markets, while more, such as Airbnb and Wish, are expected to IPO this year.

Now billed as a "startup helper" who advises new companies as they seek to get off the ground or scale, Gil has worked to start Google's mobile efforts and help Twitter scale from 90 employees to 1,500. He also co-founded a genetic testing startup, Color Genomics.

While many individual investors are considered angels that put in small amounts of capital early on, Gil is also often cited as a part of trend dubbed "solo capitalists"—a class of investors who are putting in large checks even in later stages and compete with traditional venture capital shops for highly desirable investments.

Fortune caught up with the investor to discuss how he finds and evaluates investments at a stage when many companies look very different from their final form—and why by definition, the largest and most successful startups seem dumb or low-end at first glance.

Here's our conversation, edited for clarity:

Fortune: Tell me about yourself—how did you end up in tech investing, of all things?

Elad Gil: I was getting a Ph.D in biology—focused on human diseases-related areas, for example virology, cancer, aging—and I realized partway through that I didn’t want to become an academic, partly because you get thrown into a lab for a long time. I was thinking through what is the biggest way to have an impact and do something useful for people.

At the same time, I was interested in technology. I think we forget today how hard it was to do so many things: When I was in high school, I mailed random mathematicians because I was really interested what they were doing, and they would then send back those yellow envelopes with print outs of their papers. Now, oh my gosh, suddenly any kid can go online and read every single mathematics paper that is relevant in the world.

So I moved out to the San Francisco Bay area to join the startup world near the end of the dot-com boom. I still have an old Silicon Valley ethos of wanting to do useful things for people and to impact the world positively—unfortunately that’s become the type of thing that is made fun of in popular media, like Silicon Valley.

You generally invest at seed stage—when a company's business plan may still be unclear. So what do you look for in founders?

I think the biggest determinant of success tends to be the market. I’ve seen really, really good teams in terrible markets get completely crushed. I’ve also seen really bad teams in good markets do really well. So first and foremost, I care about if you’re building something interesting in a market that is large enough.

It’s pretty rare for me to invest in a team without an idea—usually I wait until people are working on something specific so it's easier to validate whether that market makes sense. Sometimes you see teams that have built the same thing over and over again in their career for different companies, and they decide to build it as a piece of infrastructure, like PagerDuty.

On whether you're building something interesting in a market large enough, though—that's really hard to tell a priori because, if a market is big and interesting and obvious, then everyone would already be doing it. So the startups that do best almost by definition have to be in non-obvious markets. People may think it's toyish, or dumb, or super low-end—so often things start off looking weird, and you [as an investor], may wonder, “Why would you really want to do that?” And you need a bit of convincing.

So was that the case for your investment in Airbnb? I remember that Andreessen Horowitz’s Managing Partner Jeff Jordan said he thought the idea was crazy when he first heard of it.

I invested in Airbnb, I think, when it was just eight people, and they were still working out of Brian Chesky’s apartment at the time. I knew them previously, and they pinged me while raising their Series A. It actually seemed like a really good concept and idea.

At the time, I think there were all sorts of signs that something like this should work. One: The product was beginning to work on their side. And before that, when I was in undergrad and a little during grad school, I had travelled through a [travel] service called Servas, a service born out of the Esperanto community in World War II. Their basic belief was that if you could stay with people in a foreign country and learn about them directly and vice versa, we would be more likely to have world peace. So you’d sign up for the service and they’d interview you and make sure you weren’t crazy or anything. Then you’d visit a country and they’d give you a printed booklet with the names of all the people that would host you for free in that country—with rules, of course, around needing to contact them a week or a night ahead. So you could literally show up in Rome with this booklet and just call or email anyone in it to stay with them.

So when I saw Airbnb, I thought: Wow, this is Servas but you can monetize it.

But you do also think the founders and their characteristics are important.

Founders are really important. Beyond signals like perseverance, grit, and ethics, you ask: Are people learning rapidly? Are they good at sales? Not only can they sell to new customers, but also can they convince people to join them and can they raise money? You need at least one part of the founding team to be good at selling, and one part to be good at building, and later in life of the company, you need someone who is good at tight operations.

If I were to use Apple as an analogy: Steve Wozniak could build anything. Steve Jobs could sell anything. Tim Cook really ended up running the operations from the back end and scaling the supply chains.

There seems to be a new trend around "solo capitalists"—single-person “venture firms” that are able to write checks faster than their traditional counterparts. You are considered one of those solo capitalists. Why do you think it is happening?

What people now call "solo capitalists" actually has a pretty long history. Arthur Rock was the original investor in Apple and in a bunch of other great companies including Intel. Originally, he was part of a two-person outfit called Davis and Rock, but he eventually went out on his own in the '80s and was just investing as what we would now call a solo capitalist.

But now there is a set of individuals who are investing and competing more aggressively for Series A, B, C, and beyond—competing later in the life of companies rather than just the seed or the A.

One shift is that companies have stayed private for much longer. Now companies take 8 to 12 years to go public, but founders could still use help from operator investors, which many solo capitalist are, as their companies are scaling and going into growth spurts—which is often when everything is starting to break at a company.

I also think this is a trend because there’s been a sudden shift in which networks are really important. I feel like this happens every five years or so. Some networks that were relevant before are still relevant today, like Y-Combinator or Google. But there are a bunch of companies that aren’t as relevant anymore and are being replaced by, say, Stripe—which is a really important founding network—or Opendoor, Coinbase, Uber, etc. There are more and more founders [and investors] emerging from these next waves of companies.

That said, I do think there are some traditional late-stage funds that are almost late-stage dressing for people acting as solo capitalists—I don’t want to out them (laughs)—but there really are some funds with just one main partner and a bunch of junior people.

What are the networks that are less relevant today for startup formation?

There are some companies that are less central now—and then there are some that have, oddly enough, never spawned anybody, or spawned very few companies.

There’s very few companies, for example, that have come out of Amazon, Microsoft, or Apple. It’s so weird, right? And it’s part of the reason why Seattle, while an interesting city with tons of great engineers, hasn’t created a whole lot of startups in the last twenty years. There are many interesting companies up there, but it’s not reflective of the fact that Amazon and Microsoft are anchor companies.

There are sometimes small companies that have spawned half a dozen interesting companies and other dynamics. Look at the PayPal network: Every single company founded by someone ex-PayPal that hit scale was founded by a former executive. For some reason, the executive team at PayPal was amazing at starting these massive, important companies, but the rank-and-file employees, with the big exception of YouTube, didn't start much. Why is that? It could be the culture, the decision-making, the people they attract—I have no idea why. There are really strange dynamics that I don’t really understand but I think are fascinating to think about.

You are also an investor in Coinbase, the crypto exchange. Its CEO Brian Armstrong recently came out publicly with a divisive memo calling employees to refrain from politics. What is your view on how that could impact the company as it seeks to grow?

I think that question's best directed at the Coinbase team. I wasn’t involved in the decisions around that post and I don’t have insights relevant to it.

So to you, it’s not a question of whether or not he should have posted it, but that he had the right to do so.

I think, fundamentally, it is up to the founder and/or CEO of the company to determine what they want the employees of their company to focus on from a mission-driven perspective. And it is really up to the founders, really early on in the life of the company, to define that mission, and to some extent you define that mission over time. Each CEO has it within their purview to define the focus and mission of the company. And if people want to have a broad mission for their company, they can define it that way. And if they want to have a very narrow mission for their company, they can define it that way too.

What are some of the areas you are interested in investing today? You've previously pointed to defense, but you also think there is space for new social media companies—for example, you invested in Clubhouse.

About a year ago, I read a blog post about interesting things that should happen in the social-media world. Number one, there was a technology shift—so WebRTC and WebGL were baked into the browser—and WebRTC allowed for really crisp audio and video that you can build really great video apps on.

Secondly, you have a generational shift. People don’t want to use their parents' social network, and I think in five to eight years, there will be turnover on certain networks and I think we are ripe for one. And now with COVID, people crave social interactions so you have more room for experimentation and adoption.

I’m trying to keep an open mind on it—because something like Clubhouse I would’ve never predicted. I never a priori would have said an audio-only format [is the next big feature]. Some things aren’t that compelling until you try it. So I’ve been scouring obscure parts of the webs and playing with them. I think there are finally things happening with, say, in-browser VR—but again, it’s very early.

Any big misses in terms of picking out investments?

Oh, a bunch. I missed Lyft. I was a little worried about the valuation at the time and competition with Uber. At the time there was a really strong perception that every strong technology market would be winner-takes-all when in reality, many markets are oligopoly markets. I just misunderstood the market structure and the valuation.