We always end each PandoMonthly with two questions. The first is what you believe that no one else (or few people) in the world believe. A lot of people — exhausted from the ringer we’ve put them through over the last two hours — kinda phone that one in. No one more so than Chris Sacca who essentially said: Eh, I think people have heard enough of my views by this point. (Indeed, that was our longest interview to date.)

But our guest last night was AngelList founder Naval Ravikant, and he came prepared for this question. And his answer was one of my favorite moments of the evening. He came back to a theme he’d touched on earlier in the interview, about how the world would be increasingly made up of very small startups interacting with each other through APIs. No big corporations.

To see how radical this idea is, look around the startup ecosystem. All of the most promising Web companies have done mega growth rounds at huge valuations. Facebook raised well over $1 billion in private equity before even going public and employs thousands. Even VCs and entrepreneurs who believe startups can change the world, believe you have to get big eventually. Sure you can be capital efficient at the beginning, but not at the end of a journey.

Ravikant argued that Instagram wasn’t a fluke — it’ll start to be the norm. He countered criticism that Instagram didn’t monetize by saying they only would have needed a handful of people to do it. And he went a step further, saying that Google and Facebook likely didn’t need 80 percent of the people working there. He argued Facebook could be built today with just a few hundred people.

This will be possible, he says, because future things will start to be outsourced that we couldn’t dream of being outsourced today. And whole armies of workers would wake up everyday, log onto whatever crazy hardware we’re using at the time, and get a daily assignment from a variety of companies — much like an Uber driver.

Clearly a world rich with Instagrams would be good for entrepreneurs — they keep much more control and more ownership — and bad for VCs, who rely on startups needing lots of money. Money is one of the last thing many VCs have to offer, in Ravikant’s view. He talked earlier in the evening about how VCs have already been edged out at the earlier stages, not only in terms of that initial equity but in terms of brand. Everyone says “YC’s Dropbox” for instance, not “Sequoia Capital’s Dropbox,” as they might have in a previous era. Another more staggering example of a similar trend towards small playing out in the VC world: Ravikant surmised that YC’s Paul Graham probably enjoys the same personal economics as Andreessen Horowitz’s Marc Andreessen. And he didn’t have to raise $3 billion and deal with limited partners.

Perhaps I’m a dinosaur, but it’s hard for me to wrap my head around a reality where companies will keep growing but will simply stop scaling — even if Ravikant is right and they don’t need to.

Outsourcing ebbs and flows even as it generally becomes a more pronounced trend over time. For instance, the early generation of blogs all outsourced revenue. More recent startups, like the Verge and us, have hired in-house teams from the beginning, eschewing deals that startup blogs would have died for in years past.

Companies don’t always scale out of necessity. Sometimes they scale out of control reasons, security, ego, legacy, or just because it’s the way things are done. After all, IPOs are still the end game — disdain for them aside. And Wall Street wants to feel it’s buying something. And if those things are hard assets, they better be a huge staff of talented people. For the business and startup world to become this rewired, entrepreneurs and the entrepreneurial ecosystem would have to be as well.

Ravikant wasn’t shy about how traumatic this could be for VCs and society at large. But the world could still prove Ravikant right. Most visionaries sound crazy at some point. If any VCs would like to write a smart counter-point, we’d love to have it.

As a side note, this is why I love our monthly events. This is a conversation you don’t get into in a 20 minute panel with three other guys. Go here to watch the whole thing.