There are two fundamental ways to reproduce: the mammal model, where you have very few babies, and you take care of them for a very long time, and the fish model, where you spawn hundreds, even thousands of eggs and you spend zero time looking after them.
The traditional VC model has always been the mammal model. Back a few founders every year, pick very carefully, and give them resources over their entire life to grow up to reach adulthood. In order to qualify for funding, founders needed to have some extra-ordinary insight from a prior job: like Robert Noyce founding Fairchild Semiconductor after working with Robert Schockley, the inventor of the transistor, Sergei Brin for Google after having developed the core IP of page rank at Stanford, or even Mark Zuckerberg who bootstrapped Facebook to massive daily user growth before their first real VC round of 13 Million in 2005.
Since 2007, a different philosophy has emerged: be more like fish. With minimal due diligence, and a shotgun approach, Y Combinator managed to grab stakes in AirBnB, DropBox, Stripe and Reddit. The ROI was off the charts. The traditional VCs took notice.
Maybe picking winners was old school. Maybe the best idea was to back hundreds of companies, and be perfectly happy with single digit survival rates. The Unicorns would take care of the other experiments.
Maybe. Or Maybe not. One of the defining reasons Y Combinator was successful in the years post 2008 is a real platform shift was taking place with the Smartphone. When such a change happens this quickly, new success stories, like Uber, come out of nowhere as opposed to evolve out of prior knowledge (like Intel out of Fairchild SemiConductor). I would argue that in that case, shotgun works better. Right now, the platform is much more stable. With 2,000,000 apps in the app store, a random portfolio of 100 new apps stands a much lower chance of working.
And there is a second factor. In 2008, the Y Combinator companies were an oddity — and few in number. They could find access to series A or B capital much easier than today, when they are competing, literally, with thousands of similar companies. The model is a victim of its own success.
If I had to make a prediction, I would think we are going to move a bit closer to the mammal model as the fish model underperforms. More Warren Buffet, with a concentrated portfolio, and less the Russell 4,000 iShares of every app under the sun. More seasoned, second time CEOs, and less “minimum viable products” and “failing fast”.