Revisiting Paul Graham's Frighteningly Ambitious Ideas

Wednesday, April 11, 2012

Or: How to Hit Bee Hives With Pine Cones.

A few weeks ago, we wrote an article reacting to Paul Graham's essay on what he sees as the seven areas in which some highly ambitious startups would have the opportunity to get down to some serious earth-shaking disruption. We noted, however, that Graham's venture fund, Y Combinator, "mostly gravitates toward startups making relatively small, iterative progress on existing idea spaces," rather than engaging in the kind of dramatic experimentation that a fund like YC, which dispenses small amounts of cash over large numbers of start-ups, would be perfectly suited to engage in. And here's where we got into a bit of trouble.

The post found its way over to Hacker News, and not surprisingly did not sit too well with its dedicated readers, to say the least. Many of the counter-arguments to the comments made in this blog were quite well thought out, and sadly there were just too many 'harumph's and 'how dare you's for us to respond to all of them, but there were two dominant objections that we'd like to take issue with here. The first being that Y Combinator has invested in ambitiously disruptive startups, some falling within the categories outlined in Graham's recent essay, like Greplin, Taskforce and justin.tv. And the second being that, as Graham notes in his essay, revolutionary companies often grow out of deceptively simple projects.

Now, for the first issue, we'd have to concede that perhaps some of the language in the prior post might have hit readers' ears as a little harsh. Understandable. But we didn't mean to imply that there's absolutely nothing remotely ambitious coming out of Y Combinator. Dropbox, Heroku and AirBnB are all innovative, fantastically successful companies. But the point is, it seems that the relative dominance of startups coming out of YC, effecting incremental progress in existant markets, indicates a bit more playing-it-safe than we'd like to see from a fund with such an off-the-beaten-path business model, helmed by someone with such an oft expressed interest in disruption.

We know YC has to make money in the long term, just like any fund, but an outfit that plays the numbers game they do, with quite small investment figures, can afford to shake things up much more frequently than a traditional VC making million dollar investments. One of the Hacker News respondents guffawingly asked whether we'd like to see a class consisting entirely of search startups. But what's so crazy about periodically running a class in which many, if not most of the founders are trying to shoot the moon? When it was first founded, Y Combinator introduced a revolutionary funding model to the VC world, and certainly has produced some revolutionary companies over the years. But it's our opinion that it's got the opportunity to push for yet more disruption, and a call for more upheaval, for even crazier ambition, should never be sneered at. At the very least, it would be an extremely interesting experiment to see what happens to the startup landscape in the long run when an unprecedented number of wild revolutionaries are thrown into the mix.

Which brings us to the second argument raised: that revolutions need to start small and unassuming. Now it's true that, as Graham notes, anyone saying they're gonna start a revolution from the getgo is going to have to wade through a lot of "haters." And it's certainly true that many of the biggest tech companies started out just trying do something a litle bit different from what was dominating the market at the time. But, that doesn't mean the only path to a major market shake-up starts with something ignorably unambitious. When Pinterest first launched in beta in 2010, their layout and UI were worlds apart from the designs dominating the social space at the time, and yet 2 years later they've got a revolution of their own in motion. It is possible to veer well away from established wisdom, attack a space considered locked up by the existing players, and succeed phenomenally.

Furthermore, Graham has often noted in the past that founders rarely stick with their initial idea and often stumble into a dramatically different, hopefully more fruitful direction. And in his recent essay, he advises founders not to "try to construct the future like a building, because your current blueprint is almost certainly mistaken." Good advice, but that doesn't mean that it's preferable that everyone start with something safe and simple. Even if the wide majority of ideas that start out "frighteningly ambitious" from the get go are doomed to fail, if you increase the amount of experimentation with crazy going on, increase the number of people testing all the ways a revolution won't work, we'd be willing to bet you'd see a marked increase in the successful revolutions emerging from the startup landscape in the long term. And that's what we'd like to see from Y Combinator. Mix it up. Experiment some more (and more and more). Just get a bit crazier from time to time.