[jt2190]

  > The... ecosystem [of random startup ideas] would be
  > arguably more diverse, more innovative and overall better
  > for consumers. Instead of having 8 million photo sharing
  > apps and 11 million social network aggregators, you would
  > instead have an infinite number of truly bold products.

Somehow I think that there must be a better approach to creating more innovation than generating ideas at random, but I get his point: There are a lot of "me too" startup ideas. I wonder though: Is that really a bad thing?

[kybernetyk]

Innovation is iteration.

Creation of truly new things happens at most just a few times in a human life span.

The rest revolves around taking existing stuff and making it better. And there's nothing bad about it. It's how our technological evolution has worked since the "invention" of the wheel.

[npsimons]

I'm reminded of the story from "The Cyberiad" of the pirate who wanted to know everything (the name Pugg seems to ring a bell). The heroes of the story gave him a device which would continuously spit out facts. At first, the pirate was pleased, and our heroes escaped whilst he was busy absorbing facts. But these facts turned out to be random, practically useless information, and eventually the pirate was buried in them, unable to look away . . .

[davedx]

I think to some degree, it can show you what is in demand. The problem is, is it something in demand by developers or by bill paying customers?

And there is definitely a better approach, it's called "market research". :-)

[jt2190]

We may be talking about the inability of our research techniques to find all possible markets then. If companies targeted markets at random, then some of them would find the markets missed by our imperfect research. This would require that companies stop caring about survival and profit though, so perhaps the current funding models don't do a good job of incentivizing founders to think broadly.