A recent Facebook post by David Sacks is worth reading.  He said:

I think silicon valley as we know it may be coming to an end. In order to create a successful new company, you have to find an idea that (1) has escaped the attention of the major Internet companies, which are better run than ever before; (2) is capable of being launched and proven out for ~$5M, the typical seed plus series A investment; and (3) is protectable from the onslaught of those big companies once they figure out what you’re onto. How many ideas like that are left?

As you might expect, his note sparked a fierce debate in the comments (and a TechCrunch article).

I think he’s more right than not:  the unbounded “wild west” days of the Internet are winding down.   I’ve written before about the GAAF ecosystem, and how much of the Internet is now controlled by Google, Apple, Amazon and Facebook.  Those gorilla incumbents have taken a lot of friction out of the system, but they also tax, manipulate, control, and limit success at the upper-bound.  You’re not going to build the next Facebook on Facebook.

(The incumbents will get replaced, but it will be a LONG timeframe — they’re too deeply entrenched.)

However, the Facebook discussion suggests a nice framework for evaluating opportunities.  Ideas within the incumbent ecosystem (especially pure Internet software ideas) will be limited.  But stepping outside that ecosystem makes things much, much more interesting.  Uber (mentioned in one comment) is a great example:  they solve a messy, real-world problem — they’re a lot more than just a app.

Markets are smart:  Silicon Valley will figure it out, but it might take a while.