This is the best summary of where the tech cycle is at so far, IMHO.
Dave has been through a fair few cycles (remember, he was professionally developing software in 1979!) and regardless of some of his other views, I think his analysis is going to be some of the most accurate you'll see in the software / tech industry on this topic.
I think complexity is a necessary part of the equation, but you can push it into different places. The key is to keep the APIs and libraries simple. Unix got it right. Lots of small utilities that are designed to be chained together.
It seems to me that we will loose to software in this respect ultimately. We just can't fit everything we need in our short term memory sometimes.
What a coincidence. I just started working today on a social network crossed with a CMS that aims to be way simpler and flexible than Facebook and the rest. (But when I actually started writing the code, there were a few things I conveniently forgot that will definitely slow development down.)
One of the things that has occurred to me recently is that if someone was able to change the economics (even further) of processing a huge amount of data a la Google on a _much_ smaller budget, that could be one of the new sparks that kicks off the next trend.
Dave has been through a fair few cycles (remember, he was professionally developing software in 1979!) and regardless of some of his other views, I think his analysis is going to be some of the most accurate you'll see in the software / tech industry on this topic.