What to do with the money?"

Recently folks have begun openly questioning the existing Venture Capital model.

Two examples are Bill Stensrud's post on The Primacy of the New Tool and pehub's interview with Bill Stensrud.

The conclusion of many of these articles is that the Venture Capital model is broken and cannot continue to generate pre 2000 levels of returns. Pehub's article quotes Bill in the title with, "I Couldn't figure our what to do with the money."

As a fundraising entrepreneur my obvious tongue-in-cheek response is, give it to me ;)

In all seriousness.

I agree with Bill and many other people who say that VC practices need to change. Specifically, that investing in companies that provide incremental advances on technologies controlled by incumbents is foolish - always has been and always will be. The only way to take on an incumbent at their own game is: 1) for the incumbent to be asleep at the wheel, 2) to field a disruptive technology, 2) to field a disruptive business model (Innovator's Dilemma).

However I am having trouble swallowing the conclusion that after 50 years we are done with extracting all the value from the "transistor" (Bill). This seems premature and counter to the driving force behind IT innovation and disruption. In my opinion the plateau we are on is temporary and has come into play due to two factors:

1. The computing performance delivered by individual processing cores is no longer increasing exponentially the way it has for decades (Moore's Law/Corollary). In the past one could rely on generational improvements in processor performance to solve application performance problems. This means that programmers in the past could continually increase the complexity of their applications to better fit user needs without developing new algorithms or having to resort to difficult to implement parallel algorithms. For example the human genome project was possible only once sufficient computing power became available. Speech recognition has also become feasible due to advances in computational resources. The end result is that we now have to figure out a way to make parallel programming accessible to the masses.
   
2. Socially we are satisfied with current technology and are in an absorption stage.  Culturally we are trying to figure out what social network and other new technologies mean and how they'll change society so the demand for new technologies has dropped.

What we have is a situation in which we have temporarily plateaued until either we figure out how to change a business process through technological advancement or a driving societal need that needs to be filled manifests (good luck guessing what that is).

Technologically we have reached a "crisis" point (Structure of Scientific Revolutions) with respect to parallel programming.  Theory says that an answer, i.e. a good general purpose programming model, will emerge within a few years.  Alternatively we will figure out how to improve the performance of individual processors again with 3-dimensional chips, optical interconnects, or other technology and thus avoid the nightmare of parallel programming where decades of research has only been able to address a limited subset of the problem space.  Fortunately at LineRate Systems we have already devised a parallel programming model for programming network appliances with general purpose multicore hardware.

My best guess says that the parallel programming problem will be solved first and that there will be a doozy of a revolution in a few years that will make the web revolution look like child's play.  This will coincide with society's absorption of existing technology and resumes its appetite for the next big thing.

Meaning folks are already close to an answer... Invest and Carpe RoI!