(Some) IT Doesn't Matter (anymore) -- But some IT matters much more.

This is an excellent piece by Hal Varian in the New York Times, which summarizes Nicholas Carr's "IT Doesn't Matter" argument and then provides two compelling counter-arguments. First, he makes the distinction between the commoditization of IT (which is indeed quite far along) and the commoditization of IT best practices -- how to best use commodity technology (which is lagging quite a bit). Many others have also made this counter-argument.

His second counter-argument is much more powerful: commoditization of one "layer" of IT enables an explosion of innovation in the layer above it. He gives examples of how commoditization of machine parts enabled the innovation of sewing machines and other precision machines. When sewing machine technology commoditized, it enabled the innovation of clothing manufacturing, and so on.

One can see this same layered commoditization of IT: chips to systems to software. So now the question becomes, "What new IT innovations are enabled by the commoditization of some IT components?"

In my opinion, one explosive area of innovation enabled by the commoditization of systems and software is Service-Oriented Architecture, and specifically Web Services. These new, higher-level, Information Technologies, overlay the commoditizing lower layers and innovate regarding better integration across IT applications.