I agree with the view that we need to use silicon more efficiently.
However, in 20 years time, scarce industrial metals such as gallium and indium will no longer be available for manufacturing semiconductors.
We will need replacement technologies that are based on more commonly available materials, and whose supply will not be exhausted before the end of the 21st century.
Sustainability pressures - not the ideal march of "progress" or the marketplace - are what will drive the development of much new technology of the future, including new semiconductors.
We have enough "technology for the sake of technology" as it is. If anything, we need to begin simplifying our high tech, industrial paradigm.
Ironic that as a carbon-based life-form we ignored the potential of carbon-based electronics for so long!
There is a pressure to retire Si but I don't think it is dead, we just need to cut down software bloat and use the high-grade silicon fabs we have already, maybe even go back a generation to toughen it up a little.
Concentrate on using silicon smarter, better architectures, better system-level thinking, lass wasteful attitudes.
Let the researchers pursue other technologies, but in the meantime let's reap the benefit of the last decades' work and get something done with current Si technology instead of throwing it away before we have even worked out half the things it can do.
Progress is inevitable, but the pace of progress is our decision.
Researches related to Graphene are moving in really fast pace. How soon all those foundries need to rebuild themselves to adopt to the latest technology? It may be too soon for TSMC. Yet, I am sure Applied Material is paying attention to how the researches go.