I recently began reading The Singularity is Near, the contentious book by futurist and inventor Ray Kurzweil. Kurzweil argues that the rapid advancement of information technology and bioengineering will erase the distinction between humans and computers by the middle of this century. The book is fascinating but dense, and lots of its concepts go over my head. Among the more inscrutable ideas is what Kurzweil calls the best bet for creating artificial intelligence: nanotechnology. Kurzweil could have just as easily used the word "sorcery"—I'd have understood no more or less of his argument.

Until last Wednesday, that is, after I heard John Hart's lecture on nanostructures at McGraw-Hill's 2009 Innovation Conference. Hart, an assistant professor of mechanical engineering at the University of Michigan, Ann Arbor, named his talk "Making Massive Crowds of Nanostructures: From Atomic-Scale Principles to Programmable Building Materials".

A nanometer, Hart explained, is 10 ^ -9 meters across, which is about three orders of magnitude smaller than a single cell. Carbon Nanotubes (CNTs) are structures of this scale. They're created in labs like the one Hart heads by spreading a film of carbon onto a metal plate and heating it until tiny spore-like structures form. These buds grow upwards, forming a dense "vertical forest" of hollow CNTs.

Carbon is extremely versatile—graphite and diamonds are both carbon allotropes—and so being able to control and manipulate the element at an atomic level means that CNTs have almost limitless potential. Depending on the conditions present in the reaction furnace at their creation, carbon nanostructures can be grown in any number of three dimensional shapes (Hart showed images of nanostructures resembling everything from honeycomb to a vodka bottle to President Obama). A group of many of these nanotubes can be assembled into materials with "exceptional properties" like strength, density, thermal and electrical insulation or conductivity, etc.

Hart briefly discussed potential architectural uses for the technology. CNTs could yield lightweight composites and new geometries in buildings; a building's skin could be made of a CNT polymer that would become more or less porous depending on changing environmental conditions. (While I thought Hart's talk a compelling introduction to nanotechnology, I could have done with a few more examples of CNTs' potential uses to architects).

The lecture also addressed some of the obstacles that will have to be overcome before we see the widespread use of CNTs. For starters, CNTs can't be made in very long sections. They begin to buckle and warp before they can grow enough to use in something like a super-strong cable, for instance. Furthermore, running the furnace that generates them requires a tremendous amount of money and energy, and their production creates noxious pollution (a single synthesis, said Hart, creates as much exhaust as would smoking a thousand cigarettes).

So for now, it seems that architects—and Ray Kurzweil's dizzying prognostications—will have to wait until nanotechnology is made viable.