Out of left field

Scientific knowledge is on a faster-than-exponential growth curve. In the 18^th century it was possible for one person to remain current with the total advance of scientific knowledge; in the early 21^st specialists, like the Red Queen, need to run as fast as they can just to stay in the same place in their ever-narrowing specialities (relatively speaking). With all the difficulties in keeping up with the advances in one area, others are likely to be neglected - by the specialist scientists, those who track the fields and often by engineers who turn knowledge into applications. What this means in practice is that once-insurmountable problems may fall relatively simply to techniques or technologies developed for some completely different purpose.
Photovoltaic power is one of those difficult problems. It has been getting steadily cheaper over time but it is a long way from being competitive for most uses. The dependence on silicon increases the cost of materials and processing on one hand, and on the other it pushes the band gap far into the infrared and decreases the recoverable energy from each solar photon. Together this drives up the cost per watt, which improved technology is bringing down but slowly.
Disruptive technologies are coming. Nanosolar claims to have a cell based on nanoparticles of titanium dioxide (a very cheap material, used widely as an opaque agent in paint) (hat tip: Futurepundit), using an innovative self-assembly system. At 12% efficiency and $30 per square meter, such cells could cost as little as $.25/watt-peak, a far cry from the $4.00 that current silicon cells go for at retail.
$30 per square meter is cheaper than many varieties of roofing, but the cost of fabrication and encapsulation is still going to place a floor beneath the cost and limit the applications. Greater efficiency can reduce costs by getting more watts out of the same investment in encapsulation and packaging.
This advance has already been widely covered: lead selenide nanocrystals which may allow 60% efficiency. The difficulty is creating a process to turn out nanocrystals of suitable uniformity in ton quantities and assemble them into working devices. This brings me in turn to something which appears to have received no press despite having been published just the week before in Science News: RNA Molecules Made to Create Tiny New Inorganic Particles. In this case the researchers created highly uniform particles of metallic vanadium out of aqueous solution. Could the same technique be made to evolve RNA molecules which make PbSe nanocrystals of 10 nm size instead of vanadium at 2 um? Are there processes - whether surfactant-driven self-assembly, DNA linkage or other - which can precipitate a bath of nanoparticles into a layer of photovoltaic cell?
At even 50% efficiency and $100/m^2, the rules of the game become unrecognizable. Twenty cents a peak watt means power at a penny or so per KWH; 500 watts per square meter means that an electric car surfaced with such cells can recharge itself for tens of miles per day on nothing more than the light falling on it. Such advances coming out of left field will change our lives in ways we cannot anticipate, and the likelihood is all but certainly greater than our narrowing fields of knowledge would lead us to believe.  ¶ 7/15/2004 10:03:00 PM