Last Friday, Bill Gates was at the TED Conference in Long Beach and told the audience that climate change was the world's most vexing problem, but that it would take "energy miracles" for the world to zero out its carbon emissions by mid-century. What sorts of miracles? He suggested that we'd need radical new technologies that barely exist right now, like a "traveling wave reactor" that would turn spent uranium into electricity. (Apparently, a company called TerraPower is investigating this possibility.)
I looked into this topic for a TNR piece last year, and the question of "energy miracles" is a complicated one. On one level, Gates is right. There are very good reasons to fret that the low-carbon technologies we currently have at our disposal—energy efficiency, solar, wind, nuclear, coal carbon sequestration—will one day bump up against some fundamental limits, at which point we'll need big scientific breakthroughs if we hope to keep cutting emissions enough to avert a climate crisis. So there's a strong case for pouring more money into research for those next-generation technologies—particularly since federal spending on energy R&D pales beside spending for, say, health R&D. My sometime co-blogger Mark Muro of Brookings has some terrific ideas on this score.
That said, in his talk, Gates suggested that we should focus on spending the next 20 years inventing new tech and then spend the next 20 years implementing it. That seems somewhat misguided. As I mentioned in the piece, radical breakthroughs in the energy sector don't come along often—and it can often take a very long time for a truly novel scientific discovery to translate into widespread technology (I used the example of the high-temperature superconductor). Expecting that a game-changing technology can be invented, perfected, and prepared for mass deployment in just 20 years is a real gamble. This isn't a task akin to the Apollo program or Manhattan Project, because the end project has to be cheap and widely accessible.
What's more, some of the best innovation comes from broadly implementing already available technology—your humdrum wind turbines or efficiency measures. (The policy mechanisms for doing so would likely involve a carbon price and some government support.) As those power sources get deployed, prices start tumbling and incremental tweaks get made. The technology improves over time. Indeed, as Joe Romm notes, that's how PCs became so ubiquitous, as Gates should well know. Given the need to start cutting emissions quickly, that sort of innovation will be far more important in the short-term—and far more likely to make a surprisingly large impact. So, yes, a miracle or two would be terrific, but that can't be the only plan.