> With this proposal, there really are no small scales. That's not my understanding from the article. For example, "There are lots of experiments that will t

> With this proposal, there really are no small scales. That's not my understanding from the article. For example, "There are lots of experiments that will tell us something useful about the risk and efficacy of [geoengineering], and these experiments are so small that their physical risks are negligible." He's not suggesting launching right into full-scale geo-engineering, but rather to start with small-scale tests and related research.