> When we ban things after they prove to be a distraction, that usually means we proved them a distraction via a body count. Waiting for a hundred or a thousand people to die so that you have statistical certainty is the shittiest kind of science.

I don't know. You make this sound really bad, but isn't this how a lot of science works? How do we know if a medical treatment doesn't have disasterous side effects? Sure, we start by assuming it's safe, and maybe we test on animals to gain some confidence, but at the end of the day you can't really be sure until people have tried the treatment.

Obviously, there are other safer ways to study whether things distract drivers. Perhaps we could use driving simulators to test reaction time. I probably should have worded my comment differently, and asked for proof that Google Glass makes drivers quantifiably more distracted or less reactive rather than asking directly for proof of a higher accident rate.