This is pretty remarkable. I have no idea how much power is need to make 100KW of diode lasers actually run, or how decent a beam you can get from such an array, or if it would actually work around, say, dust. But that it should be doable at all is something of a tour de force. (OK, and I’m a sucker for lasers.)
Congratulations to Fran Allen for the Turing Award. Better late than never (and she’d already won all the other awards named after anyone who was anyone…)
Wave power setup actually in the process of being installed.
Behavioral economics: taking account of the irrationality of homo economicus to do good things for people.
Small revolutions: the drug-policy people finally imagine that alcohol might be thought of as a drug. This is nice in some ways but pretty sad in others, because in non-official circles the notion that the legal status of a drug by itself has little or no moral/practical valence has been around pretty much forever. (There’s also another interesting omission, namely thinking of tobacco as a drug.)
Yet another erasable paper. Not as stupid as the last time someone tried this in the mid-90s, but I’m starting to wonder if it isn’t a symptom of really bad business-process design. The same thing could probably be better done with some kind of smart paperclip (and why aren’t those everywhere?)
Penrose Mosques. Wow. (I wonder if the theoretical underpinnings were actually as stated, but who knows.)
Customer service done right.
The supposed efficacy of “natural family planning” done perfectly doesn’t impress me as much as it might, for a couple of reasons. First, the “perfect” practice apparently involves abstaining from intercourse for 10-14 days of every 28 (if you use a barrier method your failure rate goes up by 50%, which suggests to me that some of the perfectly practitioners might have been abstaining for rather more than the recommended period). Second, with 900 subjects, you’re talking about only a handful of pregnancies in the “perfect use” group — remember that’s supposedly 0.4% per women per year, so the confidence intervals are going to be several times the size of the reported result. So it’s interesting, but not what I would call “news you can use”.
Speaking of controlled experiments, this is pretty cool: if you put rats genetically predisposed to certain kinds of blood-vessel problems in cages where they exercise more, they don’t have nearly as much of those problems. Says a bunch about how closely you have to replicate some experiments to get good results, and also about how much “genetic predisposition” may actually mean for various disease.