Links for June 2014

Adorable ten year old Indian child sends a $20 bill to India’s central bank to “help the economy”. Here’s the letter they sent back. This is something I imagine Ozy having done as a ten year old.

Did you know that Kaiser Wilhelm II lived until 1941? Or that he had strong opinions about the Second World War?

Jack Chick’s anti- Dungeons and Dragons tract, “Dark Dungeons”, is being turned into a movie.

Still no flying car, but a pretty neat hoverbike is going on sale in 2017 for only $85,000.

The Netflix Summary Bug accidentally combines summaries for two movies, creating some very interesting films.

Your job is already being automated. Including some doctor jobs. I wonder if I should try to concentrate more in psychotherapy because that will be less automatizable than psychopharmacology.

If we psychiatrists ever have to fight a battle against automatization, I hope we can come up with something at least as hilariously bad as these ads by nurses complaining about medical algorithms.

Oxfam: New research that performance monitoring of charities can make them do worse. I bet this is the sort of thing where if we were able to tease apart “performance monitoring” a little better we’d find some types are good and some types are bad. But it’s useful to keep in mind.

Giant tower in the desert could generate as much power as Hoover Dam.

Apparently whatever I have is contagious, because one of my girlfriends writes things like this, this, and most recently this.

If obesity is rising at the same rate in every demographic group, what does that tell us about the causes? Article points out a lot of interesting things – people are eating better and exercising more over the past couple of years, but obesity rates are still going up. Food deserts have no relationship to obesity whatsoever. People (even the poor) spend a lower percent of their income on food than in the past. Basically, all the social explanations that never work for anything don’t work here either. My money continues to be on either increased sugar consumption or some kind of weird endocrine effect from the chemicals in our environment.

Study shows that less scientifically literate people are more likely to believe in a threat from climate change. This doesn’t surprise me. If you know nothing about science, your best bet is to accept whatever real scientists say. If you think you know something about science, you may be more tempted to try and analyze the problem yourself, tragically unaware that you still don’t know nearly as much as real scientists. On the other hand, this should put the final nail in the coffin of the “anyone who disagrees with me is scientifically illiterate and hates knowledge” theory of politics.

Speaking of climate change, there’s a lot of debate around the study that said 97% of scientists support the consensus. Although I usually approach climate skepticism blogs with an entire ocean’s worth of salt, I was pretty impressed by their investigative reporting on how many of the scientists cited as part of the 97% consensus in the study vehemently deny being part of the consensus and say their papers were misclassified as supporting climate change when they’re actually against it. An author of the study replies, saying that anyone who complains about this is a science denialist and is “cherry-picking”, which doesn’t make much sense to me – kind of like saying “It’s unfair of you to cherry-pick only the data that we falsified”. Possibly a better explanation is that she’s saying her classification process will naturally misclassify a certain small percent of papers, and if you find those and get the scientists involved to make public statements about it, it will cast doubt on the accuracy of a generally very good method? Anyway, a much stronger argument is that the 97% paper used two different methodologies, and the other was asking the scientists involved to classify their own papers, and that one also independently found about 97% consensus. Or that of three different attempts to quantify the consensus by three different teams, they all got about the same numbers – 97% vs. 97.5% vs. 98%. That makes me more confident that, even if some mistakes were made in this study, the general point stands. But as always your best bet is to read the original paper. Also, I continue to feel like political partisans with very very very strong opinions on a subject should not be the ones studying it.

Related: Climate Change Now More Divisive Than Abortion. I will always link articles with terrible stock photos.

The latest thing I learned by living with Ozy is that Texas governor Rick Perry is inexplicably refusing to take common-sense steps to end prison rape, even going so far as to return money the federal government was offering him for free for the purpose. I don’t like calling any issue one-sided, but Ozy’s suspicion that he just wakes up in the morning thinking “What is the most evil thing I can do today?” seems pretty plausible.

New pathway involved in autism discovered. Then again, new pathways involved in autism are discovered all the time. I’m linking this article because it starts with the most forced metaphor I have ever seen in a piece of published writing.

Liberal bloggers are more likely to acknowledge alternative points of view. On the one hand, it’s a plausible claim, and the methodology – looking through liberal and conservative blog posts and having people check how many times opposing arguments are mentioned – seems like it could work. On the other, it’s an article on a liberal website based on a study no doubt by liberal researchers, it’s all based on a grand total of twenty-four blog posts from a small number of blogs, and it’s hard to see how you could blind the raters as to whether a site is liberal or conservative. I would like to read the paper, but the Internet stubbornly refuses to give it to me.

Finally we’re starting to get good evidence that e-cigarettes boost quit rates. Called it! Now we just need people to use this to turn back the tide of trying to ban e-cigarettes.

There have been some good articles on licensing lately. As they put it, almost 1/3 of U.S. workers need a permission slip from the government to get their jobs. This is not just jobs where poor performance could be a disaster, like surgeons and nuclear plant operators, but everything from florists to coffin-sellers. When states with and without licensing requirements are compared, job growth is 20% higher in unlicensed states. And licensing probably increases inequality, because it prevents the poor (who can’t afford the years of education and licensing fees required) from getting good jobs.

Extremely relevant: Uber gloats that although regular cab drivers are paid only about $30,000/year, its drivers in NYC and SF (they don’t give national averages, making all comparisons apples-to-orange) make $70,000 to $90,000. The big difference seems to be that regular cab drivers are employees who give most of their earnings to the large businesses that have ponied up the $1 million necessary to buy a taxi license, whereas Uber drivers mostly just work for themselves and give a small percent back to Uber. ARE YOU STARTING TO SEE WHY PEOPLE CAN CARE ABOUT THE WORKING POOR AND STILL IDENTIFY AS LIBERTARIANS?

…except that Uber is planning to replace all its drivers with self-driving cars in the near future anyway, so I guess I can only claim a partial victory here.

Trust Your Doctor, Not Wikipedia says the BBC, based on a study showing that Wikipedia contains lots of things that contradict medical research. “90% of the entries made statements that contradicted latest medical research.” Yet I notice it doesn’t even come close to proving its headline. I bet that if you asked doctors to talk about a medical concept in as much detail as a Wikipedia article dose, WAY more than 90% of them would say something that contradicts the latest medical research. I’m not even sure that Wikipedia contradicts “the latest medical research” less than “the latest medical research as identified and interpreted by a different person” or “the medical research of six months ago” or “the medical research of six months from now.”

And while we’re on the subject – you know that group that found that non-celiac gluten sensitivity existed, and then did a different study and found that really it didn’t exist? Now it exists again. And it supposedly causes depression.

It looks like mainstream psychiatry is gingerly accepting the mutational load perspective on schizophrenia and other illnesses.

Study finds that cerebral blood flow differences between men and women start at puberty, which is interesting because a lot of male-female gender differences start at puberty as well, and a lot of the work trying to debunk such differences has been taking schoolboys and schoolgirls, finding they do about the same at some young age, and assuming that any differences that later accrue must be cultural.

Kate at G&H makes (see #3) an exactly symmetric argument on the other side. She points out that you can’t conclude that, just because some male-female difference is present very very early (let’s say at age one month) that it’s biological, because differences from differential treatment could have accumulated in their one month of life or even in the womb. I agree. On the other hand, the article she links to says this means we should “stop looking for hardwired differences”. To me, this sounds like a particularly egregious example of the fundamental rule of confirmation bias: if I find evidence I like, I ask “am I allowed to accept it?” and if I find evidence I don’t like, I ask “Am I forced to accept it?”. Just because some male-female difference is present at age one month or one day or in the womb doesn’t mean anyone can “force” you to accept that it’s biological. But if you were going around saying it was definitely because of who played with Barbies more at age ten, you at least have to accept that it’s some evidence you were wrong. No one can *absolutely force you* to accept that some brain differences are innate, but finding them at ever-earlier ages with ever-less opportunity for cultural contamination should sure point you in that direction if you’re a Bayesian. And saying “Stop looking for the differences because you can never 100% prove to me that I am forced to admit they exist” is placing the entire burden of proof on one side in a very un-Bayesian way.

Speaking of gender, Alas A Blog presents lots of statistics on why feminism has been good for family values and improved the lives of men and women alike. Interested in seeing what my more conservative readers have to say about them.

Speaking of men being hurt by insufficient feminism, Female-Named Hurricanes Kill More People Than Male-Named Hurricanes Because People Don’t Respect Them, Study Finds. After eliminating outliers, female-named hurricanes kill almost twice as many people as male-named hurricanes because people don’t fear the female-named hurricanes enough to take precautions. Hilarious if true. But another article more plausibly suggests this might be an artifact of change in hurricane name patterns. (caveat bolded after EVERYONE IN THE WORLD started posting this article to Facebook)

And I will stop talking about gender after this, but did you know that there is a whole subreddit for polite, productive debates between feminists and men’s rights advocates? More cats and dogs snuggling together!

The decline of Detroit in time-lapse.

New research shows, contrary to every urban planner everywhere but totally in accord with my personal biases, that suburbs have healthier communities and better relationships with neighbors than cities.

The CIA Says It Will Stop Doing Fake Vaccination Campaigns. As someone – I think Leah – said, very much closing the stable door after the horse has escaped. But still welcome.