Believe the evidence, ignore the cranks

Isn’t this a fairly obvious strategy? To set transgender policy, look to the evidence. It’s what I would recommend. Unfortunately, that’s not how policy is established.

The current spate of anti-trans positions has little to do with evidence-based research, science or data.

Here’s one example. Anti-trans campaigners often argue that allowing trans women to use women’s toilets and changing rooms will increase sexual assaults. In fact, research has shown the opposite. One study tallied criminal incidents related to assault, sex crimes or voyeurism in public toilets, locker rooms and changing areas in parts of Massachusetts that had laws against trans discrimination, and compared them with those that hadn’t. It found no evidence that these laws put women at risk, and concluded that “fears of increased safety and privacy violations as a result of nondiscrimination laws are not empirically grounded” (A. Hasenbush et al. Sex. Res. Soc. Pol. 16, 70–83; 2019). Furthermore, there is evidence that transgender children who cannot use toilets and locker rooms that match their gender identity are at increased risk of assault (G. R. Murchison et al. Pediatrics 143, e20182902; 2019).

Nevertheless, a false ‘protection’ argument has been used to justify anti-trans ‘bathroom bills’ in Alabama, Minnesota, Oklahoma and Tennessee, and to buttress trans discrimination in the United Kingdom.

Politicians’ claims also have little to do with empirical evidence when it comes to gender-affirming care. Alabama’s law banning provision of such care to minors described the use of puberty blockers and hormone therapies as “experimental”. It did not mention that 22 US medical associations endorse these medications as well-established treatment for gender dysphoria in young people.

Case in point: Graham Linehan. He says we literally don’t know who to believe.

He’s against trans rights, and now he’s doubting the existence of COVID (citing Bret Weinstein!) and climate change, all because he’d rather listen to hucksters lying about the evidence than accept something contrary to his biases. He is getting sucked right down into the conspiracy theory pipeline, and you can trust that he’s going to get worse and worse. He is actively demonstrating crank magnetism.

That Nature article has a good suggestion on where to start. There’s plenty of research and evidence on these matters…the problem is that it all says Linehan is WRONG, so he won’t look at it.

Much evidence-based research is already available. More is still needed, but it is either a lie or a cop-out to say that there’s not enough research to make informed policy decisions. Instead of whipping up arguments to churn culture wars, elected officials and those around them should look to the evidence.

You could also start by unsubscribing from Weinstein’s podcast.

Musk’s demo was bad, and he should be ashamed, and his cheering stans should be embarrassed

Yesterday was Elon Musk’s “AI Day”, and I don’t know whether to say it was a letdown, or that it was exactly what I expected.

The Optimus robot was not ready for prime time. Its motors worked: the curtains were pulled back, it took a few steps, it waved mechanically, it turned around, done. It did not interact with anything or anyone. It did not have to deal with any novel situations. It did not exhibit any behaviors beyond what you might get out of a Disney animatronic.

I’ve seen more impressive animated mannequins in the Christmas window displays at ZCMI in Salt Lake 25 years ago. Come on, guy, show me something that requires intelligence and flexibility on the robot’s part.

I guess I can call myself an artificial intelligence researcher now. When I was 13, a friend and I built a Frankenstein’s monster in my bed, which could sit up (with the aid of strings we pulled) and roar horribly (tape recorder in its chest) and best of all, spurt red blood when I pushed a button on a pump. We invited my siblings to witness our creation in a brief little skit and then chased them out of the room as the monster rose up.

Oh, what I could have accomplished with a bunch of actuator motors, some shiny strings of blinking lights, and a Raspberry Pi.

Musk also showed some videos of Optimus picking up boxes and of using a watering can to water some office plants. Very short videos. Again, nothing that demonstrated any artificial intelligence capabilities at all. He brought out a second robot that waved at the audience, and it wouldn’t be a Musk demo without some embarrassing gaffe.

Yet he was promising that a home robot would be available in a few years, that they were going to mass produce millions of them, and that they’d cost around $20K each.

Musk himself is a terrible showman. He mumbled and “ummm”ed his way through a poorly practiced, short speech that had no pizzaz — he relies on overpromising rather than actually expressing some genuine enthusiasm. He’s no Steve Jobs, who could get up and announce “one more thing” that was an overpriced watch and get people rushing to spend a thousand dollars on it. I’ve been coaching undergrads in making scientific presentations for the past few weeks, and they’ve been so much better than Musk — spontaneous, smooth, well-organized, and demonstrating some genuine passion for their projects. They actually rehearse what they’re going to say and are careful about being accurate. I’d suggest that maybe Elon Musk would do better to hire some UMM students fresh out of graduation, except that I hope to Dog that none of them ever end up in a job that requires them to go through the motions of empty hype and lie with more enthusiasm than their over-valued billionaire boss.