Possibly untrue science news

jerwin · February 12, 2020, 6:55am

I interned with a guy who had hundreds of citations. https://en.wikipedia.org/wiki/Leland_Clark

jerwin · February 13, 2020, 12:15am

AndyHilmer · February 21, 2020, 2:22am

kxkvi · February 21, 2020, 3:58am

It’s tempting to believe that physicians are logical, meticulous thinkers

AndyHilmer · February 21, 2020, 4:40am

Aheh. My only experience with doctors while dulting was of uncomfortable young men out of their depth with no clue how to deal with their daily workload of body-mechanic Standard Work.

AndyHilmer · February 22, 2020, 1:46am

Putting words around strategically fraudulent science:

AndyHilmer · February 22, 2020, 2:28am

From the comments. Almost there, but after describing a regular engineering process the commenter falls flat on their face with righteous indignation.

well, I am “high-decoupler” by nature, I instinctively go for stripping everything to base issue, solve that and look how simplified issue changes when we include more and more variables.
And I am often annoyed with people who cannot decouple things even for the moment, even just to see how the issue changes with each simplification.

Because such decoupling is useful in the same way as statistics are useful – you have to know what the numbers are saying but even more you have to know what they are not saying.
So decoupling is just a way to untangle the messy issue to look for different aspects and how they interact before putting it back together.

But that’s an issue, you cannot do that in a tweet, so you shouldn’t.

“For shameful low-decouplers like myself, though, I am also able to hear the obvious implication that Y is an unimportant complication that they don’t want you to think about, and when Y is something that leads to misery and suffering for large numbers of people, I tend to want to say “But you can’t dream about X while ignoring the inevitable disaster of Y that it will bring about!””

But that’s the trick, in most cases when I “high-decouple” I am not assuming Y is unimportant, I am just considering how X would look like without Y, so I can compare it with how X looks like without Y and how it would look like with Y or with Z instead.
It is an issue of different approach to problem solving. I worked for a long time with a coworker that was great at creating overall solution that I couldn’t even imagine, but I was able to dissasemble his plans into smaller pieces, analyse them and reconstruct it.
Both approaches (analytic and synthetic) were much better used simultaneously (or in turns) than any of them alone.

I do not defend Dawkins here because twitter is pointless and I don’t have context to know what he even means.
But I see your rant as an attack for the way my brain is wired to look for the solutions. And that is unfair.

kxkvi · February 22, 2020, 2:33am

Decoupling is probably OK when you’re doing a fluids simulation (not compressible? Throw that out. Temperature the same everywhere? Throw that out. Etc.)

But human traits? Behavior? Social structures? A little harder maybe.

AndyHilmer · February 22, 2020, 5:40am

Try unworkable. Engineers don’t do psych unless they’ve ever been on the outs. By outs, I mean fucked by some power structure. The onion, their parents, corporate manglement, way before you get to the structures that enforce underclass status based on race or being disabled or just a weirdo who had a bad day.

When an engineer does psych without knowing real life, the rabbitholes are endless.

chenille · February 22, 2020, 8:16pm

Trying to simplify things is an important part of figuring them out. Thinking it’s the only part – really, even assuming it must be the most important part of complex things like societies – is basically what they call engineer’s disease.

See also the libertarian tradition “of course all policy should be decided by purely Socratic reasoning from premises that sound obvious to me”, which turns out to have an awful lot of overlap. Real understanding is less about knowing how to apply tools then where to apply them.

MarjaE · February 22, 2020, 10:53pm

I have a hammer. This problem is just an unusual nail.

AndyHilmer · February 23, 2020, 1:57am

It’s a rivet. Need a backing plate of some kind.

kxkvi · February 23, 2020, 3:45am

What, we don’t have a working psychohistory yet?

I guess I was being a leetle sarcastic with my

jerwin · February 23, 2020, 4:54am

Linear differential equations? How quaint.

kxkvi · February 23, 2020, 1:39pm

Actually, they’re nonlinear. But computers can solve anything!

jerwin · February 23, 2020, 5:27pm

My point exactly. Harder to isolate each component in turn.

smulder · February 24, 2020, 5:36am

somebody does, but they’re making billions of dollars with it instead of submitting it to peer review

kxkvi · February 24, 2020, 2:12pm

Good point.

Will a hologram of Mark Zuckerberg in a wheelchair appear every 50 years?

jerwin · February 26, 2020, 5:28pm

Dr. Stiassny’s team suspects that the mondeli bureau live in these deepwater canyons, and are occasionally plucked up by the currents that churn along the rock walls. They are rocketed hundreds of feet up to the surface, the nitrogen bubbling out of their blood. The same bends that kill them make their diets a mystery, bursting their swim bladders and pushing out the contents of their guts.

Nonentity · February 27, 2020, 5:11am

For the first time in its 174-year history, the Smithsonian has released 2.8 million high-resolution two- and three-dimensional images from across its collections onto an open access online platform for patrons to peruse and download free of charge. Featuring data and material from all 19 Smithsonian museums, nine research centers, libraries, archives and the National Zoo, the new digital depot encourages the public to not just view its contents, but use, reuse and transform them into just about anything they choose—be it a postcard, a beer koozie or a pair of bootie shorts.

And this gargantuan data dump is just the beginning. Throughout the rest of 2020, the Smithsonian will be rolling out another 200,000 or so images, with more to come as the Institution continues to digitize its collection of 155 million items and counting.