Ineffective Theory

Links for September 2021

The obvious. I suspect there’s a general principle: Alice will never convince Bob unless (Bob thinks) there’s a chance that Bob will somehow convince Alice. This makes all advice addressing “how to convince people” kind of pointless.

The long road to a faculty post.

Lance Fortnow on the rise of machine learning and its relation to expertise:

There’s a more personal issue–people spend their entire careers creating their expertise in some area, and that expertise is often a source of pride and a source of income. If someone comes along and tells you that expertise is no longer needed, or even worse irrelevant or that it gets in the way, you might feel protective, and under guise of our expertise tear down the learning algorithm.

A website for comparing different sources of weather forecasts, depending on where you live. For most places (including where I live), NWS seems to be solidly middle-of-the-pack.

Tom Ricks on U.S. military leadership. Is there any reason to believe that military leadership is fundamentally different from leadership of any other large part of the government?

Various studies, reported by Mankiw, on the effect of increased tax rates in Europe as compared with the U.S.

Molecular Dynamics In Your Browser!

Click here. A few points follow, from least to most technical.

You can click and drag the mouse to push particles around. The “brush” is deliberately wide and weak, to keep the simulation as close to the hydrodynamic regime as possible.

The parameter $n$ controls the dimensionless density of particles; that is, what fraction of the screen is shaded white. The result is that changing $R$ (the size of a particle) zooms in or out without changing the underlying physics.

Enabling the hydrodynamic overlay will highlight regions where particles have a net vertical velocity. You can use this to watch shear waves. Shear waves will decay more quickly if you make things sparse.

The average velocity of the particles is fixed to be zero. You can think of this as the camera panning to keep pace with the average velocity.

The underlying technologies (aside from the obvious) are WebGL and WebAssembly. The source code to the WASM binary blob is available here.

Evolutionary Pressure on Weddings

When you hold a wedding, it’s likely to be influenced by weddings you’ve attended in the past. Mostly this has obvious boring consequences; nice features of weddings you’ve been to (open bar) might be copied, horrid features avoided. Slightly more subtle, but still unsurprising, is something like anchoring bias. The weddings you have attended in the past define a “typical” wedding, and good or bad, any new ceremony is likely to be heavily based on that model.

In summary: good features propagate, bad features die out, and culture persists.

One feature of a wedding, though, is a little different: its size. More people attend larger weddings. If, on a given day, ten weddings are held, of which half have ten guests and half have two hundred, then twenty times as many people will have attended a large two-hundred-person wedding, even though only half the weddings that day were so large.

This has consequences for the idea of a “typical” wedding. Former guests, when planning their own ceremonies, will look to the set of weddings they’ve attended to determine what is typical. But that set is inherently biased towards larger weddings. As a result, when guests try to design their own weddings based on what is typical, there’s pressure for the size of weddings to grow over time. Of course, this growth is not without bound, because there are practical reasons why a wedding can be too large. But we’d expect more weddings to “err” on the side of being too large than too small.

After a quick rephrasing, this is the same as the friendship paradox. Most weddings you’ve attended have had more guests than yours had (or will).

Finally, note that there are other reasons why weddings might grow (a large wedding is a nice signal of wealth/status, for instance), but those are all dependent on external societal factors. The evolutionary pressure on size is independent of cultural, laws, or customs.