Open thread, Feb. 13 - Feb. 19, 2017

Psychopaths prefer explicit communication because it allows them to find loopholes, abuse people, and then say: "heh, your fault, you didn't explicitly communicate your expectation that I wouldn't do X" and have their behavior socially accepted

I don't think that explicit communications automatically leads to more abuse. If someone is forced to make a demand explicitly it's easier to call the person out for making an unreasonable demand then if everything happens in guess mode.

When it comes to sexual activity ask culture often means that there's a requirement for explicit consent. "Yes means Yes" is a demand for ask culture instead of guessing whether the other person wants to have sex. It can lead to less abuse because a person can reasonable complain that they haven't been asked for consent.

Guess culture rules can be abused by assholes as well.

The levels of misunderstanding in these types of headlines is what is scary. The paper is actually about a single simple model trained for a specific purpose, unrelated to the hundreds of other models various deepmind researchers have trained. But somehow that all too often just get's reduced to "Deepmind's AI", as if it's a monolothic thing. And here it's even worse, where somehow the fictional monolothic AI and Deepmind the company are now confused into one.

If you read it, I'd be interested to know what specific techniques they endorse, and how those differ from the sorts of things LW writes.

The general 4 categories of goals/subgoals Wikipedia lists seem right though. I've see people get stuck on 3 without having any idea what the physical problem was (2) and without more than a 1 hr meeting to set a "strategy" (1) to solve the problems that weren't understood.

1. In consideration of a vision or direction...

2. Grasp the current condition.

3. Define the next target condition.

4. Move toward that target condition iteratively, which uncovers obstacles that need to be worked on.

I enjoy Every Frame a Painting on cinematography.

It's really going to depend on your interests. I guess I'll just dump my favorite channels here.

I enjoy some math channels like Numberphile, computerphile, standupmaths, 3blue1brown, Vi Hart, Mathologer, singingbanana, and some of Vsauce.

For "general interesting random facts" there's Tom Scott, Wendover Productions, CGP Grey, Lindybeige, Shadiversity. and Today I Found Out.

Science/Tech/etc: engineerguy, Kurzgesagt, and C0nc0rdance.

Miscellaneous: kaptainkristian, CaptainDisillusion, and the more recent videos of suckerpinch.

Politics: I unsubscribed from most political content a long time ago. But Last Week Tonight and Vox are pretty good.

Humor: That's pretty subjective, but I think everyone should know about The Onion. Also Fitzthislewitz.

I have one YouTube channel on game theory, and another on microeconomics.

The "Web of Stories" channel has interviews with notable scientists (Freeman Dyson, Marvin Minsky, John Maynard Smith, etc.).

YouTube has tons of good stuff, it's a question of which addiction you want :-) I'm a longtime fan of Accursed Farms.

Computerphile has a lot of interesting material. But my favorite is interviews with famous scientists and Elon Musk.

Someone on Brain Debugging Discussion (the former LW Facebook group) runs a channel called Story Brain. He decomposes movies and such, and tries to figure out what makes them work and why we like certain things.

A tetrahedron of edge length e has volume e^3/sqrt(72). So the 600-cell has surface volume 600e^3/sqrt(72). Let r be the distance from the centre of the 600-cell to the centre of one of its boundary tetrahedra. If we increase r to r+epsilon, the content increases by a shell around the 600-cell of width epsilon. This shell therefore has content epsilon*600e^3/sqrt(72) (plus terms of order epsilon^2 and smaller). Therefore we can find the content of the whole 600-cell by integrating 600e^3/sqrt(72) from r=0 to whatever r is when e=1. It remains to calculate r in terms of e.

Suppose we have a vertex of a polyhedron where n angles of size theta meet symmetrically at a point. Then the angle between adjacent faces (the so called "dihedral angle") is given by 2arcsin(cos(pi/n)/cos(theta/2)). This isn't too hard to see if you draw a little diagram. Thus the angle between two faces in a tetrahedron is 2arcsin(cos(pi/3)/cos(pi/6)). Since five tetrahedra meet at each edge of the 600-cell we can calculate the angle between two tetrahedra to be 2arcsin(cos(pi/5)/cos(theta/2)) where theta is the angle we already calculated between two triangles in a tetrahedron. Call this new angle phi. In a tetrahedron of edge length e the distance from the centre to the centre of a face can be seen to be e/sqrt(24). By considering the triangle formed by the centre of a 600-cell, the centre of one of its boundary tetrahedra, and the centre of one of the tetrahedron's faces, we can therefore see that r = (e/sqrt(24))*tan(phi/2).

After much calculation, all of the trigonometry cancels and we have r=e(sqrt(5/8)+sqrt(3/2)). Then doing the integration we get the content to be (25/4)(sqrt(5)+2)e^4.

Vs lbh glcr "ibyhzr bs 600-pryy" vagb Jbysenz Nycun, gur nafjre V nffhzr lbh'er ybbxvat sbe pbzrf fgenvtug bhg.

What do you mean by "general formula for the bulk"? I mean, (1) are you using "bulk" as a synonym for hypervolume or does it mean something else, and (2) along what axes are you wanting answerers to generalize? Obviously allowing for different edge length is trivial. Are you looking for a formula for arbitrary convex polytopes, or what?

driving up grocery prices:
flour
salt

lost all credibility to me.

There's some in the Logic/Mathematics section here.