SEMs in the wild
Storytelling as lossy data compression
This video I stumbled upon shows how commonplace it is to interpret equations as structural equation models. They just make more sense that way, because they tell a story.
Like all stories, causal stories may be misleading, but they are memorable. The particular story here is that V, the difference in electrical potential between the ends of a conductor, causes I, the current. It is not by chance that the person who posted this video is an engineer: in artificial systems it is usually easier to set one variable and observe what happens to the other, than vice-versa. Even though equations are formally symmetric, their usage is not. And use defines meaning…
This is exactly what I was getting at with this post a while ago:
Causality from outside
I have been invited to give a talk at the 17th Marcel Grossman meeting, to be held this week in Pescara, Italy. Even as I am writing this post, I still did not finalise my slides, and my talk is this Thursday. The topic I will cover is –of course- the application of causal discovery to astronomy. I just gave a talk about it at the annual meeting of
In the comments to the video, glandeokrayo9956 noted that
[…] ALL real voltage sources exhibit an equivalent series resistance. Because of this, by using the well known theorems of Thevenin and Norton, any real source can be represented either as a voltage source (with a series resistance) or a current source (with a parallel resistance); they are equivalent.
which undermines the main point of the video from the point of view of its physics content, but leaves its psychological value unscathed. In fact most of the other comments are praise by people who finally managed to grasp the meaning of V=IR presumably by leveraging neural primitives of causality. Data compression feels good.