I wonder whether this definition of relevance is always valid. Which definition?

That interest, I mean relevance, is controlled by the gap between causal complexity and description complexity. My sister bought a bike.

What?! Ha, you see! It’s not relevant.

And? It’s not causally complex, so it is not relevant.

You mean, it would be more relevant if she were in prison for the next ten years. Or if she were blind? Yes. That would increase causal complexity considerably. What could bring her to buy a bike in such a case?

But what about an eclipse? ??

I mean, eclipses are events, and yet they are perfectly deterministic. By saying "deterministic", you mean that their causal complexity is zero?

Yes. I just read something about this example. I mean, about why the occurrence of an eclipse might be unexpected, and therefore exciting.

I can’t see how an eclipse can be unexpected! It all depends on the viewpoint you take.

What do you mean? Well, it doesn’t happen everyday!

Oh, you mean that there’s some causal complexity to get an eclipse precisely today? Yes.

But that’s just probability! More or less. You’ve still to consider the description complexity of the concept "eclipse".

I thought it rather had to do with the causal complexity of having the Earth, the moon and the sun perfectly aligned. Yes, that’s another viewpoint. It’s all explained on the Website.

I’ll definitely have a look. Yes, you should. It’s cool!