Prime spirals (early view) (Patreon)
Content
Hi Everyone,
As usual, I'd like to give you an early view of the current project before it goes live this coming Sunday. I really enjoyed getting your input last time on the Windmill puzzle, it made the ending message better than it otherwise might have been, so I wanted to do something similar here.
At the moment, the video is left without a real ending. What's linked above is simply the core puzzle and a solution. What I'd like to convey after that solution is some message on the value of mathematical play, in particular when you keep asking deeper questions even after finding an initial answer.
In this case, after the initial explanation, there are a few types of questions I could imagine someone asking:
- Once you've ruled out various residue classes based on simple divisibility considerations (e.g. 44k + 12), what can we say about primes in the remaining residue classes? They look somewhat randomly distributed, but what do we mean by that? Will all of the remaining ones necessarily include primes? Is there a statistical bias for any particular spiral over another?
The answer here is deep, but it is known. While I wouldn't talk about the proof of the key theorem here (Dirichlet's) I could go into a little bit about what it's saying and the kind of approach used to prove it, and why mathematician's care. - Where exactly do these approximations for pi come from? What does it mean for one like 355/113 to be "extremely good"?
This could lead to a little discussion on continued fractions, which gives a nice way to show why other famous irrationals like e, phi or sqrt(2) won't have an approximation anywhere nearly as good as 355/113 if you limit the denominator to be, say, less than 200. This leads one to wonder why the continued fraction for pi looks the way it does, which again, is highly non-obvious. - Does this have anything to do with Ulam Spirals?
I'd introduce what an Ulam Spiral is (some of you may have seen a Numberphile video on this topic) and address whether what's going on there is related to what's going on here (short answer, no). I'd talk about how while this shows which residue classes contain primes, Ulam Spirals show a much more non-obvious propensity some quadratic polynomials have for being prime.
All in all, the aim will be for the concluding message to talk about how often otherwise pointless and arbitrary mathematical play, say with some data visualization like this, can surprisingly lead you to deep core ideas from math.
Let me know which, if any, of the possible directions above appeal to you.
-------------
If you're curious, as mentioned in the previous post, there are two other projects in the pipeline. One is a full story of e, which I originally started as part of the differential equation series (where it will likely still end up), though it will likely end up being multiple full videos itself. That's the one I think of as most important, so I'll probably hop back on that train for the next few weeks. The other is the hyperdarts puzzle, teased in that post, which is the fall-back during writer's block.
-Grant
Files
Previews only