About

Shannon Capacity is a personal technical blog about things that I find interesting. The subjects range from astronomy and physics to mathematics, engineering, computer science, and philosophy.

I don't have many opportunities to discuss my interests with friends, so in a way this blog is an outlet for me to share the topics I'm exploring with anyone who's interested. If you've ever had late-night deep philosophical discussions with someone who shares your interests, I'd encourage you to think of this blog as a stand-in for that.

The name is a reference to information theory and the work of Claude Shannon. In that context, "Shannon capacity" is the term for the maximum rate at which information can be transmitted reliably across a channel (e.g. a Wi-Fi connection).

I liked the idea as a name for a website whose goal is to communicate complex topics with accuracy and brevity.

Quite a variety. Astronomy, physics, mathematics, engineering, computer science, philosophy, and occasionally whatever subject happens to lead down an interesting rabbit hole.

Some articles may be highly technical. Others may be more conceptual or historical. The goal isn't to stay within one discipline, but to discuss topics that are interesting to me.

Neither. This is a personal blog.

I care about getting the technical details right, but the articles are written to explore and explain ideas rather than to serve as formal academic publications.

It depends on the subject. Some articles are intended as approachable introductions, while others assume a stronger background in mathematics, physics, or computing.

I try to make my articles approachable regardless of your background, but I don't want to sacrifice the technical substance to achieve that. You'll find a pattern in my writing of starting with a simple overview, so that a novice can get an idea of what's being discussed even if they don't understand all the details. I also try to introduce topics by name so they can be researched independently by anyone who wants to learn more about them.

That's not to say everyone is going to follow section part of every article. The articles move fast, can be dense, and some sections get into substantial technical detail. My goal in providing background is to prevent readers from becoming disoriented rather than helping a novice learn advanced material. You're welcome to investigate concepts that are new to you as they get introduced, but there isn't a quiz after the article. My intention is to drag readers through a gauntlet of technical depth to show them what's there, not necessarily to give a cohesive academic background to everyone who encounters the article.

I'm more interested in understanding things than in studying specific subjects. Many of my favorite textbooks are about specific problems and approach them from several adjacent disciplines. I like to do the same.

There isn't a rigid publishing schedule. I'd rather spend the time necessary to make an article worthwhile than publish something simply to maintain a schedule.

The Articles page is the best place to browse. You can choose something based on its subject, or simply pick a question that looks interesting.

No

I wanted to make sure the blog had some narrative personality by featuring myself in some of the imagery. I don't want to use photographs since I prefer to be anonymous. I would have loved to hand illustrate all of the imagery, but I'm not very good at drawing faces and hand-drawing the imagery would end up taking much more time than writing the article. So, AI was the solution I settled on.

You can almost think of the AI-generated imagery like doodles in the margins. It's not intended to add anything to the technical content of the article, but some of the articles are conceptually dense and the imagery is intended to provide a cognitive break. I find that encountering something unexpectedly non-sequitur can help bring my attention back when I find myself reading on autopilot.

I'm aware there's some controversy surrounding AI "art", so I try not to hide the fact that I'm using it. The imagery I use is intentionally recognizable as AI generated and does not attempt to replicate any specific artist's style. I sort of enjoy the quirkiness it gives to some of the images, it compliments the "doodle" goal quite well.

No. All of the text on this website was written and edited by me.

I used a bold cobalt-blue and black color palette for the outfit to make it distinct. The symbology on the pleats of the skirt is intended to resemble the spokes on the pulsar map from Voyager 1's golden record, but the AI imagery typically messes it up. The charm on the hair ribbon is a reticle; I chose that for its relationship with measurement and precision. I figure it's much less cliché than something like a key.

The cut of the dress is based on a dress I actually own (well, it's a bit shorter. Being anonymous has some benefits).

This is a tough question because I tend not to use languages I don't really like, so I don't usually end up forming very specific negative opinions on them. I'll also have to exclude esoteric languages like Malbolge.

I would say that my least favorite is Java. I don't like that it's so commonly used in industry despite its historical corporate ownership, and I don't like that the division of a project across multiple files is so closely coupled with the object-oriented framework of the language.

The percentage is the absolute content of milk fat in the mixture by weight rather than the percent relative to whole milk's fat content.

Whole milk isn't 100% fat, that would be butter (specifically clarified butter if you want to be pedantic). Usually the milk straight from a cow has a fat content of around 3%, so the options are approximately even steps up to the fat content of whole milk.

Cream usually has a fat content around 30% or so, so if you'd like you can buy skim milk and cream and mix them to get whatever fat content you want (up to the cream's fat content)!