January update, (manim upgrades, part-time hire, etc.) (Patreon)
Content
Unfortunately, no new videos for January. I've paused the Patreon billing for this month since in general content production has been slower recently, and you've all been very patient.
--
A lot of my time this month went towards manim, the open-source python library I wrote to do the animations for the channel. More and more, I've seen other people create math/CS explainers using manim; to name just a few there are some channels like Reducible and VCubingX animated entirely using it, and there are a fair number of one-off uses like this video by sentdex.
On the one hand, seeing all these is inspiring. It's touching to know that something I made is finding use from others putting out visual explanations of tricky topics. At the same time, I often watch these with a pang of guilt knowing of the many ways the tool could be improved which would have made the lives of those using it easier.
I started last year to meaningfully overhaul how things work under the hood, rewriting it so that it's built on top of OpenGL. This does a couple of things, most notably making rendering faster, and making it possible for a much more interactive development process. There's also more flexibility that comes from writing everything with shaders under the hood. While the fruits of that flexibility have not really shone themselves in recent 3blue1brown videos, there are a couple coming this year that would have been dramatically harder to make in the previous mode.
Work here is never done, of course, but a lot of my time this month went to clearing up many known bugs in this new mode and reorganizing things so that it's (hopefully) nicer to work with. For those curious to play around with it, there's a bit of in-progress documentation here.
One question a patron asked a while back when I mentioned work on manim was whether this is exactly the kind of thing that would make sense to hire out for so that more of my own time goes toward active video creation. It's a good point! I actually have recently hired some help, though not for manim (more on that below), and on the manim-front, I'd like to become a lot more attentive to the pull requests and contributions from the community moving forward.
I'll admit that part of my desire to be hands-on here is that I simply have a lot of fun developing manim. It feels less like a chore and more like an enjoyable side-hobby, analogous to how some people enjoy fixing up an old car, which is part of why I'm pausing Patreon billing for this month. Also, and in slight contradiction to the previous point, work on manim is often heavily intertwined with work on videos. Knowing exactly how every part of the library works, and why, makes it a lot faster to develop videos, especially if it involves extending the library on the fly with some new feature useful for a particular topic.
I'm curious to get the pulse from all of you on this. When you choose to support 3blue1brown, is any part of that motivated by supporting the underlying tooling which is open for others to use, or would you see that as unrelated? Would you like to see some video tutorials on manim, even if this meant less time toward new math videos?
--
In other news, I recently starting contracting a fellow YouTube creator James Schloss in a part-time capacity to help with various 3b1b projects. He and I worked together a bit on the MIT course last year, and have known each other a bit longer through the general YouTube education community. There are a few non-video projects I've been meaning to have done for a while now, like having written versions of all the videos as nicely-illustrated blog posts, or a few other wild ideas that I'll update you more about as they start to unfold.
--
As to new content in the pipeline, I did some reflection looking back on 2020 with a feeling of...well "disappointment" might be too strong a word, but something in that direction. There are some videos I feel proud of, like the one on the monster group, but for many others like the last one on medical tests, it felt like I was covering a topic out of a vague sense that it's what the channel should cover, rather than from having something novel to say, or out of simply delighting in a nice visual perspective of a clever piece of math.
At the moment, I've turned back to continue the differential equation series which I left off over a year ago. The next chapter will be on matrix exponentiation, which certainly is a beautiful topic. Also, it sets the stage for understanding a lot of other topics.
-Grant