3Blue1Brown
1 month ago
tl;dr: What would you most enjoy watching next? Many times, I’ve fallen prey to the sin of saying “in the next video, I’ll describe _____”, and then never actually making that next video. Or rather, I’ll start producing that next video, and decide for whatever reason that it wouldn’t actually be good. Past examples include describing how the Riemann Zeta function relates to primes, describing the Laplace transform, describing the beta distribution in probability, etc. The list, to my embarrassment, goes on. Most recently, in the quantum computing video, I referenced my intent to make a follow-on about the math of quantum mechanics that underlies what we described there. I’d aim to outline a concrete two-state system, and the mathematics used to describe it in quantum mechanics, and say a bit about entanglement between multiple such systems. While putting together an outline and notes for this, I find myself dreaming about doing a full series on quantum mechanics at some point. The math feels a little hollow until you put it into a bigger picture: Position/momentum operators, Schrödinger’s equation, harmonic oscillators, etc. Wallowing in these dreams, I’m wondering if it makes more sense to wait until I have a full plan for such a series and incorporate the lesson I planned to do next, namely, two-state systems + entanglement, into such a series. I wouldn’t make such a series immediately, because I know it would be a huge project, and one of my goals this year is to extend the linear algebra series by this fall, covering the singular value decomposition and the many subtopics it helps elucidate. A bit further down the road, though, it's a topic I'd love to visualize and to present more as a course than as a stand-alone video. I’m debating right now between reining in those dreams (for now) and doing that follow-on as promised, or instead turning to the older list and paying off the debt of a different promise from the past. Or instead, I could continue postponing all promises and jump straight into my linear algebra plans. What do you think? What would you like to watch next?
130K votes
Vote now
130K votes
The quantum mechanics underlying quantum computing
How the Riemann zeta function encodes the "chords" of primes
Intuition for the Laplace transform
Extension to the linear algebra series (emphasis on SVD)
4.7K
1.1K
3Blue1Brown
2 months ago
Pop quiz time! Try to answer without looking it up. Suppose I have a mystery function, and I tell you that there’s some secret value among all the numbers from 1 to N, where if you plug that value into my function, it returns True, otherwise, it returns false. For example, maybe it simply looks something like "f(n): return (n == 123)". But you can't see the inside of the function; it's a black box. How long does it take to find this secret value? For a classical computer, there's no better method than to guess and check all N possibilities. Sometimes you're lucky and get it early; sometimes you're unlucky, and it takes closer to N; on average, it's N / 2. In CS, we call that an O(N) run time, disregarding the constant because we just care about how things scale as N grows. Here's your quiz: How long would it take a quantum computer to do the same task?
183K votes
Vote now
183K votes
O(√N)
O(log(N))
O(log(log(N)))
O(1)
6.2K
1.4K
3Blue1Brown
3 months ago
I've been spending most of this week writing/rewriting the script for Grover's algorithm video, which aims to highlight a connection between quantum computing and block collisions. Man, I thought this would be relatively simple, but it turns out to be very hard to explain quantum computing if you don't want to assume any background (shocking, I know!). It's one of those topics where once you know all the setup and premises, the result itself is swift and satisfying, but all the effort goes into wrapping your mind around the premises! It's a fun challenge, but the turnaround time may be a bit longer than what I initially anticipated.
19K
402
3Blue1Brown
1 year ago
Oh hey, it looks like the channel passed Tau Million subscribers! This is cause for some kind of celebration. What do you think about a simple behind-the-scenes style video on how I animate, both showing what manim is and how I use it? If so, do you have any requests for a mathematically interesting thing to animate as a demonstration? Preferably something where the code can be short, but the final result has something interesting to look at. For instance, something like the illustrations of Lissajous curves would be perfect (see the animation and code below). Also, the next installment on Transformers is coming soon, there were a few other things going on in my life in the last month that pulled me away from it.
11K
869
3Blue1Brown
1 year ago
Shorts barrage update: Thanks again for your understanding while I've been moving the pile of shorts adapted from old lessons onto this channel. There's still a large pile more, and what I currently have scheduled is to keep posting some daily between now and the end of the week, then after that, have them go out only weekly every Saturday. I don't like how posting them necessarily spams existing subscribers' notifications with snippets of old content they may have already seen, so I'm open to suggestions here. The goal is just to get them to exist in the short feed, spaced out enough to give the recommendation algorithm a chance to learn which ones people like. For any of you who are curious, I was just looking at the analytics for this last week, and here are some conclusions - Even early on, it's pretty clear that shorts are an effective way to introduce new audience members to old lessons, as measured by watch time from non-subscribers on (non-short) videos. - People who land on a long-form video by clicking on the "related video" thing at the bottom of a short tend to spend more time on that video than those coming from other traffic sources (like suggested videos). I'm not sure what I'd expect here, but it's nice to see that people discovering lessons in this way evidently come in more invested. - If you compare the age breakdown on shorts vs. long-form, contrary to what I was expecting, the age range where you see disproportionately more people on shorts is 25-34. The percentage of viewers in the more youthful range between 18 and 24 is about the same for both formats. I'll be curious to see if those hold up, say, 6 months from now, when activity on shorts is more purely about discovery in the shorts feed, with less contamination from subscribers landing on them through the current barrage. --- In the meantime, I'm keeping busy animating the next new lesson. It'll be another physics one (though not optics), which is probably still at least two weeks out. Stany tuned!
7.3K
419
3Blue1Brown
5 years ago
Slight delays on the next video, which I was hoping to have out today, but it'll be published tomorrow. Get excited for some number theory!
3.1K
78