Using Fractals to Deliver Babies |  

When these babies grow up, they're going to take over the world, being such good friends from birth 😁

You say 'about as good as we are going to get'. I would love to see the optimal baby friendliness solution. Is it more irregular? Is it known for arbitrary grid sizes? Even for this small grid an exhaustive search seems impractical.

Very good video. When I heard the premise, I immediately thought "huh, he's gonna mention Hilbert Curves," Very interesting take on a familiar topic

What I take away from this video is that it sucks to be baby 0 or baby 63, because they will only have one best friend (baby 2 and 62 respectively) but all of the other babies will have 2 best friends (The baby number n will be best friends with baby n-1 and baby n+1).

Is there a binary way to draw Hilbert curve?

Hey! Found your video on the SoME voting page. My process didn't allow me to write official feedback, but I liked your video and wanted to give you my thoughts anyway :) First of all, the graphics were great. I'm not sure if you were actually using manim, but your use of animation would fit perfectly into the 3b1b catalogue. I also felt that the animations were well-suited to the subject; for instance the final Z-curve picture is compelling on its own, but seeing it constructed live makes the details of the recursion particularly tangible. Also, as someone who's not particularly interested in fractals, I still felt like this video had something to offer me. The motivating metaphor was a good hook, and the substance delivered. I'm not sure how I feel about the explanation toward the end. To me it felt a bit gratuitous, but I can imagine it being useful for the intended audience. One mathematical question lingered by the end. You do a really great job explaining how the interleaving algorithm improves on the deficiencies of the lexicographic ordering, and I was expecting a payoff like "Indeed people have proved that this is (nearly) the best possible baby-delivering method." Instead we got a less committal version at 9:28 that didn't make it clear to me what the status of this solution is: best possible? epsilon-close? well-known? well-used? I wouldn't say that was a disappointment per se, but you got me interested enough in the topic that I was hoping for that kind of closure.

This is actually so cool! I actually heard about space-filling curves before, for example in quantum computing and ML, but I had so much trouble conceptualizing them, so thanks for this video :)

The nicest part to me is the way you manage to explain the dragon's curve with the practical example you started with. As you said, it gives more to the fractals than just telling how to construct and admiring their beauty!! Thank you for the video!!

What happens if you convert the interleaved bits to grey code? That would remove carrying. Does it then produce the hilbert curve?

I’d love a follow up where you come up with a “score” for each curve and see which is the best according to that score!

I think it is a pity that you have not expressed the task in concrete terms so that it could be solved mathematically. For example, you could have used the largest existing difference between two neighbours as a measure of quality. But it was also about larger distances. I'm sure that 15 and 48 wouldn't be good friends and when you showed a few examples in 9:05, you deliberately only ever stayed in certain areas to make it look very effective. So when you said "But unfortunately..." I thought you were addressing that, but then it was just about there already being a name for it :/ But after all, cool topic

wow, i learned a lot from this video. Thank you.

When you realize you're the center kid on the block with the shitiest friends that hate each other

MOAR please. I love these types of channels. I subbed.

Absolutely excellent. Really enjoyed that.

I don't remember where I have this from, but pretty useful is the place reserving identity of fractals. For example, if you increase the resolution of a fractal, for example a screen, the place where it was will only get more exact. But if you Increase just the pixels, the place will change. Thats a pretty need thing for user interfaces.

Super super nice SoME3 entry!

HAPPY BABIES..HAPPY FAMILY....

chapado

👏👏👏

Nice

Thanks. Very interesting problem, didn't know that fractal can be that useful.

I wonder what would the path be on a n by b hex grid

9:30 The Hilbert curve would be better.

A bit like 3b1b’s video on the hilbert curve, very nice!