Life in life
Phillip Bradbury6 HN recommenders · 2013–2025
4.1M views · demo · evergreen · Uploaded 2012-05-13
A two-minute video of Conway's Game of Life running inside Conway's Game of Life, built from OTCA metapixels and Golly. It zooms smoothly between scales.
The canonical visual for nested simulation, Turing completeness and self-reference; HN readers call it mind-blowing and reach for it constantly.
9:52Vectors | Chapter 1, Essence of linear algebra
3Blue1Brown
53:22Ruby Conf 12 - Y Not- Adventures in Functional Programming by Jim Weirich
Confreaks
19:42But what is the Fourier Transform? A visual introduction.
3Blue1Brown
18:40But what is a neural network? | Deep learning chapter 1
3Blue1Brown
24:26The Beauty of Bézier Curves
Freya Holmér
1:13:50The Continuity of Splines
Freya Holmér
1:00:43Jim Simons (full length interview) - Numberphile
Numberphile2
25:16But how does bitcoin actually work?
3Blue1Brown
20 comments from Hacker News
Not only that, but it can find lowest common denominator, solve a system of linear equations, and even run the complete RSA algorithm (albeit with a pseudorandom RNG that's deterministic). Or run a game of live's simulation [VIDEO]
I'd guess that GoL and Cellular Automaton Theory preceded and inspired rule 30 etc. So, maybe the architects implied metarecursion, that GoL metaphorically generated rule 30. GoL is Turing complete. I guess that means there is a starting condition that converges to a rule30 cellular automaton. edit: video of GoL emulated inside a turing machine in GoL [VIDEO] (link replaced)
Here's a nice video showing OTCA metapixels in action, implementing Game of Life in Game of Life. [VIDEO] It would be amazing to see this Tetris computer rendered into a video like this.
Molecular biology: [LINK] Game of Life: [VIDEO] Sure, one is 2D and discrete, the other is 3D and continuous... oh wait, Lenia is 2D and continuous. That’s interesting.
2017 post has a much-needed youtube video of the metapixels as the top comment: [VIDEO] I haven't been able to find one for tetris tho
I find this initially unintuitive, but it makes sense when you think about it. Look at any other system that produces aggregate emergent behavior out the behavior of smaller parts. For example, here's Conway's Game of Life implemented using... Conway's Game of Life: [VIDEO] It takes many many steps of the inner game of life to produce a single iteration of the outer simulation. In general, with emergent behavior, it takes a lot of inner steps to produce a single meaningful step in the outer system. So it makes sense that our cells are much faster than we intuititively think of as "fast". Because our intuition about speed is itself the product an emergent system. We are that larger Game of Life there, so it looks incredibly fast to watch the inner one according to our own time scale.
I remember seeing that video 3 years ago and being blown away. In case someone doesn't already know it, here is Conway's Game of Life simulated in Conway's Game of Life: [VIDEO]
My secret wish is that they start working on Factorio 2 with a new item state-memoizing game engine. Let me explain. Conway's game of life organisms' different states can be memoized. Starting the simulation slowly, an engine can memoize the following computations. This is what Hashlife [1] does. Quoting from Wikipedia: "the rapid increase in speed is often described as "exploding"." So much so that this kind of thing is possible: [VIDEO] Simulating the next ticks just entail looking up the future. Except for things the users touch. Since Factorio is a game about reaching the next scale of automation, I believe this is a worthy follow-up to Factorio 1. [1] [LINK]
They are a novelty that is often surprising in the how their simple set of rules can lead to quite amazing emergent complexity. That the one-time chaotic game of life was eventually shown to be able to stably infinitely expand and be turing complete is quite interesting. I don't know that they'll ever be useful for efficient computation, but they will likely always remain fun. [VIDEO] I have not read it myself, but I know that Stephen Wolfram ( of Mathematica/Wolfram Alpha) laid out some arguments for the usefulness of studying cellular automata in his early 2000s book "A New Kind of Science".
Right, obviously as scale goes up, much more complexity is needed to maintain the same level of behavior. A bit like how unicellular organisms can replicate simply, but humans need to go through a whole process to make an offspring. This Game of Life video shows what I mean perfectly: [VIDEO] Yes, there's a ton of complexity in there, but when you zoom out enough, it looks and acts very similar to the lower level.
Unknowable. And no, you'll have no way to break out of simulation and access api of the simulator, because if it's in place it works similarly to this: [VIDEO]
I missed that, thanks! I'll have to take another look. When I think of neural networks that evolve changes in their topology and the weights between those connections, over and over towards optimizing a solution to function, I get the vision of the recursive Life in Life video[1] where Conway's Game of Life is emulated in Game of Life. [1] [VIDEO]
One thing it doesn't mention is Conway's Game of Life itself. So of course, you can implement Game of Life within Game of Life: [VIDEO]
I think the author is misunderstanding the XKCD comic about rocks [1] (or maybe I am). Just because it's possible to run a simulation of the universe on rocks, doesn't mean that rocks are conscious or turing complete. You can't forget about the person who is manipulating the rocks. The system as a whole is turing complete. Likewise with the bar of iron example that the author gave, you can't forget about the person interpreting the atoms in the bar of iron. The system as a whole is turing complete (and also naturally conscious, because the person doing the interpretation is conscious). And there is nothing physical necessary to represent such systems. You can simulate turing complete systems inside turing complete systems [2]. So I don't see why consciousness has to be a "physical phenomenon" as the author claims. [1]: [LINK] [2]: [VIDEO]
FWIW: [VIDEO] If you allowed toggling particular points manually as input from a controller... probably not too far away from possible!
This is amazing! I've seen "life in life" as a video from ages ago[0], but this is just suck a slick implementation of it! [0] [VIDEO]
This is an excellent link, well known in the "community" around this. To add to it, here's a famous and really clear sub-2 minutes representation of the scales involved if it helps make things clearer, for those who've never seen the scales involved: [VIDEO] (and that's "just" the Game of Life in the Game of Life). I think I would enjoy helping push the state of the art in this field, but it's far beyond my abilities for the moment. If it wasn't, I would love to work on a new library specifically built for the kind of things like your idea, with the API aimed at working with metapixels-based "hardware", but with the lowest-level constructs (maybe by far) being thinks like RAM or CPU caches as building blocks - while still being able to zoom to the lowest level of cells in the renderer. But it's basically R&D and HPC combined if one wants to offer an interesting alternative to the established libraries. From what I know of upcoming hardware, such a library may have to wait a couple of years and basically target (as in, full buy-in, maybe at the programming language-level, certainly in tooling) the next Apple Silicon-like (or beyond) breakthough in commonly accessible hardware for a nice, "free" performance boost. Maybe GPUs are next?
Computers implemented in other computers. Specifically: Minecraft running in Minecraft [LINK] or Game of Life running in Game of Life [VIDEO] Bonus: Digital clock implemented in Game of Life [LINK] Surreal because it makes it apparent that our universe could really be emergent behavior from a simple set of rules.
> This combined with the fact those interactions have been happening for billions of years in parallel at such a massive scale does not make it surprising that persistent and self replicating patterns emerge. On a large enough scale, even improbable events become likely. I don't think anyone has proven the scales necessary for life to happen by chance. What you have done is worked off the base assumption that abiogenesis happened accidentally and therefore the time and scale needed to do so is the age and scale of the observable universe. Under the infinite monkeys theorem that the works of Shakespeare would eventually be produced, but we are not working with infinite time and resources here. We are working "just" 9 billion years and "just" 100 billion galaxies, which, for abiogenesis seems pretty short especially considering what % of planets have zero chance for life due to their composition and star proximity. How many billions of years of compute time across how many parallel computers do you think it would take Conway's game of life with random RNG seeds to simulate itself ( [VIDEO] )? Regardless of your answer, I still find it more likely Conway's game of life simulating itself points to an intelligent force of some sort having iteratively created it.
It might be argued that one of the dynamic programming (caching) ones would be the best by a reasonable standard. These can be used to simulate absolutely monstrous grids, and is where we see simulated CPUs and whatnot. Here's the game of life running in the game of life: [VIDEO] - that one uses Golly.