Particle Life: simulating "life" with 200000+ particles

Added: 2026-06-01

[00:00:00]The idea behind our simulation will be pretty simple. We're going to have a bunch of differently colored particles.

[00:00:05]And these particles are going to be attracted or repelled by each other depending on the color of the particles.

[00:00:10]For example, maybe red particles are attracted to blue particles but repelled by green particles. Maybe they also weakly repel themselves. Once we decide on all the rules, we just throw a bunch of these particles into a soup and complex things will emerge out of that soup.

[00:00:27]Now, I didn't come up with this. Someone else did. It's called particle life, and I'm not entirely sure whose invention it is, but it's not mine. Anyway, since the rules and details have been very well documented already, what I want to do in this video is explore the different types of patterns that commonly show up in the simulation, as well as what causes those patterns to exist. When you load up the Particle Life website, which I've linked in the description, you will see seven different colors of particles with 49 different forces acting on those particles. Each of these 49 forces will be randomized as well, so you will get a simulation that no one else on Earth has ever seen before. And you may ask, what are all these things? And why are they moving around? Why is there a snake?

[00:01:04]Which of the 49 forces are responsible for the snake?

[00:01:08]And what is going on over here? It looks like the moving blobs are splitting apart and merging together. Why is that?

[00:01:14]And do all the other simulations with different rules exhibit similar behavior? Remember that that was just one simulation randomly chosen out of countless other options?

[00:01:23]As with any puzzle or problem, it's easier to start with the simplified, albeit less interesting, version of said puzzle or problem. So, let's start with only one type of particle, which means only one type of interaction, green on green. The green particles either attract themselves or repel themselves.

[00:01:37]But if the particles were really repelling each other, you would expect them to spread out. So, why are they grouping up? The reason is because of something I haven't mentioned yet. If every one of these particles exerted a force on every other particle, I would need a supercomput to run the simulation. So the people who wrote the code made it so that particles will only interact with particles that are somewhat close to them. There is a maximum radius and our particle will not interact with anything outside the maximum radius. Now for reasons I don't really understand each particle has a minimum radius as well. So if there are particles within our minimum radius, our particle will not interact with those either. So that's what's going on here.

[00:02:14]All these particles are close enough together to not interact with each other. And all of these blobs are far enough away from each other so that their distance just barely exceeds the maximum radius. So they don't interact with each other either. Nothing is interacting with anything. The system is stable and resistant to change unless I change the rules.

[00:02:32]If I increase or decrease the minimum radius, the blobs get larger or smaller correspondingly. And if I change the maximum radius, the particles rearrange themselves until the distance between the blobs matches the new maximum radius. The blobs are now further apart because the maximum radius increased.

[00:02:47]Now if I make the force attractive instead of repulsive, pretty much the same thing happens except the blobs are circular. Now the diameter of each blob is about the same as the minimum radius and the distance between blobs is greater than the maximum radius.

[00:03:00]However, the fact that each blob's size is limited by the minimum radius has an important implication. If you keep adding particles to the blob, it becomes unstable. There is a critical mass past which blobs will suddenly destabilize. They do eventually stabilize again after they eject enough material, but that can take a while. We will call these stars, and they're going to be quite common. Stars do not form when the force is repulsive because in this case, adding particles to a blob will just cause it to split apart. All right. Now, if we have two types of particles, we have four different forces, which are shown on the left in a matrix. In case you're curious, here's how to read the matrix. Let's say you want to know whether or not green particles are attracted to violet particles. We would find the green particles row and the violet particles column and we would look at the corresponding squares color. If it's blue that means attractive force and if it's red that means repulsive force. The intensity of the color corresponds to the strength of the force. So we know that green particles are strongly attracted to violet particles. It should also be noted that each of the four forces has their own minimum and maximum radius. All right with two types of particles spaceships are now possible.

[00:04:06]The idea behind a spaceship is green is attracted to violet, but violet is repelled by green. So green just ends up chasing violet around. Now in an ideal spaceship, three things should be true.

[00:04:17]The attractive force should be slightly stronger than the repelling force because otherwise Violet will run away faster than green can catch up and the spaceship will fall apart. The attractive force should also have a larger maximum radius than the repelling force for the same reason. Each type of particle should also be attracted to themselves to keep the two parts of the spaceship intact. If the attraction force is not strong enough, the spaceship becomes fragile and more prone to splitting apart. A consequence of this third fact is that whenever you see spaceships, you will almost always see stars. In general, stars form whenever particles are attracted to themselves.

[00:04:52]The fact that there are spaceships moving particles around also helps stars grow faster. And it turns out that the more mass a star has, the more violently it erupts. Stars also retain the properties of the particles that compose them. So, a green star will chase a violet star around to form an unstable spaceship. Stars can also contain both violet and green particles at the same time. These tend to move around erratically, as you can see. Or maybe you can't see. I don't know what YouTube compression is going to do to this video. Anyway, snakes. Take a guess.

[00:05:21]What combination of forces is going to result in the formation of these snakes?

[00:05:26]The answer is when the two colors are attracted to each other but not to themselves. Sort of like magnets or electric charges. Here is what the matrix looks like. Why does this form a snake? Well, remember earlier in the video when there was one type of particle that repelled itself in that case, blobs formed and we are going to have something similar with the snakes because both types of particles repel themselves. What I'm trying to say here is a bunch of blobs are going to form.

[00:05:51]Now blobs of opposite color attract each other. So now we have a pair of blobs, but this violet blob is going to attract another green blob. But remember that green blobs repel each other. So these two green blobs are going to end up as far away from each other as possible.

[00:06:04]This cycle repeats itself with more and more blobs joining the snake. This process does take a while though, so snakes don't typically appear until later on in the simulation, especially when you have more than two types of particles because those extra particles are going to run around and interfere with the snake formation. Anyway, some more things about snakes. The maximum radi affect how spaced apart your snake is. and the minimum radi affect how thick your snake is. More specifically, the minimum radi of the green row of the matrix affect how thick the green part of your snake is, while the minimum radi of the violet row of the matrix affect how thick the violet part of your snake is. However, too much adjustment of these values will cause the snake to break apart or become extremely fragile.

[00:06:42]So, snakes are not as commonly seen as you might expect. Rings, on the other hand, are extremely common. They tend to show up whenever two types of particles are attracted to one another. This may seem counterintuitive because if two blobs are attracted to one another, you might expect them to just stick close together. But in most cases, one blob will swallow another to form a ring. The particles with stronger attractions to themselves or a smaller minimum radius are more likely to go in the center of the ring. The particles on the outside typically don't really attract or repel themselves, which is the reason they are able to spread out into a thin ring. If they were to interact with themselves to a significant degree, for example, let's have all these green particles repel themselves.

[00:07:21]the ring becomes broken. Each of these three parts repel each other which causes them to be equally spaced out.

[00:07:27]The size of each of these broken parts is dependent on the minimum radius and the distance between broken parts as well as the number of broken parts is dependent on the maximum radius. Rings are a pretty complicated business. The last pattern I want to look at is rotation. Rotation is somewhat rare and can only occur when you have at least three types of particles. I'm pretty sure it happens when color A is attracted to color B, color B is attracted to color C, and color C is attracted to color A. So, they just chase each other around in a circle. And also, each color needs to be attracted to itself, otherwise the whole thing would fall apart. The matrix would look something like this. Rotation is quite unreliable, though. Sometimes a matrix looks like it'll have rotation, but then I run the simulation and nothing happens. Maybe one of you can find a reliable way to determine whether or not a simulation has rotation. I know it has something to do with the minimum and maximum radi, but I don't know what.

[00:08:17]Anyway, now that we understand the basic patterns in particle life, we can look at all of them together. The settings I'm using here are 200,000 particles evenly split between seven different colors. I also increased the world size and made the boundaries wrap. By the way, all the forces in all the rules were randomly generated. Here is our matrix.

[00:08:36]Anyway, you can see rings and blobs starting to form, which is generally the first thing that happens in any simulation. Spaceships are also forming.

[00:08:43]They are rather slow right now, but they will increase in speed as they accumulate more particles. In general, the more particles a spaceship contains, the faster it will travel.

[00:08:52]Let's zoom out a bit.

[00:08:55]If we look at one of these fastmoving spaceships, we can see that it is composed of pink, cyan, and red particles, which is strange because the slowmoving spaceships that we saw at the beginning were also composed of blue particles. Where did they go? Anyway, if we look at the matrix, we can see that red is attracted to cyan, cyan is attracted to pink, and pink is repelled by red, which is basically the same thing as the spaceships we looked at earlier, but with extra steps.

[00:09:18]Surprisingly, red is strongly repelled by itself, and the spaceship is still able to hold itself together. Notice how most of the blue particles are just floating around in the background. This is mainly because they repel themselves and do not have any strong attractions with any other colors except with cyan.

[00:09:33]You can see that small pockets of blue form near cyan blobs. This is the reason why slowmoving spaceships contain blue particles. They just latch on to the scan and stay there. But past a certain speed, the attractive force isn't strong enough and the blue particles fall off.

[00:09:47]If you want to determine how active a specific color will be, first look at how that particle interacts with itself.

[00:09:53]Then look at the corresponding row in the matrix, then the corresponding column. If you see a lot of repulsive forces, that means the particle will likely sit in the background and not really do anything. If you see a lot of attractive forces, that particle is more likely to be active and participate in star formation, especially if it is attracted to other active colors.

[00:10:10]Anyway, you can see there's a new type of spaceship moving around slowly, and we'll call these bugs. This bug is actually quite resilient. In fact, earlier it survived two collisions with the pink spaceships. It can do this mainly because pink particles are repelled by green and yellow. So, whenever a spaceship crashes into the head of the bug, it simply gets deflected. And whenever a spaceship crashes into the tail of the bug, it gets eaten because the tail is made of the same stuff as the spaceship. After a few more minutes of the simulation running, there are a lot more of these bugs now, and some of them have accumulated enough mass to become unstable. If we zoom out a bit, we can also see that there seem to be less pink spaceships than there used to be.

[00:10:47]This is probably because pink spaceships are made using cyan particles, and there are a lot less of those available now since the bugs have a large scan component.

[00:10:57]Some of the cyan is also locked away in these stars. And because pink is repelled by red, the cyan in these stars is effectively unusable for pink spaceships. In general, stars become more and more common as the simulation progresses and particles get moved around. If you look at where my pointer is, you can see one of the stars turn into a bug by attaching to a yellow blob. The main propulsion method of these bugs seems to be cyan and yellow.

[00:11:19]Cyan particles are attracted to yellow particles, but yellow is repelled by cyan. The other colors don't seem to contribute much to movement and instead are only there because they are attracted to cyan and yellow. It is possible though that the red and green particles evolved to be there because they make the bugs more resistant to pink spaceship collisions. Pink is repelled by both red and green, but neither of those colors feel a significant force towards pink. So, red and green particles can deflect pink spaceships without being moved themselves, meaning a bug is more likely to survive if it contains red and green particles. You may have noticed that this whole time magenta particles have been doing absolutely nothing. And if you look at the matrix, this makes sense. Magenta's row and column are both filled with red, meaning nearly every color is repelled by magenta and magenta is repelled by nearly every color.

[00:12:06]Magenta is attracted to itself, which is why there are circular blobs of it everywhere. I suspect that if I ran the simulation for a few more hours, the spaceships and bugs would push the blobs around and they would eventually merge into magenta stars, but that's only speculation. I later ran the same simulation but with 500,000 particles which my laptop did not appreciate.

[00:12:25]Pretty much the exact same thing happened. I don't really know what I was expecting. All right, that's all I have for you today. If you want to try all this stuff out yourself, the link is in the description. Hope you enjoyed and I'll see you next week.