You can always add the "complicated" options behind toggles. Have it start simple by default. :)

Just learned about this program today, and find it interesting that even though it isn't specifically coded for this that certain "biomes" seem to appear in the simulated world. For example, when you start out most of the world is a "desert" in that there is no food out there. If a mover evolves and finds itself in a desert all alone it will most likely die. So after observing here are the Biomes I see: Desert: Nothing out here. A mover that comes out here will likely die alone and hungry. Plenty of room for producers to expand into. (Can become anything depending on what expands into it) Field: An area with bountiful food. A mover that finds a field will likely have a population boom. No room for producers to expand into. (Once movers find it, will likely become a Savanna.) Heavy Forest: Packed with producers. Nearly impossible for movers to navigate. Little room for producers to expand. (If it gets too heavy it'll likely die out and become a field) Light Forest: Has both producers and movers. Can exist for generations in equilibrium as the two life forms co-exist. (Can become a heavy forest or savanna if the equilibrium is disrupted through evolution) Savanna: No producers, but plenty of movers. Movers will wander around this area hunting other movers or just waiting for one to die of old age. The food that the dead movers produce keeps the population alive. (If food runs out entirely might become a desert, but if a producer is introduced could grow into a light forest.) It's actually interesting seeing what biomes tend to form next to each other. You can have a bountiful field protected by a very small heavy forest. Right on the other side of the heavy forest, you might have a savanna where your movers all roam starving for food unable to break beyond that heavy forest. Further away you have a desert just waiting for the world to expand into. But then that all changes when a single mover finds its way past the heavy forest and into the field. A population boom happens and the forest spreads into the area that the movers just cleared. Now the field has had a "climate change" that has taken it all the way to a light forest.

Told my biology teacher about this since I thought that it was a great way to represent evolution and natural selection and all of the factors that play into this. She said that she would probably use it next semester to teach about evolution. Keep up the good work!

It is an amazing project, with a lot of ways it can be utilised. That is why it is loved!

Very nice! Could you perhaps, even if only on a trial basis, change the eye cell so that it doesn't have to decide on a viewing direction, but looks in all four directions? Advantages: 1. the cell does not have to be rotated manually each time, which distinguishes it from all other cell types. 2. the cell gets rid of the little line that sticks out from the other cells and somewhat disturbs the pixel style. 3. since the cell is always part of an organism, some sides are obscured by other cells anyway, which makes it not too overpowered. Evolving visual range would also be exiting.

Nice work! Very curious to see how this project *evolves* over time!

It's a very nice project! At one point I tried creating organism sort of like eukaryotic cells, with small creatures moving inside and dependent on the larger ones. It sort of worked if you add little gaps for the smaller creatures to move between cells and made it more of a parasitic relationship than a symbiotic one. However one thing I noticed with this experiment is that the second a small producer develops, they have the tendency to "block" the reproduction of the larger organisms. So it's difficult for organisms to survive past a certain size because small creatures tend to just take over the whole map. So even a successful large organism can't reproduce if small creatures are blocking all of their reproduction options. Just an observation I noticed.

Please! This game should definitely have neural networks in it. I wish I'd have more tools to observe its NN state in the mod that you've linked. Btw, I've spent weeks fiddling with your game. I do believe that it has a tendency to fall into "perfect" organisms on a long run and an NN might bring some long term diversity to them.

I love the new additions! I was contemplating to add my own contributions, but at the moment I have too little time to get into this. One thing I was thinking of that the killer cells instead of harming the entire organism, just would harm the cells they come into contact with, in other words, damage parts of the organism, which is more what it is like IRL. That may be tricky to implement, though, because it has to interfere with the genetics, I think. Another cool option would be gravity, but that may become too demanding for performance.

You deserve way more subs man, the work you have put into this project really shows, very neat, right now I am screen-shotting my favorite evolutions, would love a method to save them in game!

I think adding other peoples stuff means its harder to update and maintain, but to keep it simple their stuff can always be made optional in some advanced features menu, so I don't think its bad in that regard.

This simulation needs a digging cell to dig through walls. I would love to see something like that.

Just did a quick read through of your github project and it's a breeze to read through. It also cleared up most of my neurosis about building cellular automata since I just kept seeing implementations iterate over a 2d array. I get that in a pure cellular automata, each cell is binary but once cells become multicellular, not multiple adjacent cells forming an organism but a cell that is nonbinary, a stack for the non null cells makes way more sense and I've yet to see that implemented in code until finding this. Thank you for all the hard work going into this and I can't wait to see what else goes into this I the future

Rn record for largest evolved organism is 2246 cells, good luck for anyone who wants to beat that record

I would like the feature where if there are cells outside the center cell (with mover cells) they could move on their own and be mutualistic by finding food for both of the creatures.

It is so fun to play this during school. Thank you!

This is an absolutely awesome project, and I'm very happy to see that it continues to develop! I've spent countless hours testing all kind of evolutionary concepts, just watching the life flow, or trying to overcome some challenge (like, to create a natural predator for purple flower, but no success so far on this one 😅) I'm very passionate about evolution and ecology, but I'm no programmer, so I can't implement anything in the code... But if I could, it would be probably one simple change - nothing revolutionary, not a gamechanger, but the one that I believe would make simulations more realistic and more diverse. As far as I understood, having a single mover cell is no different from having any number of them, and all the creatures with movers are moving at the same speed. I've witnessed multiple times, when a predator chases a prey, and a prey runs away, and they just move as a single unit across the space until the prey crashes into something (predator wins) or changes direction (predator loses). So... my idea is - it would probably be much more interesting, if multiple mover cells increased the speed of a creature, at least slightly. I guess that could lead to some new type of arms race, like an appearance of "cheetahs" and "gazelles" with over-the-top mover cell count that would prove beneficial for them 🐎🐆😀

Hey, i love your concept and how fun it is to just experiment with different setting and different starting organisms(i love games like this). I have a quality of life suggestion that i think shouldn't be that hard to implement and at least for me would be very nice. You see, i always play with the window at the bottom of the screen open to see changes to population etc. but as of now this window takes like 1/3 of the screen which means i either have to zoom out further, not see 1/3 of the map or set custom grid setting each session. I propose you let us make the "info/editor/evolution controls" window smaller, either freely or with a few different modes. Another even easier solution that would still go a long way is under "grid size" make an option for making the map short enough to have it on the whole screen(and not outside of it) while having the window open. I know im bad at explaining i hope that made sense and thank you for this project :)

I noticed this before I saw the video because I love this program so much!!!! Thank you everyone for this! I'm a traveling nomad so I don't have access to a computer often. I cannot thank you all enough.

Environmental factors would be interesting like having the chance to produce being higher on one side than the other as well as a difference between plant material and flesh based food

Love starting worlds and watching them evolve in the background while I'm working on computational cell biology!

Thank you for this cool project! Keep it up!

The walls feature is really cool because it allows for the evolution of several distinct stem-isolated ecosystems in the same simulation.

First time stumbling on this. Very interesting. Could probably watch this for hours!

This is really cool, for all the reasons you and commenters say, and I'm glad that a community sprung up around it!

It's a really interesting program. From the simplicity emerges complexity, and in that way it is a beautiful simulation of the basic components of life and ecosystems. I suspect if the right pressures for survival and parameters to account for evolutionary changes, even more advanced ecosystems and interactions could easily be explored. Like if organisms could grow new cells at life stages like metamorphosis, or adding true binary fission for reproduction, more cell types like a reproduction cell or such to add a heightened mutation factor. So many cool ways to expand the concept

This is going to be my first project to follow and learn ai and coding. i hope i can catch up to you one day

fire bro keep doin ya thing this is amazing

Well done man!

How this works is far beyond me, it’s extremely, extremely cool!

We need some way of tracking the evolution!!! Like a philogenetic tree. But it would need some naming system. Maybe simething like: first species: _A._ Then the next two species would be named _AA_ and _AB._ Sorry for my bad English.

Do you think adding new food types and mouths would allow for more interesting simulations? The way I see it it would allow for specific niches to be eventually developed

Some ideas: 1 Movers drop half food when they die. 2 Non movers cant have spikes/shields. 3 Non movers cant have mouths but instead have leaves/roots which can give them food. These are just my ideas so please add to more if you would like. -TheBlockyGamr

Yo sick i noticed this yesterday but i love this

Really good game keep up the great work

Loving it. Recordings of evolution leading to interesting organisms would be awesome. I often use it as a background on the tv when People are around, but its CPU is not great. Cell size 5 is already laggy. This means extinction is pretty rapid.. Just an idea

I think what everyone really wants is a modlist tab. That way it's up to the people who want the features to update. And you wouldn't have to maintain it.

this is my new favourite thing

New sub here, man this is very interesting.

its an awesome game, totally luv "playing" it i wish the colorings would be customizable tho :)

It would be great if you put the organisms to evolve on dry land and they laid eggs and you put new cells like sting, camo, healer and it can also have a realistic version, like having a button in the corner of the screen and then when the person clicks on the button they would go to the realistic version of the game and the person could choose what the realistic shape of their organism will be like.and you can add that people can play with the organisms they made

very cool

Cool new update!

Intriguing

In the vein of having a simple option, why don't you just create a default interface/settings alongside an advanced interface and settings? This would allow for the simplicity and ease of use that you intend to preserve while also allowing people interested in more advanced functions to have a central location to explore those more complex ideas.

I'm planning on starting a little project with this simulation

This remind me about Ken Stauffer's Evolve4.0 an old very nice proyect

One thing that would be cool would be to create a way to save templates of all surviving organisms at a point in time (e.g. after you ran it for a long time), and then import those templates into a new sim to battle it out against templates from other sims.

Can add small holes with gate and key mechanisms to the perlin noise walls, like how our cells filter.

sick

i've really enjoyed playing iwth this for the last couple of days! i look forward to updates! i noticed a bug on the editor tab, sometimes it disappears and i have to sample a lifeform in order to be able to use the editor again. i'm not a coder so i don't know best way to explain it. cool project though!!

I had an idea in regards to the movers movement, instead of having them move randomly each iteration have a variable that determines the chance of changing direction (overridden by the eye). Then if triggered they can only turn left or right 45 degrees unless they run into something. If they hit food they don't change direction, if its bad or a wall they do a 180, if they hit something else then random except in the hit objects direction. I think this will make them act more like ants.

I think it would be neat to have either metamorphosis or polymorphic generations(think fungi or fern with different forms for every other generation). Maybe we could see some creatures live as plants in odd generations, and "motile seed animals" in even generations. Or plant-polyps that metamorphose into animals at a certain age(or the other way like sea squirts)

Hmm...maybe creating a second perlin noise distribution (or perhaps a linear transformation rotating and flipping the original one) and using it as a negative would be useful for generating breaks in the walls. I'll check out the source code to see how complex that would be.

I was thinking about adding an armor cell which persist when a cell dies. kinda like how the outer skin cells are dead but still serve a purpose in protecting the insides.

I really enjoy the life engine. But in the next update could you make a tab with the history of all the creatures and when they come into existents and when they went extinct. Thanks for considering.

I'm addicted to this. I've found one pattern that seems to take over and remain dominant (I let it live for 4 million iterations and it never evolved away from it). Trying to find a "killer" that can overthrow it now

More complex life cycles wold add the opportunities to fill more niches It would also be interesting for the neural networks to be able to differentiate between spices and not just cells for communal relationships and parenting instincts

Could you possibly implement some sort of entanglement like behaviour into the cells. So that some cells can still interact and share information with each other despite being great distances apart

If you decide against neural nets for brains, you could make it a bit more probabilistic. Probability that it moves left, right, forward, backwards, turning, or not moving at all (relative to orientation), and maybe different probabilities depending on which eyes are currently active. I've noticed that the simple brains they have now aren't particularly versatile, as it were.

I had an idea for a new cell which just seemed super intuitive to me. I see that communication is just completely lacking here and i feel like it should be possible to implement. My idea is that there's a simple cell that can send out and emit signals and the cell's brain can randomly determine when to send out a signal and what to do with the ones it recieves. This way organisms might start sending out signals whenever they find food so other organisms of the same type can develop a behavior where they move to that direction when they recieve the signal. Also mover type organisms could use that cell to warn each other from threats so the individual organisms would be more likely to survive if they have this cell type.

One thing I noticed when playing around is that large organisms don't seem to develop much because they are easy to kill and have a hard time doing anything. Maybe instead of instant-death, killer cells can kill just 1 cell of an organism instead of instantly killing the creature (without changing the genome of course). For mover balancing, could make `movement_speed = amount_of_movers / (the size of the organism / 2)`. Single-celled organisms aren't going to move very fast, but neither are large creatures with only 1 leg.

What factors would be needed to develop really big and complex organisms? What is refraining it? Is it bidemensionality, world size or just queston of enough machine time?

I've been learning to write shader code recently, and I bet it's possible to accelerate the simulation by running it on the GPU. I think you could speed things up tremendously by making everything run in parallel. Maybe I'll contribute soon, who knows?

This is the next stage of computing. Just like how hobbies started the personal computer revolution. Hey @Emergnt Garden, have you talked to the guy from he @The Bibbits: Digital life channel? Ya'll have a lot of great work in common.

Would be interesting if something would be added to make organisms a bit smarter in navigating walls and compact spaces filled with other organisms.

Congrats! Which language did you use? I have similar ideias, but for complexer organisms

Any chance for a download? I would love to be able to run it in the background.

This may be a big suggestion, but you could streamline the modding process, so that users would be able to install them? Kinda like steam workshop?

One of the things I've noticed on every browser I've tested it on is a heap size overflow error that crashes the simulation. It doesn't matter whether its at a high resolution or low, after a few 100k iterations it crashes. It would be nice if there were multithreading support for javascript but I think that won't realistically ever happen for browser performance reasons (you don't want one tab bringing the whole browser down). I've shown a bunch of people in my family and friends this project but they're biologists and not code monkeys so its hard to get them to understand whats going on. So far the reception has mostly been "I don't see how this is any different than Conway's game of life", which is frustrating.

Do the organisms have some sort of short term memory? Or long-term yet...

Can you make it like Ant Life?

It would be great if you added blood/vain cells

hi everyone try the golden ratio in the cell size it makes path tracing somehow

What about a cell that can move the wall blocks

This would be super cool in £d but then again 3D can actually be harder to view sometimes.

Is there a way to run the code on my computer instead of on a server?

You should add the option for food to decay so that complete extinction would be more difficult to occur.

Maybe you could add different environments like ocean,desert, grassland etc.

Have you seen foo52ru life simulations?

I have a idea for The Life Engine: a cell that can move and rotate other cells of the body. Like a piston in Minecraft.

"Generate random walls" works only of default grid sizes. Try to use 1k x 1k grid and generate walls

Is this proyect full open source?

Still working on that please

Hey dude can you please add herds and predator packs

Is this still being updated?

In wich language is it written in?

The game is possible on mobile

generate random walls is hype

Is this on app store and can you breed

If you make everyone fork the project to make their changes, this makes the life engine itself go through some kind of "meta evolution".

How to play the game give me the gode to play

Step 1) Set cell size to 1 Step 2) Set lifespan multiplier to 1000 Step 3) ???? Step ?) Dopamine

A brain block that makes everything they do including Mutating Movement And Thinking

it can build a human brain?

DONT MAKE THE MOVER REPRODUCTION COST -999 IM WARNING YOU

The joel haver of coding

a*