"aww look at this tiny island it looks like someone died here" -cary 2016

@slate1496

5 жыл бұрын

Two years ago

@randomderpyfamily4795

5 жыл бұрын

Hello Linxcion!

@chasington5102

4 жыл бұрын

Auquix 3*

@thunderboltpoplays7758

4 жыл бұрын

Auquix The Boat *three

@moises3545

4 жыл бұрын

Bbbbllllleeeeaaaarrrrruuuugggghhh

My two changes that I would make would make would be: 1) drastically remove water, or make water a needed resource along side food 2) death makes a uniform color that isn't present in grass (meat). Maybe remove red from grass and make it exclusive to death.

@Bibbedibob

7 жыл бұрын

right, that way scavengers could evolve

@speadskater

7 жыл бұрын

meat would have to degrade over time to make it so the whole map isn't covered in red. the rate of this would change evolution I think.

@irchonite1953

7 жыл бұрын

I agree, I think the color of death should be red and not the color of the creature, or maybe a slight mixture of both, but mostly just red. Adding a hydration amount to each creature where they need to go on water to increase it might make it better and much more interesting, but it also might be too complicated.

@Luaporleafcutterant

6 жыл бұрын

Archeryte that way predators can camp near the water.

@revimfadli4666

5 жыл бұрын

Would be interesting with recurrent networks such as NEAT or MGU(or both - why not?)

I think the color change from the dead bodies should decrease over time if it currently does not so that the chaos in the north will slowly return to normal grass colors ant it would be like the grass is growing off the Yumms of the dead body

@phlimy

7 жыл бұрын

You are totally right!

@Alice-mo4vs

7 жыл бұрын

Decomposition.

@TheBasikShow

7 жыл бұрын

+

@irchonite1953

7 жыл бұрын

Great idea!

@hdef6602

7 жыл бұрын

+

Hey man... I just found your videos and I'm completely in love with your evolution simulators. As a Biologist... it's fascinating seeing these creatures (even though you get the biology wrong sometimes). I can't wait to show this to people... One question though... Why are you letting the creatures able to change their own colour and the colour of their food source? Here's my thinking as a Biologist to make this more "realistic". The neural network should only be able to control the behaviour of the creatures. While their physical characteristics (colour and food source) be only driven by natural selection. This way you also can determine what "species" the creatures are. I would also add 2 more variables... Strength and Agility. Where "Strength" makes the creatures able to kill others in less hits... and "Agility" makes the creature able to sprint faster out of harms way. And maybe this could give rise to some predatory behaviour. But who knows... I know jack shit about programming.

@noahgorny8534

7 жыл бұрын

I totally agree with you, in fact the colour-changing also means random food drops which are usually toxic to the enviorment, making the whole point of predators useless. Also about the Strength and Agility- if something like this was added, those abilities should have drawbacks to deny to creatures from simply massing them both.

@TheLeeringMachinist

7 жыл бұрын

You would have to put a soft cap of some sort on strength and agility such as reduced metabolism (more energy loss per frame) or reduced growing speed else you would just have infinitely fast ad strong creatures. A lower hard cap (e.g. a certain creature dies when their energy reaches 50 rather than 0) and upper soft cap (e.g. a certain creature's agility is exponentially reduced when their energy goes above 150) also be put on a creature's size which would increase biodiversity and make distinct "species" more obvious. Another change that could be added if you stopped creatures from changing their food hue is reducing the range of food colours to ~4 with 3 being different types of plant life (cellulose, fruits, vegetables) and the other being meat which would be dropped upon every creatures death. Meat should decay over time to reduce the random colour splotches seen in the video. Having creatures with a tolerance to multiple types of food could again be soft capped by increasing metabolism.

@Jazzafritsch

7 жыл бұрын

i dont know if he removed it but he did have a feature that increased speed but it was only really effective to run from harmful environments such as water or low food

@AndreRhineDavis

7 жыл бұрын

I figured that the colour change would be a way to introduce communication. So like, the creatures could evolve so that certain colours have certain communicative meanings, e.g. as a mating call, or to intimidate others, etc.

@teovinokur9362

7 жыл бұрын

Good idea

I suspect that some of the complex(ish) behaviors that you hope to see, are simply not possible with a neural net of that configuration. You can probably get much more complex behaviors by having more hidden layers and/or more nodes per layer and/or even a totally open-ended neural net configuration that can evolve arbitrarily (but make it so that having more nodes require more energy upkeep so that they don’t just grow ridiculously without achieving anything).

@TheAmbientNoise

7 жыл бұрын

I concur. A single hidden layer is not enough.

@ZweiSpeedruns

7 жыл бұрын

oh my gosh it's timwi. I recognize you from ppcg. Hello!

@JakeSandlandTheJirachiAKARoyNe

7 жыл бұрын

Hallo there

@maixck

7 жыл бұрын

I was thinking the exact same thing, thanks for saying it.

@maixck

7 жыл бұрын

I don't understand your point, the cost of having neurons is pretty expensive evolutionary, our brains take 20% of our bodies energy, the entire nervos system must take a ton. Or are you saying that it's easy to implement but it would be hard on the simulation that´s true, but i would argue that it is more important than being able to run 10 thousand generations on designs that simply can´t exhibit the complex behaviours we want.

Here's an idea, Adding erosion/whatever over time so the land changes (not just with food-related things). That island being empty kind of annoyed me, so maybe after a while a land bridge can form between it and another island, or an existing land bridge somewhere else erodes away!

@voomroom8697

7 жыл бұрын

Andre Wee That works too.

@eriks1765

7 жыл бұрын

+Andre Wee (Anduin1357) look at the human migration to America though...

regarding the types of food being poisonous to the creatures you could make it that certain colours are just harder to digest (eg take more energy to eat) becoming less rewarding but not fatally harmful.

Wouldn't it make sense for yums dropped by a dead creature to decay over time. This would reduce "pollution" between climate zones

@Keldingull

7 жыл бұрын

or have them drop the opposite color to what they are since the current scheme makes them cannibals! :V that or simply make it increase the yums under it without color change.

@dmarsub

7 жыл бұрын

if mouth and body color would be more static, it could implement the chance of a carnivore to develop. in nature dead animals spoil the food around them for a while, but other creaters transport them away (bugs, scavengers) it needs to develop a balance, between a cleaning troop, and the plant eating population. multicolored mouths would be interesting, so either an animal can focus on one kind of food and be very efficient at digesting it (cows) or spread out on 3 or 4 kinds of food but make it cost some more enrgy to maintain, or be a bit less efficient.

@MsHojat

7 жыл бұрын

That, but in my opinion creatures shouldn't drop food of their own color anyway. I think it would make sense for creatures to drop food of a color that doesn't normally grow, which represents meat. Having the automatically growing food be any spectrum of color seems strange and problematic.

@Keldingull

7 жыл бұрын

How about yums and carni-yums, One's vegitation and the other's meat. This way we get carnivores, herbivores and omnivores. simple.

@aliciabaumgartner1406

7 жыл бұрын

Yeah but that wouldn't work with the colour scheme. I think just raising the square's food content ("fertilising" the soil) would be easiest.

Please open source this, I would love to play around with it.

@swizard

7 жыл бұрын

If you actually watched the video, he literally said he would release it soon

@danielbaird1295

7 жыл бұрын

He has been on the fence about it for a while, so it's relevant to voice our opinions on it.

@swizard

7 жыл бұрын

ah in that case my bad. I too hope he releases it soon!

@RubyPiec

4 жыл бұрын

@@coderk8298 yearlate, but thanks

Great work! That's a comment that's not posted as often as it should have been here.

I just watched all of your evolution simulators videos today, the work you're doing is remarkable!

Wow, you really are an amazing person. This is really awesome work and the fact that you're even thinking of open-sourcing this is really generous. Keep it up man, you're doing great!

could you make "bodies" decompose and turn the grass back into it's original color?

@chasecomes

7 жыл бұрын

you could also make the fertility on that tile improve for a short period of time

@Adam-zt4cn

7 жыл бұрын

It actualy makes sense, as the climate only favorites one type of plants, and it should get reverted to what it was. The Hue could slowly get closer to the climate value, as they are the same at the start, but the YUM value should stay the same - energy can not be destroyed (if we do not take heat loss in consideration)

@octosaurinvasion

3 жыл бұрын

its

I LOVE HOW YOU SPEAK. don't change

My worry is that, now that the food on the ground can be basically any color, and so can be the other creatures, it will become more difficult for the creatures to recognize that they are looking at another individual, rather than just a random patch of grass in that particular color, and thus that social (be it hostile or benevolent) interaction becomes less likely/possible.

@Phagocytosis

7 жыл бұрын

I understand that's how they see the color of individuals, but if they only recognize color and both the ground and inviduals can have colors in all hues, there doesn't seem to be a way for creatures to distinguish ground from other individuals. I think they can see either one, but presumably they can't tell the difference. That may be what you meant in your last paragraph, as well.

@Phagocytosis

7 жыл бұрын

Andre Wee No, that's not right, look at 7:13 in the video.

@Phagocytosis

7 жыл бұрын

Andre Wee No worries :)

@MsHojat

7 жыл бұрын

The big problem I see is that the colors of creatures vary so much, and accordingly change the food they drop by 100%. Considering that these creatures are all assumed to be rather similar sort of creatures (even if they have very different behaviors or traits), they should all drop similar color of food. Many predators don't care if they're eating crab meat vs chicken meat. Crabs and chickens are insanely different creatures, but it's still meat nonetheless. Auto-generated foods should have only 2-3 different sort of huge ranges that it can be, while meat should be a whole other hue range (specifically quite narrow), which is either always the same, or varies just a bit based on genetic makeup. As it is now, likely behavior to see for smart creatures would be those who eat both themselves (cannibals) and their natural environment food which matches their body color. It's a rather nonsensical thing, at least to Earth biology - at least as far as I know.

You should make it so the mouth hue is an input as well as an output. Unless I'm understanding incorrectly, the creatures have no means of figuring out what their outputs are presently. Additionally, it'd be cool to see you make the hue things be on an add-to and subtract-from basis rather than being from 0 to x, so we'd see less red creatures

@pixequil

7 жыл бұрын

Also, make it so the dead bodies fade after a while. Less toxins

@Phagocytosis

7 жыл бұрын

Excellent point... they definitely need to understand what their mouth is like. I mean, I suppose they could always change their mouth hue such that it fits what they have beneath them as they eat, but that's one strategy; for an alternative strategy whereby they focus on one food type, they need to know what they can actually eat.

@doctorpc1531

7 жыл бұрын

but the "what is beneath me" is ALWAYS a better strategy, no? Even if all the grass was one colour, that would still protect you from dead things.

@Phagocytosis

7 жыл бұрын

Yeah, that seems to be the case, but it might be tricky to evolve, or something? At any rate, I think it's good to give them such options when possible. Sometimes when you simulate evolution, things that seem intuitive and obvious turn out to be wrong, and that's why you model it. I very much doubt our intuition that the "what is beneath me" strategy is better is wrong, but you never know!

@doctorpc1531

7 жыл бұрын

Vincent Oostelbos No, you literally just need a connection between "OHue" and "MHue". That single connection gives you immunity to food poisoning and allows you to eat on all coloured tiles. If you want to see the maths, go down to [1]. The thing with evolution is that it finds the path of least resistance. If you are trying to solve a practical problem of eating coloured grass and mouthchanging is an option, then by all means keep it in. What we are looking for, however, is recreational problem solving. We aren't as much interested in if the solution is "the most efficient" as we are seeing if it is *interesting*. That's why it is important for us to block off paths that are so obviously better that they kill off diversity. Evolution, by nature, creates dull solutions. That's because they work the best, And it's our job to create arbitrary rules to limit this dullness. [1] Assuming they can connect directly, it happens every 144th time a new link is born. (=> a common mutation with instant and major benefits). If not,it "sets up" every 132nd time and then has a 1/144 chance of activating. (=> A common mutation with no immediate benefits, considerably slower to spread). (Unless, of course, both mutations happen at the same time, which is 1/19008 of the cases)

Cary when watching this i thought creatures need water to survive too. So you should add that.

Please release a download link!

@rapids7841

7 жыл бұрын

plz do dat

@Toad_Hugger

7 жыл бұрын

Yes, please XD

@blazingnest2988

7 жыл бұрын

please do

@FainaruMr

7 жыл бұрын

yup pls do

@stephenwatson7349

7 жыл бұрын

please yes

Ooh! Just hours after I found this channel and fell in love mostly with this simulation/game!

Wow. I remember BFDI... Binged it in one summer. Great to see you're still active. Can't wait to see more 'Volvio.

make the map bigger please it will encourage regional evolotion think about it

@timothymclean

7 жыл бұрын

That makes it slower. TANSTAAFL.

@speadskater

7 жыл бұрын

with much less water

@phlimy

7 жыл бұрын

Yeah it would be really nice if he made it bigger after he optimized the yums calculations! :)

@speadskater

7 жыл бұрын

I'd like to see dependence on water if water were kept at this degree.

@Tinigame

5 жыл бұрын

nice

Hi, I love this simulator! And also your videos!

Omg! I love your evolve simulator! You're awesome! And your voice is totaly fine, very pleasant even

Great series. This game has inspired me to make my first evolution game! Keep up the good work!

it would be cool if instead of only eating food they'd eat each other when they encounter someone that is the color they like :D

@lukefrancois6173

7 жыл бұрын

that would be cool

@martinkunev9911

7 жыл бұрын

Seems like cannibalism to me. I think it would be more interesting to see predators.

@BxPanda7

7 жыл бұрын

Martin Kunev Or maybe just some species evolve to become meat eaters while the rest eat the grass, see if it would balance itself out or if one race would dominate all the others !

@alexwang982

5 жыл бұрын

Fight

Im here for evolution and your voice is fine

Holy shit, you bringing up Battle for Dream Island brings back so many memories of when I had watched it a long time ago

Great videos man! Love the evolv vids, evolution sim too. Subbed!

Damn, this needs to be open sorce. It'd be an excellent teaching tool like the first one was. Lesson 1: a good evolution sim deals with most things that make coding fun and challenging. Lesson 2: don't put entire source in a single file.

When he releases the source code someone should run a billion generations with a large population on a super computer.

@quin2910

7 жыл бұрын

yeah it would be very interesting what happens after 100'000 years.

Cary, this is awesome. I can't wait for the next one! Good luck with the efficiency increase~

I am really excited for you to release this. I intend to thread off simulation so that I can use a cluster of raspberry pis to process full time and see what results can be achieved.

I was just looking at how old Dojhilr got...

@sixadamra

7 жыл бұрын

I was waiting for him to reach 100 years old but the video was cut before that :')

@redstonetechnician1993

7 жыл бұрын

***** Yeah.

@quitting_the_internet

7 жыл бұрын

Redstone Technician Don't mind me I just went to your channel to be disappointed by no redstone videos

@redstonetechnician1993

7 жыл бұрын

Grimm Aww, you wouldn't be able to sleep if you heard my voice.

Cary shouldn't it have its own mouth hue as an input so it can compare it with the tile below it?

@animowany111

7 жыл бұрын

Well, if the simulation ran long enough the creatures should be able to set a direct neuron connection from the hue under them to the mouth. That would make them be able to eat anything.

@kevinclon

7 жыл бұрын

Not quite tho, Becasue it can be eating upto 4 tiles at a time

you're getting better! keep it up, man.

This is pretty cool. Keep up the good work!

Ok, first I must say I love your videos. Thinking about this last one in a biology course, I had an idea that you might appreciate. One of the problems of your simulation, and of evolution simulators in general, is that you always get homogeneous population, because every cell can bread with each other and therefore traits spread much too fast in all the individuals, leaving very similsars individuals and very sad subscibers. I suggest you add the notion of specie to end this. Every specie, in my idea, would have a specie name of three or four letters added to his short individual name, for exemple Yul-Ghud, Yul-Afae or Yef-Arag. Individuals of the same specie would be able to breed together only (for exemple, Yul-Afae can have progeniture with Yul-Ghud but not with Yef-Arag or Gal-Axelod). Modifications, of one letter only, would happen randomly and very rarely (like one out of 1000) so that you can have a pretty large group of one specie evolving together. Having this specie system would have those advantages 1-Differenciation of population. Supposing you have a large group one specie (let's say the Yul). At some point there is a «sexual» mutation and one individual becomes a Yuf. Supposing he survives and breed (assexualy) well, he and his desendants will not be affected anymore by the neighbour population, so one hundreds years later, you may end up with very different Yuls and Yufs and you are happy 2-Carnivorous species and other specializations I think that having different species would make carnivorous animals possibles, because there could be an equilibrium between carnivorous and herbivorous, whereas right now if it happens, every creature may gain those behaviors and there will be a huge decrease of population. By the way I think, instead of getting every corpse the color of it's owner, you should get them a specific colour (logicly red, and red being excluded from grass colors). I'm sure some may very soon evolve into eating red and eventually become carnivorous). 3-Creation of Darwinist trees Right now it would be insanely chaotic. But if you included in these, let's say only species with 100 of more individuals generated, it would be insanely great to understand the evolution process. 4-Better differentiation of relations between individuals This one is pretty obvious I hope you read and consider this (long, I know) post. If you have questions (or want to collaborate a bit with me (OMG the dream)) just leave them in the comments. Keep doing this insane stuff!!!

@nizwiz._.5906

5 жыл бұрын

This seems really cool, I'm going to try to change the code of my download to fit that, wish me luck!

What if with the addition of predators, you added a noise level and hearing to each animal? Each organism would produce noise either leaning more towards quiet or loud that could be 'heard' by other organisms. You could develop predators who are blind thus rely on their hearing to hunt and a number of other interesting outcomes!

I'm so happy you posted this :D

THIS is amazing!! I LOVE it!

you could try to add an option that disables the rendering part of the simulation, this way when you're away it doesn't have to render the entire screen. if this is possible or not probably depends on your way of programming the simulation but it would probably speed it up quite a bit

Imagine we could eat grass then we would save a lot of money

@kuko-lukas

6 жыл бұрын

vegans

@TheFlutterQueen

4 жыл бұрын

Except for in deserts.

@Poetawesomendo

4 жыл бұрын

Miluna_Mimi I made that comment three years ago

This actually one of the coolest things I have ever seen. People have spent years of their life doing this and you coded a program to do it. wow. imagine this thing on a super computer. You could right a thesis on this.

@alexwang982

5 жыл бұрын

DevotedFire *write

Cant wait til the next video. You earned a subscriber

Are creature able to kill ohter creatures? if not will this be added? I think creaturs shoud be able to look around

@sebas200425

7 жыл бұрын

1st yes

7 жыл бұрын

Yes, they can kill other creatures and yes, they can look around.

13:13 Poor Jcphoh

Heya, you should totally add a funktion that makes it so that if 2 tiles that are next to each other have different colors, their colors start to match. Love the vids and the hard work, keep it up!

Brilliant! Love this kind of stuff :)

FINALLY

Often the rendering also takes much computational power. So maybe you can speed it up by adding an option to disable rendering.

@dexterovski

7 жыл бұрын

I have the same opinion. Rendering all this stuff is stupid when what you really want is fast evolution.

@MsHojat

7 жыл бұрын

Andre Wee Well, it does still depend how the graphics are being rendered, but yes if it's using efficient techniques, it would be no problem.

What you need to implement is Biomes, and biomes transition areas. That will give you the "continents" idea you were mentioning. Love your work.

a few videos ago you mentioned wanting to give them the ability to know their x and y position. what if instead, they can temporarily remember areas (this would last longer depending on how often they visited there) and know their x and y while they are there. This would mean they could remember where there is lots of food. this could introduce 2 new features, memory strength and curiosity. memory strength would make them remember for longer or less time, and they could even remember the size and strengths of other creatures which are there and how often they are there. Curiosity would determine how likely they would be to explore areas they don't know yet. this would solve the problem of the island with just one creature on it in this episode. the amount food there is in an area could affect this stat too, as a creature might become desperate for food and explore.

Hi! I've been making stuff similar to this, and discovered this channel by accident. I'm fairly new to it, maybe just a week. My first one is a bunch of randomly coloured particles which bounce around and occasionally die, and sometimes breed a new particle of average colour. You hover over them, to kill them, and you lose when the population hits 200 (Starting at 12). You almost always lose, because they particles which are close to the background colour are harder to catch, and breed out of control. Would love to make something with you!

@keksimusmaximus444

7 жыл бұрын

micomator This sounds neat.

@micomator

7 жыл бұрын

Carlos Fernandez Thanks! I need to watch more of these videos. Programming is easy, biology is hard. The one I'm planning at the moment will have creatures with 2 limbs that flap through the water. They will only be able to move a minute amount on land, so will die before they find food on land. But I will make them able to see food near the shore, and evolve to be stronger, so hopefully I can see them swap from land to shore. I'm currently programming these evolvable qualities: eyesight/range, hunger, strength, and each terrain type will have a certain 'buoyancy' which makes it require more strength to move in (i.e water is easier than land, but there'll be more competition).

@alfalfamire

7 жыл бұрын

Post these things on your channel. I will definitely sub.

@micomator

7 жыл бұрын

Okay! Anything specific you'd like to see? Are you interested in the development, or just the results?

@keksimusmaximus444

7 жыл бұрын

I would love to see the development! I'm a newbie to programming, so it should be great to see how these "evolution" games work!

Make them not be able to go in the water unless they have the ability to.

this is really cool man, cant wait to get my hands on it

what is this game I've never heard of?!?!

@S_O_O_C

7 жыл бұрын

he is still developing it, and its more of a simulator than a game

@sciencechicken7669

7 жыл бұрын

Its hes own game he made it

@tommyproductions891

4 жыл бұрын

It’s not released irl

please don't make us wait a month!!!! Any chance you could update it and make a video in 2 weeks?

@quin2910

7 жыл бұрын

You are awesome. I really want to make my own simulator but I am basic at coding right now (javascript)

@dexterovski

7 жыл бұрын

If you know JavaScript, it will be a piece of cake to learn another language. C++ combined with Python is a very good choice of languages IMO.

@quin2910

7 жыл бұрын

Mykola Zekter thanks

@quin2910

7 жыл бұрын

Mykola Zekter do you know any good sites to learn off?

@dexterovski

7 жыл бұрын

Quin Proctor Python is pretty easy (at least basics), pick something from these maybe www.quora.com/What-is-the-best-online-resource-to-learn-Python . C++ is a bit harder. Here is a list of great books on C++ stackoverflow.com/questions/388242/the-definitive-c-book-guide-and-list . Good luck!

That's a really awesome project you've got here. You should make a game which involves a player that controls like a village of humans versus animals and evolution or something like that. Like it'd be completely different from your generic game, it'd be a game which has enemies evolve to kill you allowing there to be a range of different and interesting scenarios that the player has to react to.

Yaaay open source, thanks! Take your time, awesome video :D

Does anybody know how he does graphics?

@Phagocytosis

7 жыл бұрын

I do not, and would also like to know.

@genericnameman7899

7 жыл бұрын

i do not, but do not care to know

@Phagocytosis

7 жыл бұрын

Not the most useful comment, but fair enough.

@FyJonas

7 жыл бұрын

Neither is this one or yours

@Phagocytosis

7 жыл бұрын

Mine was useful to myself, because I was interested to get notifications of responses relevant to the question posed by OP.

I immediately recognized your voice as familiar. When I saw you were one of the guys who voice acted for BFDI, I honestly wasn't surprised.

You are a genius. You programs/simulations have so much more potential than you are giving them credit for. I'm surprised some government program hasn't snatched you up yet.

Great job! It´s amazing!

Its amazing to see what obs looked like 5 years ago. It hurts my eyes.

This could become a sit and watch your world evolve with the creatures creating wood structures and beginnings wars and then upgrading their equipment

carykh you NEED to put sightlines (the cross of vision) directly sitting in front of creatures and directly behind the creature to improve the hunting system Example: Imagine that there is a creature that hunts other creature, in a chase the creature hunting needs one way to see in front of her, without it is completely inefficient watching the video I realized that no hunter creature appeared because it was completely inefficient and the creature that is hunted needs see behind her, like when you hear that there something behind you, to balance the predator-prey sistem I sorry my English is that I'm using google translator because my native language is Spanish P.S. greetings from Venezuela P.P.S. hola a todo el que me entienda

@sebas200425

7 жыл бұрын

leave a like please

@sebas200425

7 жыл бұрын

you can comment your opinion about my idea and you can comment here your ideas on how to improve mine

This is soo interesting! Thank you for sharing!!

Great Work you seem cool keep up the good work! :)

Totally love it! Please don't take another month to release the next one T.T

I found your channel last night and my cat woke me up at 3 in the morning again, so here I am. You're content is great, but you're so hard on yourself. An example is at 0:14. Relax. You're awesome.

Here's some random suggestions Cary! This would create an environment that would change and be non-static. First, you could give the tiles a height value, and then add a sea level that gradually raises and lowers over thousands of years. Second, building on top of that you could add anti-subduction zones, which yearly send out subduction waves, which start by slightly raising land, but gradually shift to lowering the land, and when they hit other opposing subduction waves they either merge or terminate, thereby creating subduction zones. To prevent a fixed final land mass, have the anti-subduction tiles shift every few centuries and have the rate of change the subduction waves have from raising and lowering on their journeys change randomly over hundreds of years. This would create a fluid environment to the game, mimicking reality. Thanks for reading all of this!

Love your programs!

I'd love to increase the efficiency of your sim once you release the source code, can't wait!

multicolored mouths would be interesting, so either an animal can focus on one kind of food and be very efficient at digesting it (cows) or spread out on 3 or 4 kinds of food but make it cost some more enrgy to maintain, or be a bit less efficient. keep it up :)

So excited! Keep it up!

16:28 Ndente has learnt to avoid water, amazing work

This is awesome! I hade such idea of doing evolution simulation for long ago, but never actually got to it. Now you inspired me for trying the same! I kind of think, the only way to get Aritificial intelegence, is kind of "grow" it inside the computer, so it evloves by itslelf. Cool, right? :D

I think mouth colour should be an input which is only changed randomly through reproduction. This would mean creatures couldn't just change their mouth colour to be the colour of the tile below them, and it would mean you would get different species occupying different ecological niches, by eating different foods.

I like the evolution videos that you have

LMAO I found that little scarce Easter egg when he looked at the voice acting segment. +1SUB

Might I recommend dramatically increasing the energy cost of moving? That makes using one's vision to analyze the environment, and respond appropriately to it more important. Additionally, I would suggest adding an energy penalty to changing one's mouth/body color, to encourage specialization of food sources, as well as preventing perhaps an easy escape from predation (change body color to no longer match the mouth color of a predator). Just some thoughts. I love these evolution videos.

11:35 "So I just restarted the simulation and then I ran it for 27 years." -Carykh, 2016 Thanks for your dedication, lol.

Once you add the optimizing feature that basically pauses calculations on tiles not being used (if you do add that), you can add a button that updates all tiles so you can see what color they currently are and their stats

Oh yes, please release the source code! I'm enjoying playing with the seed value of your evolution simulator. I'm also trying to get my head around the code, so I could maybe add some stuff. Like maybe ramping up mutations when a certain species dominates more than 90% of the population.

Very cool video! When you release the source code, can you make a video about the code? I think it would make it easier for many of us to understand the code, so the community can give you inputs too. This is so an awesome project!

You should let he creatures vision angle evolve, as well as the length of vision (focal distance), this would likely give very interesting results... Subbed! Also, is it possible for the creatures to decide whether to mate with another creature based on the other creature's colour? If so you might see this trait evolve as would likely be beneficial.

blue grass? cool...its better than getting stoned

This is pretty awesome :P

Oh, also, is there any way you could show us a much longer run at some point? 1000-10000 generations or something? It's already cool as is, but I'm curious to see if more intricate things evolve on the longer term. Never mind, you said that at the end of the video. Sorry, I got too excited and commented early!

men you are a genius, how the hek you made all these things, i want to tell you that you are a really **** good programer(sorry for my english)

Really nice! As a ML/DL developer I have to say though that its pretty infeasible to expect some 'cool results' based on this environment and a normal CPU/GPU. The environment and the problem the creatures (agents) have to solve have to be simplified a lot to actually achieve something like evolution of species.

This is so awesome, screw it i'm learning programming just because of this.

Gosh how I love that intro

Finally, a video! I just wished there was a download...

That is really interesting like how different animals have different diets

new sub. love your stuff

I would love predators to appear in this. Interesting to see what this simulation could hold.

You should probably code for decomposition. so when a creature dies the food begins to slowly shift towards the climate and yums grows to simulate new plant growth from the corpse of the dead thing

Predators and herbivores? This is so awesome. Good job man!

@MsHojat

7 жыл бұрын

Except that can't really occur right now because of the fact that the creatures stupidly plop down food of their creature color, which is just a random arbitrary color (hence the rainbows). Although what could happen is creatures who have the same colored body as the food they eat, so they just munch down on both their auto-generated food that matches their color, as well as any food dropped by their own kind. That said, it's a nonsensical type of omnivorous that is only congruent with the arbitrary color they've developed. What would make more sense is to have all creatures drop the same color of food regardless of what their color is. That would represent meat.

Wow! Managed to get to this back in the olde days when there were less than 1000 views