I Simulated Evolution: Something Happened…
ฝัง
- เผยแพร่เมื่อ 29 ธ.ค. 2024
- --- Links ---
TH-cam Membership: / @eightlittlebears
Discord: / discord
Thanks so much for watching!
I created artificial life along with a procedurally generated environment to simulate evolution! The ultimate goal is to turn this into an evolution simulator game, but one step at a time… let's just do our best to (try to) learn about biology and ecology!
How does natural selection lead to evolution?
How does predation evolve?
Can we find homeostasis?
In this simulation, we explore the relationship between herbivores, omnivores, and carnivores. We also allow mutations in size, speed, rate of reproduction, and diet through natural selection.
Thanks to all the great simulation videos on TH-cam which inspired this series! Particularly:
Sebastian Lague - / @sebastianlague
Primer - / @primerblobs
ThinMatrix - / @thinmatrix
The Bibites: Digital Life - / @thebibitesdigitallife
The Sapling - / @thesapling
Music by:
geoffharvey from Pixabay - pixabay.com/us...
Nullhertz from Pixabay - pixabay.com/us...
#evolution #simulator #nature
For anyone seeing this for a second time, I realised I uploaded the wrong version which had some washed out colours, poor resolution, and odd sound! This one is better :)
Algorithm may make this video gain alot of views
@@humanhuman7280 hope so!
hey what you have shown is basicly population dynamic. The emergence of predators is an option and you can see that predator and prey are adapting (thats kind of evolution) but in general you have basicly one species of different traits living all in one environment since they all travel through all bioms and basicly having the trait of canabilising each other or not.
all other traits are within the range of theire population. Or can they only mate with an individual of matching traits. if they just divide to reproduce its clear that you have such dynamic, since each one is a lone ranger reproducing like crazy with the right traits killing all the others until the perfect apex predator will have to decay or will go extinct anyways.
thats why i mean you show population dynamic and predator prey interaction.
@@liketobe2743 so they are different species (even though many of them look the same). Each creature essentially has a crude “genome” (emphasis on crude), and they can only make with a creature that is (a) 95%+ match for the genome, (b) of the opposite sex, (c) at reproductive maturing and also seeking a mate.
@@EightLittleBears cool!
Ironically, giving prey more ways to defend themselves would’ve limited the growth (in size)of the predators making the point where they’re to big to sustain themselves taking much longer which is ironically beneficial to predators
Ha - interesting point! That actually seems to happen in my predation video (although there is no meteoric rise of the predators in that one).
@@EightLittleBears so, basically, you showed the predators the valley of uber-optimized killing machines? you took a ball out of the hill and into the valley?
*too
The thumbnail looks like an peta game :v
Thats how nature works Herbivores and Carnivores keep each other in check so none of them overpopulates and die out due to not enough food
Hey ! I'm doing a PhD in evolutionnary ecology and this is a really cool video ! What you observed at the end of the video with carnivores (predators) replacing herbivores (preys) is very common in population dynamics, although not as drastic as what happened in your case. The Lotka-Volterra equations are a pretty famous way to explain and model what happens, I recommand you to check it out and what's been done around it !
Amazing! Thanks for the advice!
hey i do not have a phd! but i think thats cool.
What’s a real life example of the predator/prey replacement dynamic?
@@cloverring The textbook example is the data around records of hare and lynx furs collected by hunters from 1845 to 1935. By looking at these records, researchers were able to link the number of furs collected to the abundance of the specie over decades. In years where lots of lynx where accounted for, the number of hares was very low, and vice versa when lynx counts where low ! This results in a cycle where lynx population grows as they eat more hares, then decline because hare population get low, which allows hares to thrive again, and so on.
Do you think the severity of the shift might be to do with the size of the map? If it was four times bigger, do you think we'd see less sharp changes as the carnivores/omnivores hunt what's near them, rather than cover the whole map? Putting this question out there in case 8LB would consider observing the stats over a larger map... if that doesn't cook their machine. :D
Is there a mechanic to this that let's the environment slowly change? like temperature making more ice and desert climate, that would be interesting to see how things adapt.
I am in the process of adding evolving biomes (with decomposers, insect populations, various plant types, soil nutrition, etc.). Hopefully it will be a real “circle of life” update, but getting it right has been a bit trickier than expected, so I am not sure how long it will be before that update is complete.
Temperature changes due to seasonality or something might be a good bridge though - thanks for the thought!
k@@EightLittleBears
One interesting feature I think you could add would be to add a variable for how "Migratory" a species is. Basically, a chance for it to move, even though it has available resources. In a larger map, this could potenially give prey some defense against predators, at the cost of potentially moving to a location that's worse off resource wise.
Thought this video came from some of those well known TH-cam channels about evolution and coding. But no, just a currently underrated TH-cam channel which so happens to have an interesting well put video. This video really needs more attention.
❤️
This is pretty cool, and it makes sense when you view it as the very beginning of evolution. Like a pendulum, it swings between states based on what is advantageous at the time. This changed the way I viewed evolution. Pretty cool stuff!
Thanks! And yeah I thought the same thing!
Holy shit... evolution is just esport strats/metas over big time...
@@judah5037 I mean, esport strats/metas literally evolve, so yea?
@@judah5037 If you don't watch TierZoo already, I highly recommend it for game terminology applied to real world animals
Your mutations are really well done.
I might recommend focusing on the four F’s :
(Fight, Flight, the other 2 you have). Size and speed models don’t really capture all of the physical traits anyhow (camouflage, intimidation, communication, … there’s so much)
I like how you chose hexagons instead of squares (it’s hard math, but more fun)
Your theories about Epoch 2 sound correct too. There isn’t really an advantage to being a carnivore - over time the population would converge to 0 (it might also be because carnivores might be cannibalistic - consuming other essential carnivores - causing the extinction)
Thanks for the comment - "flee" is in my predation update video, so I am getting there! "fight" is also there, but it is a bit artificial because there is no "injured" status for either the predator or prey, but I will add that eventually :D
The effect you've got when you got a lot of herbivores, and then they were replaced by carnivores and vice versa is probably the same as "population waves", it's a pretty easy, but interesting concept. So yeah, they probably would repeatedly replace each other, just like it happens with hunters and prey in real life!
Pretty cool insight! I hadn’t heard that terminology before but it made for some interesting reading!
@@EightLittleBears yay, cool! I wasn't sure if it was called the same term in english, because it's my third language, but i'm glad it was interesting 🌺!!!
@@fantomdyablo9325 I think you nailed it!
this is video quality i dont see in channels with like 50k subs this is factually accurate informative entertaining statistical give this man some compensation for his unthinkable work
Thank you so much!
0:11 because hexagons, are the bestagons
"bees are smart" me 10 seconds later (not a joke) watch a bee crash into my window and die.
5:22 "[...] an untapped population of defenseless walking meatbags."
HK-47: "As a meatbag would say: 'I have a bad feeling about this.'"
4:12 Correct me if I'm wrong, but based off of what I've learned in biology, energy that's passed up the food chain gets less and less, making it so that carnivores, especially ones higher on the food chain, have to eat more.
My understanding is that you are right about energy being passed up the food chain. Calories are “lost” (actually just passed back into the environment,etc. since energy is never lost) through waste or the development of inedible body parts (bones, teeth, horns, hair,etc.).
It is also true that, if a tiger kills a deer that is 5 years old, the calories the tiger can obtain are limited by a single moment in time which will be a tiny fraction of the overall calories that deer has spent growing, walking around and fighting over a 5 year period.
This doesn’t necessarily translate to carnivores needing more calories on an individual level; it just means that more calories need to be spent in an overall environment to support a population of carnivores (I.e. a single carnivore needs to consume many prey animals over its life).
That said, 1kg of meat has far more calories than 1kg of grass, so carnivores can afford to eat less often even where they require more calories than a herbivore (which is goal of the mechanic mentioned at 4:12).
It’s an interesting point though! I’ll have a think about how the energy lifecycle can be more accurately represented in our environment!
@@EightLittleBears Yep, this is what this comment is for!
Also, good luck in making an improved simulation!
You have to consider just how much the prey animal has eaten to get the meat it has. One morsel of meat is equivalent to many MANY times its weight in plants in energy simply because that's how much plants the herbivore had to eat to get that meat in the first place.
Meat is extremely densely packed full of energy, and even a single meaty meal can last you, a human, a whole month without any other food. Plants will have you starving pretty quickly unless you're constantly scarfing them down.
Some exceptions of course, like extremely hearty fruits and especially beans and nuts, but even those are still nowhere near as potent of foods as a good slab of meat is.
youre kinda right
the assumption is usually that only 10% of the energy is passed up. so a cow gets 10% of the energy in grass while a lion that ate the cow will only get 10% of the cows energy
but, a cow stores a lot more energy than grass. so a cow has to eat grass constantly throughout the day while a lion may only need to eat cow a day.
calorie/energy wise, a lion eats more. but when we ignore energy amnd focus on quantity of things they eat, cow eats more
@@Zack-fu4lo Okay, that makes sense. Thanks for the explanation.
4:32 Another channel working with evolutionary simulations, the Bibites channel, had a similar phenomenon where it was much more advantageous to be a herbivore than a carnivore. He eventually figured out that the key block he was missing was digestion. Since plant matter in real life isn’t as energy dense as meat, there’s less profit in actually processing it and it takes longer. Maybe that’s what is happening here as well? It would be super interesting if that was the same case here!
It definitely makes a difference. Also I love the bibites!! ❤️
Thanks for the comment!
2:16 “and there are no islands”
*me seeing the 1 tile at the south end and some tiles at the north*
The bears: "mmmmm plants"
Also the bears: "hey uhhh what if we ate each other? "
Bears one more time: "nahhhh plants too good"
Haha silly bears…
Very entertaining video! Good luck on your future TH-cam endeavors.
Thanks!
Very interesting! I am intrigued to see your simulator and channel evolve. This is also some superb video quality.
Thanks so much!
This is amazing. Would love to see some new variables like:
- age of maturity (could be earlier if enough food and later if they're struggling to build energy)
- reproductive age (could lose virility over time)
- lifespan (obvious)
- socialising (how willing/eager to mate)
Can't wait to see where you take it
Thanks for the comment! ideas 2 and 3 are already in there but in very simple terms.
@@EightLittleBears can't wait yo see even more
Very well made video and simulation enivronment creation! I'm doing my Bachelors Honours Project in Multi Agent Reinforcement Learning and its alway interesting to see others environments to get inspired especially how user friendly it looks with unity
subscribed great work, you can see the quality of the video took a long time.
Thank you!
This is so awesome! So exciting and inspiring to see people just out here doing awesome things just for so :D
Thanks so much!
I just randomly got reccomended this channel, but i love your handle.
Just some bears in a trenchcoat!
This is a great video! Keep up the amazing work!
When you add more animals, I'd buy this as a game. Heck, I'd love it more if those animals were dinosaurs
Dinos (and other pre-historic animals) will come eventually! 😃
give the predators the ability to understand to not mess with much bigger creatures, the ability of fleeing and escape for smaller creatures and the ability to choose help others for herbivores (only if they re big enough and rarely anyway)
Check out my predation video! Some of this is in there :D
Amazing work! I thought that omnivores would thrive due to being able to get food more easily, I was right for a bit lol
Ha - I thought the same when I first made it actually!
My problem with experiments like this is that it's so limited and closed. You have a preset list of statistics that definite the creatures and they can only mutate within those preset parameters.
There was an experiment in the early 80s by Thomas Ray that really showed what digital evolution could do. There was a Virtual Machine called Tierra. Programs would run inside it and Tierra had two important points, when a program was running, every so often a bit would get flipped, 1 to 0 or 0 to 1. The other was that it would track how long each program was running and delete it when reached a certain age.
Ray set one program running in it called 80 (so called because it was 80 bytes long) which simply copied itself into a new memory location. Most of the copies that mutated simply stopped functioning, and occasionally he would get a 79 or 81. But eventually he got a 45! On examining a 45 he found that what it was doing was using the code of an 80 to reproduce itself, like a parasite. So 45s and 80s needed to be in balance, but apparently 79s were immune and could breed without that limitation. Eventually a 22 showed up that was not a parasite. A functional, self contained program that was almost a quarter as small as Ray was able to make (smaller programs take less time to run so can reproduce faster).
Eventually other types of programs evolved like groups that passed the pointer around to each other do they would all have more opportunities to copy themselves. Parasites evolved in these groups that once they received the pointer, didn't pass it back.
He tried turning off the mutation and he STILL got evolution, from two programs trying to copy themselves into the same location at the same time.
So many traits and tactics evolved that Ray never could have imagined and didn't build into the program at the outset.
The one disadvantage is that it doesn't give you a landscape-like graphic that is immediately understandable by the average layman.
Thomas Ray’s work is pretty fascinating (though I am only vaguely familiar with it), but I see two differences when compared to my work:
1) he is seeking to explore the limits of evolution, whereas I am seeking to bring the outside world into a playable environment; and
2) I am not as smart as Thomas Ray 😂.
Thanks for the comment though - very interesting stuff!
My first time watching your channel and I already love it my dude btw why is your profile pic a frog when your channel name eightlittlebears
@@GrayscaleMC-y4f glad you liked it! And honestly the frog is just an old picture I had that fit well into the display pic window haha
It would be interesting to see species evolve into aquatic or semi aquatic species if resources were limited on land
Definitely on the roadmap! The main issue is that the Whittaker biome system doesn’t work with aquatic biomes, so I would need to rework that system. Not impossible obvs. Just time consuming!
Maybe add some form of social communication among herbivores, carnivores and omnivores respectively and see how that changes things? Like having some carnivores move in packs while others are loners and the like?
Thanks for the comment - this is something I am going to implement at some point. But I think I need to implement some form of intelligence first so they can develop teamwork over time!
Thanks for the comment!☺️
It's so cool that the habitat affects the evolution greatly mostly at first the predator evolved in harsher environment and less green but after moving to the green tiles they evolve into omnivores and slowly to herbivores us the population grow and the competition of food greatly expanded they evolve to omnivores and it repeated again , sadly you can't see more in the carnivore because of the environment they don't really need to evolve on that path.
Yeah! I’ll do lots more with different environments over time so hopefully we’ll see some variation!
Man this is cool good job man
Thanks!
This is sick dude the production quality is insane
Thanks!
This pattern remind me of the role lemmings play in their ecosystem. Several predators rely almost exclusively on lemmings to breed and rear offing during the spring and summer months. Lemming populations will grow exponentially until they crash, only for the population to recover and the cycle to recover. When the lemming population has crashed, predators such as the Arctic Fox will often struggle to find enough food to rear their offspring that year and few if any young will survive from that season.
The one thing I'm noticing and tilting my head at is how the vegetation in the plentiful biomes remains unchanged over the different generations? Wouldn't there be some areas that are over-consumed, thus diminishing the amount of plant life available in certain biomes? I would think that would create a more reasonable place for carnivores/omnivores. Since with less plants to consume, the mutation of consuming meat would be more necessary.
Fascinating all the same.
This is something I want to build in eventually. At the moment there js just a maximum amount of plant life in each biome (with ratios loosely based on biomass in real biomes on Earth). However my next update is adding the concept of soil quality/nutrients, waste, decomposers, and various plant types. So it is working towards the concepts you are talking about (though it won’t be all the way there yet).
@@EightLittleBears Oh absolutely. You are doing it in a much more advanced way than I initially envisioned. Keep up the good work!
I feel like a point to make possibly is that it seems strange a herbivore and or carnivore just switching diets completely, in the sense that if you had cow which is strictly a herbivore and gave it meat as its only food supply it wouldnt just adapt to the new food source, it would die due to not being able to digest it.
So that’s a good point. The creatures don’t just switch over. What happens is that herbivore offspring can mutate to be a bit more omnivorous along a sliding scale. If this keeps happening for several generations then they will eventually start to prefer meat and then after that they might slide all the way over to carnivore. Most of the time what happens is they hover somewhere in omnivore and then slide back to herbivore.
A theory I have for epoch 2 is it looks like majority of the terrain supports plant food meaning that by rule of natural selection herbivores would dominate as shown and maybe I missed something but If carnivores and herbivores can't mate in the simulation it would lead to a lack of ability to reproduce for the carnivores due to limited numbers spread too thin and lead to them eventually dying off.
Thanks for the insightful comment! Yeah it is definitely tougher to be a carnivore, though one thing to bear in mind is that the herbivores are also split into different lineages (even if they sometimes look the same due to the limited number of avatars), so the herbivores can all mate with each other.
This is amazing but in a future video maybe make so the type of plant evolve and change to because just like animals plants change over time too like amount of food produce or grow time
On the todo list (and will be partially set up in the next release, which is focused on energy/nutrient cycling)!
bro your voice is so good, you could commentate grass growing and i'd still watch it
Thanks so much! This is my new favourite comment!
One thing that would interesting to track in future experiments would be the the herbivore population and speed that plant food respawns. Theoretically, the faster the plant food respawns, the less effective carnivores are by comparison. This video inspired me, I would love to try to do something similar
Definitely have a go!! You don’t need to invest a lot of time to get a basic simulator up and running, so it should give you an idea of it you like do it!
Also tracking plants is something I want to do at some point. The funny thing is that faster growth makes the world better for herbivores, as you suggest, but more herbivores makes the world better for carnivores… and more carnivores (I think) makes the world better for plants… the circle of life!
interesting to see that species:ALRE isnt the only evosim that experiences the herbivore bias phenomenon.
Basically something in evosims ive seen is it takes a VERY long time for carnivores or even stable ominvore populations to develope. Something in the simulation causes a bias towards herbivory. My theory is that its the population size. If you could simulate something more akin to a continent than an island in scale, then you might have carnivory pop up more easily. Most evosims opperate in population sizes of hundreds, not thousands or hundreds of thousands. This puts them in similar scale of ecological complexity to small islands, which might only have a few hundred observable animals. Stuff like the innumerable dots in the pacific. Just like those islands, you rarely see macrocarnivores, just insectivores or maybe piscavores. So maybe if you increased the scale to be bigger, youd have more complex systems emerge
I think you are definitely on the right track with this thought. I will be doing a much larger one at some point, but need to refactor my code to make it run more efficiently first… but I want to get all the core systems in place before doing the optimisation… will see I guess 😁
@@EightLittleBears from my own experimentation publicly available evosims, ive noticed that the bigger the sim both in environment size and population cap (if the sim has a performance cap), the more likely it is to see macrocarnivores.
But i also suspect, since these are a simplified system ignoring factors like micro habitat ecology, and use simplified analogues to things like how diet works (usually a spectrum between two values representing herbivore and carnivore, with occasional factors for specific preferences of fresh meat over carion or fruit over grass) that these simplified factors are resulting in potentially aberrant behavior.
To test that second hypothesis, youd likely need more complexity than one person can code, and one desktop computer can process, or youd need to experiment with real specimens, which is not feasible till we develope o'neal cylenders and temporal manipulation technology to allow us to run controlled environments for thousands or hundreds of thousands of years. Maybe one day we will have the tech to run such experiments, more likely we'll have the technology to simulate it before we get the technology to make it happen tho, as youd just need a more powerful computer and more complex simulation for that.
i think its because the real world doesn't have invisible walls, and prey develops ways to evade predators, like hiding in the water, air, trees, and soil. I played Species ALRE, and what often ended up happening is the species that bred fast enough to max out the population cap either thrived through making it so no new babies could be born, or that some herbivores would become omnivores and eat the dead herbivores opportunistically, but occasionally become hyper lethal and kill everything, creating a cycle where the parent only lived long enough to reproduce, and then the child would eat the mother, and thats probably because the scale of time is summarized. Any time true carnivores that hunted appeared, they would eventually die because the herbivores that lived were the big strong ones and would one shot the carnos.
Interesting - I have tried to avoid "invisible walls" where possible. The only one really is the maximum plant biomass that can be supported in each biome. That said, that wall is essentially an abstraction of the limits of physical space and solar energy that can be received, so it sort of is a "wall" in the real world as well!
Thanks for the comment!
Amazing video dude!
Thanks!
would be cool in my opinion if herbivores could kill or drive away each other when trying to compete for the same tile. Size should not guarantee win in a fight but it should give better odds for it. perhaps there should also be a stat that determines the size of potential competition that a creature would flee from.
The “bullying” mechanic is definitely something on my list. It is important for things like bears as well!
Re the size point - check out my predation video which completely changes this mechanic! ☺️
I love this! Evolution simulations are awesome!
I wish there's a game in which we can just watch our little creatures evolve and mutate... If anybody knows a game like that, lmk...
There is a game called "the sapling" on steam which allows you to design animals and plants on an array of planets, and watch them survive and evolve
Man who simulated evolution surprised that something evolved
What can I say, didn’t expect it to work 😂
I wonder if you let the simulation run for really long would we see lotka-volterra predator-prey cycles?
Yeah I think it probably would. I am planning to run some very long ones at some point when I get all the core systems to place I am happy with!
Thanks for the comment!
Fascinating and brilliant ❤ You succeeded where I failed so many times, Life simulations are a project that I worked on programming at least a few times and I’m super curious what program you created this in? Also, are you an art student? You definitely seem to have some really good knowledge of animation and programming.
I used Unity for the programming and Procreate for the art. Nope not an art student - just something I’ve had an interest in for a long time 😊.
Thanks for the comment!
I believe another food source should be added - high sugar/high carb plant matter like fruit, but is not easily available or all the time. Sugary fruits are one aspect of what enabled human brain growth to have its potential, sugar powers a big brain
thats crazy! isn't non-hardcoded predation almost impossible great work!
Great video, great simulation! If I understood it right, you have a set of fixed traits that can be present in greater or lesser amount which changes with mutations. I know that one important facet of organism phenotypes is that immune systems and protection against prey are extremely important. They evolve the most rapidly of any of the traits of any organism. And I don't think you had a trait or traits that represented the ability to defend against predation. I'm sure you're gonna be adding that and I'll be watching your subsequent videos.
I really love ecosystems and such. I even have two jars with small ecosystems in them at home, looking forward to see more
Tiny ecosystems are awesome!
Hey just an idea what if you code it so they can migrate into groups and have colonies, wars and international races like the Olympic Games and what if you made it so they can evolve to go to other planets and moons? Anyways great video
This seems pretty advanced!
@@EightLittleBears thanks for responding but i know mayby try to get chat gpt to code it😁
yes, if the simulation continues, the species would enter an oscillating pattern, where there are die-offs and resurgences, till they can find a cyclical balance.
It reminds me of a different life simulator that had a similar issue of predators being wiped out. The problem was that the prey evolved specifically to evade the predators until they died out or adapted into omnivorous scavengers. While that one didn't result in a later predator boom (probably due an initial amount of meat from the vast numbers of herbivores that died naturally) the solution to the problem there was to add a digestion mechanic that meant prey couldn't overly adapt without sabotaging their energy efficiency. If course it was a far different system, so you'll probably need a different solution.
Interesting- it wasn’t bibites by any chance was it?
@@EightLittleBears Indeed. With how complex their system is its a shame they don't do more videos showing how they adapt in different situations.
You’re not just some guy on the internet you’re the guy on the internet who’s video the viewer clicked on lending you credence
Hello! I really love your videos I would like to run some simulations for my colleagues in the university. How do you create this simulations, what programming language and environment do you use?? Thanks!!
Thanks for the comment! It’s awesome to hear there are more simulations waiting to be made! I use the Unity engine and c#.
It depends on how you programmed their diets to evolve over time. If your "Carnivore/Herbivore" is a single gene or slider and the animal essentially has to choose between herbivore and carnivore, it will always choose herbivore.
The solution to this, if this is the issue, is to have meat-eating and plant-eating be two separate sliders, from 0-100 or however you do it. This way, an animal can evolve INTO carnivory without first having to lose its herbivory. This transition between herbivory to carnivory could be what is causing predators to be unable to evolve, because if they have to first drop their extremely successful herbivory just to get carnivory, they would never do that. If you allow them to evolve carnivory BECAUSE of their success by having it be a separate gene, it should happen more frequently.
Realistically speaking, most animals on this planet are some level of omnivore. The only true carnivores are either cats, or basically anything arctic/oceanic due to the zero plants thing. Most other animals are omnivores, with some level of preference.
Plus it's important to note just HOW MUCH BETTER meat as a diet is over plants. Meat does so much more for the animals that can digest it than plants do for the animals that can digest those, including healthier bodies, increased healing rate, more dense muscle mass, so SO much more energy per bite of food, carnivores are also generally able to be smarter due to the higher energy intake, and in general can make an animal that eats meat last so much longer lifespan wise. Plus they don't have to eat very often as you said. Gators can go like a year without food. We humans are only omnivores and we can go like a full month without food after having eaten a hearty meat meal.
And finally, the big one that requires entire simulation changes: Certain types of creatures will just naturally need a specifically sized plot of land to be able to work as a population, otherwise the "population fluctuations" mentioned elsewhere will end up destroying themselves due to land not being expansive enough, and your simulation zone is pretty physically small.
Very insightful. A couple of points you might find interesting:
- in this sim there is a single slider for diet, but it ranges from -99 to 99 with 0 be a perfect herbivore (I.e. it can obtain 50% of the nutrients from either plants or meat) - I am currently reworking this for the next video but too long to explain here 😁.
- the sun also takes into account (to some extent) the point you make about meat being higher-reward. A single animal provides more energy than some of the biome tiles (though it decays rather than grows over time). Diet also scales with how often a creature actually needs to eat, so a carnivore will get more out of 100units of meat than an omnivore will, and both will get significantly more than a herbivore would get out of 100 units of plant food.
One thing to consider is that a consistently low population of obligate carnivores is not necessarily a sign of failure I think. Like you say, there are very few terrestrial obligate carnivores in nature (because it’s hard and you need a much bigger population of prey to sustain it). Basically what I mean is don’t think of it being carnivores vs herbivores with the higher number winning - think of it like “does this reflect a type of balanced ecosystem and how does that compare to the real world?”.
Re: your final point. I designed the simulation with “territory” in mind for creatures, so this is something I can add without completely rewriting the core logic 😁. It’s a great point and definitely on my todo list!
i wonder how a much bigger map would effect it?
i thought you had 3.6 mil subs XD you should have more like 360k atleast
Ha thanks! My channel is pretty new so I’m grateful for all the support!
I would suggest creatures to move before tiles run out of food completely. It's unrealistic they would scour it that thoroughly, and a completely depleted tile is a strong barrier,v so this has the potential to change the simulation dramatically.
How about a probabilistic motion where the more food is available, the more likely the creature is to stay?
@EightLittleBears -- Does the simulation account for creatures with a genetic predisposition to grow large starting out their life small? If not, you might want to add that factor, perhaps along with some sort of growth speed Factor, which of course would also affect how much food they would likely require, and be affected by how much food they actually eat. Also, if some species can begin reproducing before they've necessarily reached maximum size, then being a predator that can grow extremely large wouldn't necessarily mean inevitable Extinction.
It's an interesting point! At the moment the starting size is just a percentage of the max size which scales up or down based on reproduction rate (e.g. fast reproduction results in more immature/smaller offspring that growth faster to reproductive maturity). It's a system I want to enhance as it is pretty basic.
Yeah, looks like with ample plant food sources there is no strong pressure to flip to a carnivorous diet for most of the run.
Eventually the herbivorous population hit a limit of the food supply and that provided conditions that favored predation.
With the simulation limited to size as defensive/offensive trait, there was a very strong pressure for everything to get big, but herbivorous hit a dietary wall from being to large driving the population crash and 'reset' to a stable state.
It's very impressive the simulation population didn't go into a terminal decline when it crashed.
Thanks for the comment! I think you nailed it!
This is cool ive seen some evolution videos but this is the coolest!
❤️
Really interesting. The visuals help a lot. I'm excited to see more of this.
Thanks!
I wish this was actually a game I learned a lot for AP bio
One day it will be!
i would be very interested to see what is different if the map size is massive like 3000+ tiles and an almost equal part of every biome and liike 2 or 3 pawns in each biome to start
I would LOVE to see a world populated only with rabbits wearing vibrant mustaches. Subscribed!
awesome video, immediate subscribe!! i'm going for a degree in computer science and a minor in biology exactly for this reason, i want to simulate stuff like this too!
Thanks, and wow! With those credentials you are gonna be blowing me out of the water!
@@EightLittleBears haha, thing is though, you have a skill that's a lot more valuable: perseverance. i've been "meaning to" get around to all sorts of projects, but i still haven't really made any significant progress, whereas you're actually doing the thing. that's one of the reasons i admire your work so much, because it's more than just an idea for a thing, it's a real thing that you can put out into the world and show people! so cool.
They changed is so that any youtuber can do polls now! I hope this helps in some way.
That fast moving model is unBEARable to look at XD
TH-cam has shown my video to over half a million people (though not all clicks of course), and this is the first bear pun comment I’ve had, so not even mad that it’s mean! 😂
@@EightLittleBearsCouldn't come up with a better pun, I'll try harder next time
Whoa, I love evolution simulations, made one myself but it was bad 😂. Yours extremly good though :3
Thanks 😅
highly enjoyable and interesting
thank you
Glad you liked it!
Great presentation and topic. You've earned yourself a subscriber!
Thanks!
I think herbivores dominated because the carnivores had no way of sustaining themselves at a larger size on a sparser diet of meat, whereas the herbivores could grow larger and remain in lush areas where food is plentiful.
Had the carnivore had the ability to learn cooperation or develop more means of capturing prey, perhaps they would have been more successful.
I’d love to see the same simulation but with plant/meat based diseases that decrease lifespan allowing for a possible fast evolving but non-immune, or slow lifespan, but immune.
Interesting ideas!
This type of channel makes me instantly subscribe
Thanks so much!
Okay sooo more Video for me?
yeah, i know im late to the party but i was curiouse if you concidered that it may have been that the hurbavores may have depleated the food supply? just curiouse of your thoughts
They can’t deplete it completely because plants regrow, but individual tiles can be taken to 0 and then they take a few turns to start regrowing (and quite a few more turns to fully regrow).
So it certainly is possible that the herbivores took the food to a breaking point which led to some significant amount of deaths, and then this could have had some knock on effects (e.g. making it harder for creatures to find suitable mates).
Ideas:
terrain speed effects
shorten life spans 75%
Increase map size
Reduce initial population by 25%
Thanks!
Awesome Creation and Video, I’d say your qualified enough by the looks of it 👌
I am interested in the coding part of it is there a git repository you have shared, and which algorithm you have used like NEAT module, or any reinforcement learning techniques?
wow this is nice and motivative, what was the environment made with? python or a separate tool?
Make the plants carnivorous. To better state this fact, make the plants themselves have species and have them develop mutations that make it harder for the animals to adapt.
I’ll do my best! ❤️
@@EightLittleBears Ballin'. Have some fun!
Neat
🤩
Neat
Almost 900 subs 🎉
🥳
Awesome video! The algorithm spat something good at me for once 😊
Thank you!
Quite an interesting simulation but it is limited by the size of the simulation and the detail.
A larger simulation would likely help prevent a complete predator extinction as well as turning down the rate of mutation.
The other thing would be changing the success rate of predation, predators in real life have shockingly low success in hunting which helps prevent them driving their food to complete extinction, which ironically helps them stop driving themselves into extinction.
The main thing I noticed was body size, which normally has drawbacks depending on the ambient temperature of their environment.
Yeah I’ve done a couple of these things for later videos. Thanks for the comment!
really cool it kinda looks like niche
is there the possibility of downloading this simulation somewhere?
Edit: Just took a little look, seems like we'll have to wait, I'll definitely stick around
Yeah - I want to put a version in Itch.io at some point for people just to mess around with, but at the moment there is no UI, no gameplay, and I am not sure the reporting would work in a build version. Also, I am making the changes in code at the moment, so you couldn't even add your own creatures or anything. I'll get there though :D
This is amazing thank you so much also it’s really cool listening to your commentary and I hope you keep it up - sincerely teen aspiring to become a zoologist/ecologist
Thank you so much!
Okay first of all. That map is amazing. Bro it satisfies my brian.
This could be a game man.
Haha thanks! The goal is for it to be a game eventually. Just want to get the simulation to be pretty solid first and then I think making it playable should be relatively easy.
Maybe adding a variable where vegetation density also changes
I think you’ll like my next video!
Great video
Thanks!
Bigger map could lead to more sustainable carnivores population. Some pseudo isolated areas. Possibilities to cross water tiles for enormous cost. Different energy burning rate dependent on biome etc.
what program are you using? it looks really good
Thanks! Unity an c# for the code, Procreate for the art
What did you use to program this masterpiece?
The main tool I use is Unity, but I draw all the art in Procreate.