@@TheBibitesDigitalLife Damn, this must be really blowing up right now if even giants like kurzgesagt are getting this recommended just at the same time as me. Awesome work btw, can't wait for more Bibites in my life!
Since you're already making the predator-prey relationship a continuous gene pool, why not make the prey-food one as well? In other words, the food should be Bibites too, where the photosynthetic gene greatly limits mobility. These "plants" could then grow, divide, and adapt to the herbivores (since the plants don't want to be eaten either). I feel like this would unify things and make the life simulation more realistic and fair. And since the prey need to adapt to both sides of a changing ecosystem, it might balance predators since they get a slight advantage; they only need to adapt to one other species, not two.
Yeah, It's part of the roadmap! It's just that there's a lot of other systems that need to be addressed before it would make sense to let plants evolve too!
@@TheBibitesDigitalLife Maybe your simulation could need another class/layer of objects, like rock or different kind of obstacles, where animals need a certain set of "limbs" to get over that more efficiently and certain plants might grow faster on or around them, because of higher concentration of nutrients. Further more you could add patches of "nebulous" stuff or "inorganic" matter, like particles or dust clouds which contain nutrients, that could harden to "soil" patches of different quality. You could add something like "fungi" or "microbes", where certain nutrient particles are gathered and are being concentrated into those "soil"-patches for plants to grow. Evolving different "root"-systems might change intake of different nutrients for the evolving plants. Adding a path for meat and poo to be consumed by destruents to set those "inorganic" particles free for the fungi to use to congregate it to those patches for certain plants then having an easier time getting by fast, if they have evolved the roots for it, should do the trick maybe. That could be enough to establish a lot of dynamic niches, and hopefully some "stable" ones. If the Map/World is big enough, and the way how the nutrients are transforming is broad enough for many creatures to step into the chain, you will end up with different biomes, depenend on number of parameters.
@@Chareidos This is an interesting point. I think the simplest interpretation could help a lot. At the moment plants grow anywhere and everywhere randomly. but what if the plants only grew in certain spots? or if they also needed to drink water? then the preditors might be able to catch the herbivores on their way to food source towards water or vice versa.
Something that would also help is make rotten meat spawn bigger plants. Maybe make it so only weak plants (moss) spawn naturally, medium planet (grass) spawn near poop, and big/richer plants (bushes) spawn near rotten meat. This gives a path to Predation by making it advantages for even herbivores to kill.
Also right now the only thing to evolve for is how well you eat, and survive to reproduce. There needs to be environments so that there are more factors in play. Herbs need different environments that they need to adapt to and different types of plants to eat. There should also be things that block the view of bibites so preds can ambush.
Yes, right now it seems the herbivores have an infinite source of energy, which isn't the case for carnivores... so course there is no way they can catch up. Also, perhaps more cluttered Big sources of plants would help? Herbivores would gain more trying to eat there than they would lose to ambush predators camping around (could even be a behavioural trait-> hang around clumps of plants).
This was incredible, seeing the herbivores delevop into herds really took me by surprise. You clearly already have a great system in place to allow such detailed interactions between the same species, I'm looking forward to seeing future videos!
@@TheBibitesDigitalLife but shouldn't the carnivores have an incentive to become smaller and work in groups too? Like wolves chasing down bigger prey in groups
Honestly I think a really important missing part of this ecosystem is the capacity for the plants to evolve, even something small like adding a gene to them that controls color could help, because it would make it harder so become hyper specialized
predators very often target the weakest members of the herds. e.g. wolves and lions cause a mass panic to seek out sick and old individuals. or they target a young ones of that year. maybe you could build in life phases in the herbivors; making them dramatically more susceptible to predation at the beginning and end of their lives. you could make them slower or decrease their eye sight. very cool project btw.
Aging would be absolutely brilliant! Another feature could be to make bibits dying of age drop less meat than if they are killed by another bibit as that would encourage carnivores to actually go out of their way to kill rather than to just settle for scavenging!
Exactly what I was thinking. In addition to this I think reproduction should be factored in as well. An organism attempting to reproduce becomes slower and 'meatier' as it stores up nutrients. Perhaps allowing this various parameters to evolve -> weaker daughter cells but less energy intensive reproduction (or the opposite ofc), number of cells produced at a time (varying energy like the previous one), and whether they attempt to 'protect' offspring.
Weakness is a huge factor for predators. Hunting takes specialization and a lot of effort but scavengers make the most of whatever is available. Apex predators are often scavengers as well so they aren't limited to one food source. You need animals that can eat both some plants and meat since not all food is equal and scarcity is a major factor.
I managed to get carnivory to evolve in my simulation, and as far as I can tell, the key was turning off the negative nutrition value for incompatible food, in this case plants. I think I set the minimum at +10%. I think if you allowed an input so that they know what is in their mouth, and can more easily reject things which are harmful, it would help a LOT with carnivory. Can't tell you how many times I saw my carnivores die to randomly wandering into a plant glob, which tanked the rest of their energy, effectively poisoning them to death. It's quite unrealistic as is, because just because a cheetah ends up with a mouth full of grass on accident, doesn't mean they're gonna devour it all.
I didn't realize that! this would mean that apart from digestion, an evolution of "taste" is important for creating a pathway into different eating niches, which in retrospective makes a lot of sense!
@@makelgrax Ya, it's even possible, if not likely that taste leads, and digestion follows. Because say you have a random mutation that allows you to better digest a new food source, but it tastes like hell to you, the mutation is at best going to remain dormant until taste comes around, if not disappear before being usefully expressed. But on the other hand, so long as eating something isn't too harmful, if it's pleasant to eat, the tendency to eat it is going to eventually benefit that random mutation which makes it beneficial by a LOT, immediately.
@@Nevir202 i think more likely that if something is starving, they will try to eat different things. some will kill some animals. others wont. the ones that survive will remember the taste and know "this tastes bad, but it didnt kill me and im not starving now", that said there are certainly some things they wont eat because of the taste, peppers for instance.
@@henrymccoy2306 No, though it did lead to near universal omnivorism, as you'd expect. Like I said, didn't see any other way to work around every plant being essentially a poison landmine, which seemed to wipe out the carnivores most of the time.
From my understanding the main benefit of predation is that predators are able to quickly gain certain materials (vitamins etc) that prey have to farm a long time to get. In real life many herbivores have to spend most of their life just eating while carnivores only spend a short time. Sometimes only catching prey a few times per week. This makes meat a much much more efficient food source. Perhaps you can implement similar mechanics into your simulator? Perhaps different "vitamins" being needed for different tasks such as growth, reproduction, health and movement. Edit: Oh, i had to watch a bit more to see you mention digestion. You can ignore this comment then hahah.
While I agree with the obvious statement that the consumption of meat directly is much more efficient in general, the process of that meat developing into existence is much less efficient overall. A much larger quantity of energy is necessary to be provided to create meat over time than what the meat itself would give. So its not necessarily a sustainable option for all species. In fact, it wouldn’t be possible at all if there weren’t already that the vast majority of life getting energy from plants to then pass down through their body.
Maybe if the plants could evolve too, like they could evolve spikes or poisons, maybe even predatory like behaviour such as pitcher plants or Venus fly traps. This, I think, could help prevent the herbivore population being able to evolve beyond the capabilities of the predators as they will also have to adapt to the changing food source.
That’s a good idea. I doubt it would lead to predation, but if plants could evolve spikes/poison, then some of the interesting stuff about predation would have to occur between the bibites and the plants. A form of fruit could also add another niche both for bibites and plants.
So.. Late to the discussion but why dont the plants also evolve this would make it so the herbivores cant just optimize ito an op state and stay there, they may get op but only temporarily. Looking at plant life irl, additions/evolution of, things such as thorns or toxins have evolved bc like herbivores plants dont want to be eaten either.
Theoretically predators would stop plants from evolving countermeasures as well. Since plants without the required countermeasures to survive in an environment without predators will survive and propagate in an ecosystem that has predators. Herbivore VS Plants becomes an arms-race, Herbivores VS Plants + Predators creates a delicate balance. So I hypothesise.
I've been following the development of this game for some time, and I love the direction you've been going so far! I want to see this game succeed. *Treat plants like you do with the Bilbites* Herbivores are op because there no evolutionary pressure between plant & herbivore... Along with no relationships between Carnivores and plants. Having a variety of plants that come in different forms and a Carnivore - Plant relationship can help resolve this. Here's some suggestions I would like you to potentially consider with the digestion mechanic: *Plant reproduction through energy demand* | Mechanic - More Chloroplasts generate more energy, when a plant reaches a certain amount of excess energy, it creates a spore/seed nearby (And resets it's energy) *Plant variables stats* - *Tougher plant fiber* (Plant digestion time vs Chloroplasts) - Longer digestion means less movement, and making it easier for carnivores to catch prey that rely on speed. - *Varying plant sizes* (Energy & Endurance vs Maturity Time) - The idea is that a bigger plant can survive from smaller herbivores that nab pieces off it, but takes longer to reach maturity to create seeds/spores to reproduce. They also have more chloroplasts. - *Spikes*
Flowering and fruiting plants is also a great thing to introduce. Some plants are inedible unless fruiting, which adds to another niche being developed.
if plants could be multi celled (parts can beak of from the main plant without the plant dying). then the plant could have a thick core section, and then grow leaves or fruit. fruits could spread seeds when pooped out. fruits could also be laxative to reward plants with poop.
To deal with prey becoming too fast, you could implement a form of exhaustion using available short term energy. For example, gazelle are very fast, but they wouldn't be able to keep up that speed for long. You could also increase the cost of speed to make it exponential, while also increasing the energy that meat provides, essentially making speed more cost-effective for predators.
Ingesting plants is really labor intensive for not much in return so yes revisiting digestion is the right call. But you also need to consider omnivores and "fruits" sparse high energy plants. Or you always end up with herbivores and scavengers. It will be waiting for your next video ;-D
Fast rotting and specific demands to digestion of rotten will make the scavenger niche less efficient. At least it did on Earth. The increased lifespan of herbivores will also make free meat rare. By the way, the ability to share digested food with others may result in some interesting things like farming behaviour. Or eradicate the predator niche.
Yeah I think increasing the food diversity in general would be beneficial. We see the most ecological diversity on Earth in places where there is a great abundance of energy sources. Omnivores provide a clear path to predation, but having these niches be sustainable means increasing the base level of plant diversity to support different types of herbivores.
Having age / disease reduce the overall efficiency and vitality of a herbivore would also make it much easier for the predators to be able to hunt the weak and sick and thus provides a natural buffer to their existence rather than having to rely on catching the healthiest of bibbit or scavenging the dead.
A thought on camouflage: You could give the background of the simulation space a color gradient and allow bibites to use a similar gradient. When bibites look for each other they would need to detect a difference between the target bibite's color and the background color. This creates biomes dependent upon background color, allowing more opportunities for divergent evolution.
This is insane. Even if the predators eventually died out, the fact the herbivores evolved to avoid predators, and the fact the predators evolved into 2 separate species was insane. Cudos.
Looks interesting so far! One thing that comes to mind is a board game called "Primordial Soup", in which there are 4 colors of nutrients and your critters need them all. It looks like you currently have just one nutrient, so that the only difference between food types is quantity. You could have plants produce several types in varying amount, then create an anti-nutrient (poison) that a plant or animal can produce to ward off predation. (A fancier example: tannin in acorns makes them leathery, reducing their food value to the ruthless killers known as squirrels. The concentration of tannin is highest in the most crucial part of the acorn, so that the seed-eaters sometimes toss out the partly eaten nut and give it some chance to survive.) You shouldn't expect predators and prey to have an even balance, overall; the rule of thumb is that about 10 units of body mass supports 1 unit in the food-chain level above it. 10,000 pounds of grass --> 1000 pounds of rabbits -> a 100 pound family of skinny coyotes. So as long as there's a viable predator population at all you're probably doing something right. In general I'd focus on increasing the food value of animal flesh, since it's generally more nutritious than plant matter, and/or require some specialized digestion to get full value from plants that have some types of nutrient (like cellulose). You could also allow for biting attacks to rip away chunks of raw, steaming red pixel goodness, simulating something like parasitism and creating a stronger evolutionary pathway to the full-blown kill-and-eat strategy. It could even look like the "milking" of aphids by ants, a relationship that's kind of parasitic, kind of domestication.
If the plants could evolve as well, it would create competition between them and the herbivores, giving the herbivores harder time to evolve optimal predator dodging capability
Yup, maybe start simulations with three different plants, herbivores, omnivores and carnivores That plus some more environmental factors like temperature could help
Great video, when you announced that you were going to incorporate digestion (metabolism) I realized that is a great parameter to introduce. I am not a biologist, but love studying ecologies. One thing to be careful of is plant stagnation. In many simulators, if you have one type of food (plants) being introduced with no change, it can lead to Symbiosis, a situation where two species are living together and very few, if any third parties can enter. In this case, plant particles and the Bibites, are the only two symbiotic animals needed, resulting in any third party species to become irrelevant and go extinct. One way to challenge this bi-special simulation is to allow plants, to mutate and diverge. Having plants change shape, size, color, and energy richness at a much slower rate will allow for new species to enter the market, allowing more and more Herbivore species to generate. More herbivores means greater chances of carnivorous, parasitic, and even scavenger populations to grow. Within Bibites currently, this wouldn't be possible, however with some minor tweaks on how plants generate, it can be done. I will try to add a further detailed explanation on how symbiotic ecologies simulations can be avoided to allow diverse ecology at a later time. Best of luck and excited to see more from you.
I just found this project and I really love the work you're doing. I've just downloaded v0.3.0 and the only gripe I have is that I wish it was easier to read their neural pathways, I understand how they are laid out but sometimes its hard to guess what the inputs and outputs actually mean. But it's just a small thing so it doesn't really matter. I can't wait to see where your project goes next.
1) I think that meat is easier to metabolize than plants is how it works IRL. Meat isn't necessarily more energy dense, it just doesn't need to be transformed into useful proteins and such before use (I'm not a biologist). Make meat a larger energy source and predators might end up with that evolutionary edge they need to exist. 2) If plants didn't spawn in at random and instead were rooted in place, then your predators might evolve ambush predator hunting styles where they camp out near plants and wait for their prey to need food. 3) If plants evolved too, then they could develop different nutrient types (of varying metabolic efficiency to produce) in an attempt to defend themselves from herbivores. This would force herbivores to evolve as a predator to those plants, which might limit how optimized they could become against carnivores due to metabolic constraints. (if digesting becomes more energy intensive, then they might not have the energy to escape their predators). Plants would need a "seeding radius" evolution though, so that they can spawn their children far enough away to maybe escape immediate consumption by herbivores. 3.5) To make this whole system work, plants must also consume nearby dead bibites. maybe rotten meat? maybe bibites leave "skeletons" that only plants can eat. This absorption radius could even be changed via plant evolution.
@@Peter-vv1sb Cool idea! But the encoding might need to be limited to seeds/fruits only. Otherwise we end up with a similar situation to the one we're trying to stop (prey evolving beyond predator ability to hunt them). Ecosystem collapse due to poisonous plants.
It’s amazing that the internet can allow someone to work on a project like this full time. Fingers crossed this channel takes off and becomes financially viable, I’m excited!
I did something similar. The way I did it was to evolve plants and herbivores like you did, but then also let the plants evolve with the same rules. Eventually they’ll start moving, forcing the herbivores to hunt.
Large animal vet here with a thought based on the food. The ability to digest cellulose, comes at a huge cost of needed to rely on microorganisms to ferment the cellulose to usable molecules. This requires a lot of volume and would definitely come at a disadvantage compared to the relatively small digestive system of a carnivore because other animals are much easier to digest. I think that could provide a way to balance out the populations of herbivores vs carnivores. Awesome video and looking forward to see more!
@@TheBibitesDigitalLife Yo, take your time my guy. I've got mountains of patience for you. The quality of your work so far has earned that much at least. I'm just glad to see you've been able to dedicate everything to this amazing project, and I can't wait to see where it'll go next.
I think something you're missing is implementing the advantage of meat digestion vs plant digestion. Meat digests faster and we have the energy available to us faster and it also uses less energy to gain energy. Whereas plant eaters require more energy and time to digest their food. Or increasing lethality of attacks like bleed damage. That could increase the favor to the carnivores.
What might help is evolving plants. In the simulation, only the herbivores and carnivores evolve, the carnivores trying to become better at eating herbivores, and the herbivores avoiding being eaten better. But, the plants are static. In the real world, you also get plants evolving to be harder to eat, stuff like the spikes on cactus.
This is such a weirdly timed and interesting find. I found myself something like a week back just randomly wondering about how exactly predators would have first evolved, and what life would look like if they just never did. All that to say, this looks like an awesome channel and I'm glad to randomly stumble onto it.
Not only should you have plant evolution, you should also have environment evolution in order to destabilize and encourage alternative paths. Tectonics/volcanism, solar radiance, acidification etc.
Environmental evolution is a neat parameter to add, however, it is not necessary for competitive species simulators. I think plant simulation is a good start, while environment can come later. However, environmental parameters should be set and included ASAP, to allow for future additions becoming more easily integrated.
@@TorresTyrus For each ecosystem to evolve you need a perturbation proportional to the depth of local minima of the state and the perturbation duration must be sustained long enough for random sampling to overcome the free energy barrier. Only then can you evolve into adjacent local minima. If you apply a single static forcing function then at best you get a bimodal minima that again becomes stable. The forcing function would then need to be dynamic in order to sample multiple ecosystem states - hence an 'evolving' environment.
this project is what i hoped spore would be like when i was a kid, plz never stop improving on it, cant wait to see where it is in a couple of years, thank you so much for your effor.
Oh yeah that makes sense. Plants tend to stay viable for longer (rot more slowly) and also just simply regrow over time. BUT they are harder to digest and give less nutrition relative to the effort of eating them. So it makes a lot of sense to tackle this problem using digestion, which is a much overlooked thing in simulations like these I think
@Paul Master Gaming yeah, though an ideal version of that would somehow be dynamic in the developed organs. You don't want to fix what precise organs you got, you ideally want to have the world have the sorts of physical properties that drive organ creation For instance, very small things don't need a circulatory system or lungs to distribute oxygen as simple diffusion through cell walls will be enough to get stuff everywhere. It'd be amazing if it somehow were possible to model this accordingly, making circulation and digestion and breathing and perhaps reproduction efforts entirely emergent. But the number and complexity of systems to simulate quickly becomes intractable if you aren't careful. So I'm really not sure how best to go about this. This may actually be more difficult in 2D than 3D because in 2D you can't really have tubes without making an object fall apart. Although there surely are tricks to "fake" stuff like that
It would be fascinating to see a simulation with 3 niches, that each one hunts another one and is hunted by the other one. Like a biological rock paper scissors.
the transfer of energy (digestion or efficiency) is definitely super important when you consider predation. Another thing to consider is that predation is usually done in groups in real life, and that their are "hunting grounds". This may help you understand why your predators are failing as of yet
This is so interesting! I was blown away when scavengers diverged to fill the niche of cleaning up corpses. I am really excited to see where this project goes!
This project is so cool, dude, I am loving the development of these little guys and can’t wait to see where they go next. I’d love to see the food source and environment develop in the future, too- making for wholly diverse and one of a kind scenes.
I was a huge fan of yours 2 years ago, but I forgot what your name was and couldn't find you until today, when this video popped up in my recommendations. I'm so glad it did, and I love your content. I can't wait to binge watch your videos.
Simply amazing. I was wondering, how did you learn to code? Someday I'd love to try creating my own life simulation and I don't really know how much programming knowledge is needed. Thanks!
I started programming when I was 15, but pretty much just for fun. Honestly mostly learned by working on small project like this, I would encourage you to watch some basic unity tutorials and then dive right in trying your hand at it. It doesn't have to be perfect, it just have to be fun!
@@TheBibitesDigitalLife how did you learn to make a neural network, I’m using unity and made a small simulation but I cannot get a neural network to work properly
There an online course called Machine Learning by Andrew Nguyen that does a good job of teaching you all the basic concepts 🤔 Then it's about playing around and implementing that yourself
If you want proper predation to happen, then make plant pellets spawn in clusters in set positions, and have the ones normally spawning all over the place worse. This way herbivores will be incentivised to gather in one place, where the predators can camp. Bonus points if you make the static clusters seasonal, so eventually both have to endure times of unreliable food source.
I'm so glad this got recommended to me, this kind of simulation is the kind of thing that I've been interested in for a while. The interesting thing to me from this video seems to be that by studying real life mechanisms, and introducing that complexity, we get closer to real life. By turning these things on and off, we can see how the complexities in real life have made our ecosystems the way they are!
I'm relatively new, so I may not be aware of all the details. Are Bibites functionally immortal or is there a mechanic that lets them die of old age eventually? The latter could be tuned to encourage more omnivorous species to evolve into dedicated scavengers, due to the steady supply of meat. This would be a bridge to predation. If the meat decays somewhat quickly for example, some scavenger species may decide to increase their energy yield by following around old and sick herbivores, so they can eat as soon as it expires. From there the step to actively killing those individuals is a decently small step and allows them to expand into the larger predatory niches. Overall, having plant matter be "less nutrient dense" (less energy per same sized pellet) and "harder to digest" (releasing energy slower) than meat might be the most potent adaption though. You can see in many real-life species that herbivores usually spend most of their time grazing to maintain their energy intake, while carnivores can often go days or even weeks without food, instantly replenishing their reserves off of a successful hunt. That would encourage carnivory in areas with scarce resources, as well as limit the energy the herbivores can dedicate to evading predators. I believe that's what you're referring to with the "digestion" concept?
Technically they can live forever, but we rarely see that happen. As they age they accumulate more and more inefficiencies that make their metabolism more costly, so at some point it becomes too hard for them to survive. And agreed on the rest, I think that's one of the potential evolutionary pathways for predation 🤔
I'm a game dev and I just found your videos. I'm building a similar simulation and your work is an amazing proof of concept for me. Thank you for sharing.
WELL I'M A GAMEDEV AND I'M WORKING ON THIS SIMULATION AS NOT A PROOF OF CONCEPT All jokes aside, it nice to hear 😁 please share your project of you have links!
Another great video, man! Artificial life is such an interesting topic, and I think the holy grail of artificial life is to find a way to make organisms evolve new body parts, functions, organs and behaviors that were not programmed. I think this can be achieved someday by actually simulating the DNA molecule and all its complexity (along with simulating individual molecules, realistic physics and chemistry forces etc), but it may still be quite too far. But for what's possible right now, I believe your project is the most complex so far. Keep it up!
Let me first start out with I dont know much about coding or anything. Though I do greatly enjoy seeing these simulations. It shows how hard it is for the recreation of the system that are in place on this planet. So please take my thoughts with that in mind. 1) you might want to look into different types of "plants", think tree leaves, grass, acorns, bark. Not every herbivore eats the same parts of plants or even the same plant at all. Now if you could think of a way for the plants to evolve themselves to defended themselves. Think parts that are ment to be sacrificed, like when trees drop all at the same time so the herbivore population cant match the surge of acorns. Than you have plants with thorns. Some with toxins. But you get the point it forces more niches for herbivore types, which might help the carnivores. 2) I think another thing to apply is a season cycle were you can simply say cycle time 1 full plants, 2-3 is half plants, 4 is little to no plants. 3)it might also help to implement at a certain point carnivores cant hunt herbivores because they are to big without ether growing themselves or forming a pack. 4) I know you made it so carnivores could throw things, but encourage their coding some how to prefer that method more. I say this because many carnivores have ether lighting fast attack that allow them to get the prey before they react or honestly just shot something at them to immobilize them. 5) finally you might want to look into a way carnivores can teach their young how to hunt thus allowing for one that might of figured out to attack from behind in a blind spot to than teach future generations to do as such.
AMAZING! I love this so much! 🤩 So interesting to see how the herbivores and predators evolved to fill different niches over time! Good luck with the next part of the project!
THIS IS SO COOL HOW HAS RU-vid NEVER RECCOMENDED THIS TO MEEEEEEE! A while back before the pandemic really hit, I had a computing class and the final project was working with a group to make a videogame. We made a snake clone, but one of the ideas I wanted to implement but we couldn't was an "Evolutionary" factor that would give players different drawbacks and benefits, and tying it to a color gradient "gene". After that I had an idea of a simulator type snake game where you set up your creature like spore and play the role of the random genetic mutations and decide how to implement changes. SO THIS PROJECT MADE MY BRAIN PERK UP SO HARD! It's really cool seeing how scavengers evolved out of the pressure of the systems, that to me shows that, whatever the hitches are, it's on the right track for something at least, I look forward to this video project series!
Awesome! And thanks! I would encourage you to try your hand at it too! Programming is kinda like magic. You write up a spell and then it executes, and produces (often not 😅) what you wanted! And Artificial Life is one of the most interesting and fun things to code in my opinion!
Hi, I'm a Complex Adaptive Systems engineer from Sweden. I did a similar (but less detailed) simulation-program a while back. I got the same result as you, herbivores where op. So I decided to let predators eat each other as well. This got an interesting result where eventually all carnivores became ambush-predators, standing still until they got hungry (conserving energy). Then they waited until a prey came close enough. When they had children, the parents let the children eat them, along with all their siblings, until only one (on rare occasions two) remained. This way only very few predators where on the map at the same time, making the herbivores slow. And since predators where standing still, herbivores learned to not care about them. This actually got very stable and lasted long enough for me to end the simulation myself. Although, it did not change much at all once the strategy was set. Not sure if this helps in any way, but I found it fascinating non-the-less. I wish you good luck with your project! :)
I love ecological simulations, man. The creative possibilities are endless. For example, one could make it so that food isn't a resource, and the animals have an arbitrary objective, like maybe slap each other, to stay alive. This stuff is great. Keep on it. And my ideas for the bibites: you could make it so that they drop a little food when damaged, not just on death. Of course, you would need to keep the law of conservation of energy in mind, such as make damage drain energy, as well as drop food. This would allow for the transition to predator to be viable, so you wouldn't just have to introduce an engineered species every time.
You could try to make them have to evolve their bodies, so rather than just evolving to increase the speed variable they'd have to slowly go through all the physical convolutions to actually make themselves go faster, this could also impact the eyesight, size, weight, and so on. I especially like this idea because you'd be able to see more variety in how the creatures than just what variety gets hard-coded.
Firstly i think what you are doing is awesome and will eagerly be waiting for more vids. I would imagine one of the key drivers in predation is 1. time to find/consume energy (hours eating grass vs 1 hunt a week), 2. energy density (meat vs grass) & 3. energy loss/time for digestion. My understanding is that meat is much more energy dense 5x - 20x, requires less time and energy to digest (mins/hours) and can be done in relatively quickly compared to herbivore consumption (hours/days). I think including these three factors, if they aren't in there already would dramatically increase carnivore population size.
I wonder if it would be beneficial to the simulation to split the meat into two groups. : one that is hard to digest with a low energy content and one that is easy and energy dense but that has a fast decay rate. This sould force the split between predator and scavengers, and this way you could independently tune the simulation parameters to find a stable simulation with predators.
I agree but in a different way. I think that when the meat is first dropped it is super easy to digest. Then after a while it becomes harder to digest hurting scavenger.
I'm new to following your work, so forgive me if you have taken this into account and I just haven't seen that video yet, but... 1) Based on other videos you are looking at an energy transfer aspect. Part of the Carnivore vs Herbivore dynamic is that the Herbivores generate vitamins the carnivores cannot get elsewhere, such as vitamin B12. Perhaps your herbivores need to use a small fraction of their energy to generate a secondary type of energy that some generational mutations create a need for in place of the plant based energy. This would force come generations to hunt because he plants produce type A energy, but they need type B energy. 2) The overwhelming favoring for herbivores could be tied to imbalanced plant growth. A lower supply of plants would limit the availability of plant based nutrient, which would, in turn limit herbivorous population growth and mutation rates. 3) A meat based diet has a dramatic effect on brain development, which is why carnivorous species tend to be more intelligent than herbivorous. This could be represented in the animal produced nutrients, (Energy B as it were) increasing the odds/rate of intellectual evolution.
I think that you could add some code that dictated how hard a certain food source is to eat, and one for bite force. For example, at the start of the simulation all the Bibites could be soft bodied, with low bite force, because plants are soft and easy to eat. Then by random chance a Bibite would mutate just enough bite force to start consuming other Bibites. Then the Bibites that can mutate tougher skin would be able to survive and make more hard-shelled Bibites. From there it would be an arms race, likely with the other adaptations coming into play. I also think a health mutation would be useful so the carnivores don't just one shot everything.
i should note from time to time to consider the stress patterns of english, which are generally important: devélop, devélopmènt, encóurage, encóuragemènt, épisòde, biológical, órganism, orgánic, órgan; stressed syllables (up-áccent) end up sounding longer and have more well-conserved vowels (the other ones may relax, often all the way to the laziest vowel depending on how quickly one speaks). Secondary stress (down-àccent) also slows down the relaxing process. Stress patterns that you already perform well: journey, important, define, predation, obvious, exact, interaction, carnivore, herbivore, exploit, possible, energy (most words, really)
Cool project! Predators in the wild often have trouble catching their pray in their prime so they usually target the young or old/damaged individuals. Exploring this area could maybe allow the predators to not go extinct.
Maybe the solution to an evolutionary pathway is to introduce the niche of an omnivore - basically instead of having two food options, increase the number to 3 by introducing fruits or smg like this. Pure herbivores only gain limited energy from fruits, less from meat and most from plants; omnivores most from fruits, but still relevant energy from plants and meat, while the pure predator diet drops the energy from plants and reduces the one from fruits. This would allow a stable pathway that doesn’t punish steps in the direction of evolving to a hypercarnivore. Love the content and excited for more!
I actually did something like this at my old job, since I didn’t have anything else to do. One of the outcomes after several million timesteps was an ecosystem with the following: A predator that traded mobility for a slow metabolism, it would sit in one spot and eat anything that came into it. Stationary plant species that survived thru photosynthesis Herbivores that would walk randomly to find and consume plants. The plants had formed a symbiotic relationship with the predator. The plants attracted herbivores for the predator to eat, and the predators ensured that the plants around it wouldn’t be consumed entirely and could still propagate their genes. I think the reason stationary predators evolved was because it was too much energy to move, and two predators encountering each other would fight, and reduce each other’s energy levels further. This was all done in MATLAB on a 2D grid system, with several n-dimensional plots of the genome distribution + biomass. It was a neat projectZ
I think trying to make new optional environment further down the line, like introducing simple weather, climate and cyclical seasons factors could yield fascinating results. They could influence the plants' growth factor and/or nutritional value. Fruits/seeds could be a viable food source for long term storage to survive the winter, while predators could try to enter a type of hibernation, slowing their metabolism. Also, imagine highly territorial herds if bibites protective of their lands with highly nutritious plants or of their communal piles of harvested food. Migrating bibites, provided there's a map with climate and biom gradients. On another note, I think it would be extremely interesting to make eusocial bibites that could mimic some aspects of how ant and termite colonies are organized with having an egg-laying queen who's only concerned with one task while drones gather food, share it with others by using communal stomachs and perform other functions to maintain the colony as though it were a single multicellular organism of sorts, with further potential for complex interspecies interactions like ants herding aphids (protecting them from predation and transporting them from plant to plant and consuming their sweet secretions as if they were cows).
this is also the first i hear of your project!! i think its cool and im excited to see what you do! also i peep the kurzgesagt poster 👀, glad to see ur a man of taste
Holy shit.... this is like Asteroids crossed with Spore I thought there were no more genre crossovers left to explore but you've singlehandedly proved me wrong
Couple ideas - A movement debuff shortly after consuming plants - Anyone can digest meat but the ability to digest plants requires a tradeoff - Fresh meat is far more valuable than rotting meat. Perhaps digesting rotting meat gives a slight debuff as well - A more narrow scope of sight and slower turning - An evolvable stamina meter that affects your metabolism cost depending on your current speed, while your current speed affects the stamina meter - Evolvable stealth in a slow-to-traverse environment with gradually different background colours
what an interesting projet! already looking forward to the next steps! I'm thankful the almighty algorithm recommended to me a fellow from Québec (saw the UdeS shirt haha)
Please add simulated evolution of the plants as well, that is something many evolution simulators miss. Plants can be easier or harder to digest, grow faster or slower, near or far dispersal of seeds, etc. New models look great
Perhaps one thing to consider is the effect of age on animals. A lion will not target a healthy adult gazelle but rather single out the young, old, or injured because they are slower and weaker. Just as a lion targets and catches an older slower gazelle, the predator Bibites could target the older slower herbivore Bibites. Introducing the concept of a lifespan to the Bibites could help the predators maintain a healthy population. I.e., after a certain length of time Bibites become slower and weaker, this would allow for predators to ideally always catch a small portion of the herbivore population even if the predators cannot catch the healthy fast Bibites. I think this change would be ~relatively~ straight forward to implement and could potentially lead to some interesting results. Awesome project, 11/10!
Bein oui! Au début je pensais pas que ça allait pogner de même ce projet là... Le nom est un peu une heureuse erreur 😂 mais je suis content au final d'avoir un terme québécois comme nom pour le projet!
It is the first of you videos I watch and I have not read other comments but here are some of my ideas : - escaping predator should be costly in energy. Good escape should have to feed more in order to maintain themselves to the point where they compete for food. This way, there can't be only super good "escapers". Seeing the video, they evade by turning, so they could use X energy each time they turn Y times - depends on how things are implemented but hunting in packs is superior to hunting solo. There could be an evolution that let predators to hunt already targeted preys. This reduce the total meat for one member each hunt but let them at least have some - diversify plants. If plants evolve to not be consumable by herbivors, herbivors have to adapt to plants AND predators making evolution more costly. They will either have to adapt to predation first or plant first, making different niches. I liked the idea of color used to show evolutionary path. Maybe you could have 3 values for plants (RGB), and same for herbivores and carnivores. A species could eat something only if it's R in in range of R of target +/-10, some for G and B. That would make specialised plants, specialised herbivore and specialised carnivores
As many other people say, there is a need for calculate how easy food is to digest or how much energy meat gives compared to fibrious plant. But, an important step I think, is also energy storage. Having storage that is costly to upkeep, but can be spent to boost a creatures speed or attack. Then having storage that is slowly releasing energy, but is really energy efficient. Energy storage is a thing I see many simulation almost ignore, but I at least personally see as a very important factor in the animal kingdom.
You should give the bibits the ability to see and eventually recognize the special body parts of other bibits (ie: paddle length, size, color, mouth shape, armor)
This is the 3rd vid of yours I watched now and saw the problem before watching this video. I'm glad you addressed it. The plants are not co evolving. If the food source developed to be less eaten the bibites would have to choose between develop the means of better eating the plants or better eating the plant eaters.
