The Evolution of Predation in a Simulated Ecosystem

Welcome back!
I recently decided to embark on a quest to unlock the evasive dynamic of predation!
Predation is extremely rare in artificial life simulations that don't hard-code it directly.
For The Bibites, I want to make the EVOLUTION of predation possible.
Considering The Bibites is an open-ended simulation, there's a lot of details that need to be just right to make that possible.
This is going to be the first episode of the series where I'll try to achieve that.
In this one, I'll present a few basic principles of biology and ecology in order to show you just how hard of a task it will be.
=================================
The Bibites is an Artificial Life (Living AI!) simulation where I recreate some biological processes and let the lifeforms live, eat, reproduce, and mutate, leading to active evolution.
Their can evolve their body through a genetic algorithm and their behavior through a custom neural network algorithm.
=================================
Follow me on Twitter : / thebibites
Support the project on Patreon : / thebibites
Join the community on Reddit : / thebibites
Download and play the game : leocaussan.itch.io/the-bibites
Subscribe to the channel: / @thebibitesdigitallife
=================================
Credits:
Footages from Pexels.com and Videezy.com
Procedural Sprites by Brax - Twitter@Braxiations
Intro tune by Fireballme
#evolution #AI #devlog #simulation #ecosystem

[@kurzgesagt]

Nice notebook and poster! And what a great project!

[@TheBibitesDigitalLife]

Thanks so much! And what an honor! I'm obviously a fan, it's amazing to be noticed by a Senpai 🥳

[@scrotymcboogerballs6756]

@@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!

[@nobody6032]

The fact that Kurz said something is fucking insane. You can quit KZbin now.

[@chester9585]

Ooh

[@chester9585]

@@nobody6032 it would ironic if that actually happened

[@Yellow_Slug]

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.

[@TheBibitesDigitalLife]

Thanks a lot ❤️ Yeah, the developments are always very interesting when it comes to evolution!

[@aenetanthony]

@@Flandre-Scarlet How is that insane?

[@shannonsmith3818]

@@Flandre-Scarlet Isn’t evolving an advantage to better survive in their environment through a mutation the definition of adaptation?

[@autosneak1618]

@@Flandre-Scarlet adaptation is the filtering of disadvantageous mutations from advantageous mutations, which is evolution, so what's ur point

[@cortster12]

@@aenetanthony Ignore him, he's a science denier.

[@CodeParade]

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.

[@macslash5833]

Oh shit codeparade I love your vids man, how's hyperbolica doing btw

[@TheBibitesDigitalLife]

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!

[@Chareidos]

@@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.

[@kulgydudemanyo]

@@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.

[@astick5249]

Oh i had a similar idea but i like your take on it more

[@drJoep043]

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.

[@scrotymcboogerballs6756]

That is a great idea! If you think about it, for example lions would have a much harder time if their only food was fully grown elephants.

[@drJoep043]

@@scrotymcboogerballs6756 haha! indeed very good example.

[@cucumbercow7379]

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!

[@tacobell222122]

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.

[@jabuki2]

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.

[@SirAroace]

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.

[@pupip55]

I would say nutrients heat map for multiple different nutrients so growth rate is different based on the radios

[@Tetrix9898]

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.

[@SpydrXIII]

realistically meat, and poop from meat-eaters is bad for plants.

[@dappershinx9234]

@@SpydrXIII I'm pretty sure that's objectively not true.

[@Bone_guy]

Is it possible for them to be omnivores, then predation could be passive or active.

[@itsalexmullen]

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.

[@infiniteplanes5775]

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.

[@TheYeetedMeat]

I made a comment on plant evolution. It’s buried in the boids video

[@misspurdy27288]

Yeah, artificially forcing herbivores into more niches would help canivores a ton.

[@mrbam4739]

I just commented something like this. I should have scrolled down farther.

[@mlvluu9836]

plantsattachtocarnivorestoprotectthemselvesfromherbivores

[@peqpie]

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.

[@user-lh5hl4sv8z]

In real life, a lot of predators use strategy rather than just chasing them

[@jomamabee2481]

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.

[@JepTheLegend]

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]

Thank you so much 😁

[@warbrain1053]

@@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

[@i_inject_mercury1930]

@@warbrain1053 yea, But to a point not to make large carnivores useless. I think small creatures shoud be scared of large ones

[@warbrain1053]

@@i_inject_mercury1930 yes but smaller carnivores in herds could overpower a large one

[@i_inject_mercury1930]

@@warbrain1053 100 rats still coudnt kill An elephant

[@Nevir202]

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.

[@makelgrax]

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!

[@Nevir202]

@@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.

[@jessiejanson1528]

@@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]

Wouldn’t this just make all bibits carnivores? Seems like it defeats the point of evolution to make the value any more than 0%

[@Nevir202]

@@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.

[@theqracken4035]

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

[@BrutusAlbion]

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.

[@LittlestTeapot]

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.

[@DerSolinski]

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

[@antonstezhkin5991]

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.

[@StrawberryFPV]

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.

[@jeupater1429]

The nutritional value of meat should not be a linear decay. It should decay as a reverse exponential such as: x^(-1)-0 .1x+1

[@antonstezhkin5991]

@@jeupater1429 smart move! This will meke a fast scavenger niche.

[@khlorghaal]

this depends on rules of the sim, how accurate is the herbivore analogy

[@robot9604]

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*

[@nathanwi1147]

Wow a lot of good idea

[@TorresTyrus]

Flowering and fruiting plants is also a great thing to introduce. Some plants are inedible unless fruiting, which adds to another niche being developed.

[@nilsdock]

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.

[@TiagoH1710]

Maybe the plants of the same species grow to different sizes depending on location?

[@zmaster1481]

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.

[@Ugly_sextoy]

It’s a good idea but maybe that will make the game too complicated

[@windradyne8724]

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.

[@the_icel0rd247]

this is actually a really good idea im all for it

[@zmaster1481]

@@Ugly_sextoy not a game but a simulated ecosystem so I disagree.

[@zmaster1481]

A final thought, like in nature the herbivorous group could adapt to the presence of toxins an even harness it's power for their own gain...

[@CirJohn]

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.

[@gangsterguardsman4576]

This would be awesome, probably add so many new possible niches

[@Pyriphlegeton]

Your explanations are getting insanely good. I really love all the evolutionary theory behind the project!

[@TheBibitesDigitalLife]

Thanks a lot! I wanted to take the opportunity to make educational content at the same time, obviously the project has a big educational potential

[@aarnijarvelainen8499]

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

[@Kavriel]

Indeed, plants aren't as passive IRL than in the simulation.

[@LostBoyHaruko]

I'm not sure if people are mentioning this but your editing with your random pop-ups is really fun and unique!

[@TheBibitesDigitalLife]

Thanks for the comment 😁 I try my best, especially with this video!

[@jemmerllast8492]

Hopping on the late to discussion train (I just found this amazing project!), but can herbivores hurt each other? Some real life herbivores will fight over resources, which would provide a little more meat into the environment (in the event of a kill, that is. Just injury is more likely, with bibites learning to leave an area if injured). In addition, wounded animals would likely move slower, making predation easier in the event of highly evasive prey. Then again, this may all depend on the addition of passive offensive and defensive features, which I do not believe is a thing in your simulation atm.

[@zombielizard218]

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.

[@helldronez]

or capsaicin in 🌶plant

[@2rfg949]

haha i was there when your first started. so glad they're still alive!!!

[@TheBibitesDigitalLife]

And going strong! Maybe slowly tho 😅

[@copypolice]

One thing that struck me watching this video is that predators often can't kill the most healthy prey. They more often kill the wounded, sick or older amongst their prey. This helps keep a balance between available food for herbivores by thinning the herd and feeding the predators.

[@krisschnee3151]

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.

[@mchsprod]

I’ve been watching for a while, and I’ve been wondering: What if the plants could evolve too?

[@MaxTheMuffin]

That could make it less op

[@StandardName562]

@Paul Master Gaming Lol, plants are OP

[@Cl-2048]

That would change it up quite a lot.

[@TorresTyrus]

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.

[@EbonyWolf.]

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.

[@smorcrux426]

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.

[@connorhaley3190]

Parasitism is in the simulator now, although I haven’t seen predators evolve to specifically hunt them, but that is probably because I don’t have it

[@Ulmaramlu]

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]

Would the plants evolve a sort of encryption code to their nutrition that the bibites would have to specialize into in order to gain access to it?

[@Ulmaramlu]

@@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.

[@jonwallace6204]

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.

[@TheBibitesDigitalLife]

Do you have a link/video ? This kind of stuff is always very interesting to me (obviously 😅)

[@Tetrix9898]

Evolving plants would help. Then the herbs would have to evolve with pressure from the plants and from the preds.

[@pustota7254]

Plants running away

[@theprogrammer32]

I just happened to find this channel. The editing is top-notch! I can't wait until the next video on this series

[@anthonyzeedyk406]

this project is extremely underrated. more people need to know about this.

[@TheBibitesDigitalLife]

Couldn't agree more 👀

[@alacer8878]

I've been here for- years. And every time a new update comes out, I just get so excited to watch. Amazing stuff, my guy.

[@TheBibitesDigitalLife]

Thanks for following the project for so long, I'dd love to give you more but I'm doing my best! At least I'm gaining experience 😁

[@alacer8878]

@@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.

[@TheBibitesDigitalLife]

@@alacer8878 Thanks a lot for all the kind comments, I'll keep at it 🥰✊

[@sebastianfries274]

Finally another episode of this fantastic journey

[@SuperUltraBob232]

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.

[@minerharry]

Amazing video! Your explanation of evolutionary paths was exquisite, best I’ve seen

[@TheBibitesDigitalLife]

Thanks a lot 😁

[@Kram1032]

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

[@Kram1032]

​@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

[@lekilleur92300]

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

[@Clumbob]

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.

[@TheBibitesDigitalLife]

Great minds think alike 😎 And by great minds I mean you and all the other viewers, not me...

[@halschultz]

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!

[@phrozenwun]

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.

[@TorresTyrus]

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.

[@phrozenwun]

@@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.

[@mk1interactive944]

irl, plants have a much less value as food over meat, -which allows carnivores to go longer without food.- (this was incorrect but I'm keeping it in for anyone who reads the comments on this comment) allowing a "larger stomach" meter and giving a larger value to meat should make it so predator survive. also, I think adding stealth would open up alot of nieces! I also noticed that there are no unfriendly herbivores. they all are just like, "oh, you ate my food... guess I'll starve"

[@ebob0531]

Carnivores can not go longer without food than Herbivores. Herbivores use less energy and or more energy efficient because they ingest low energy foods, allowing them to go much longer than obligate carnivores can without food. The reason the predators die is because there isnt enough prey in the simulation. For example, IRL there are about 2,400 wolves in Minnesota USA, and about 945,000 deer in the state. Granted humans also hunt, but there were only around 60,000 tags claimed in 2019. SO basically you just need a very large prey population for a predator to sustain itself (1 wolf consumes about 30 deer in minnesota per year)

[@ebob0531]

Probably what it needs is a debuff for old age or disease. Most predators actually only kill sick or old herbivores. That is also what allows them to sustain themselves and helps keep the prey population in check as they kill off older and sick prey

[@mk1interactive944]

@@ebob0531 what if you gave the predators a hibernation period like bears

[@markoalex8819]

If you added some objects, that the herbivores could move around, it could lead them to making shelters they can hide in.

[@potentpotassium5776]

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.

[@TheBibitesDigitalLife]

Welcome back 😁!

[@sick_icarus]

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.

[@doompoison2365]

I am so stoked for thid, the models look captivating!

[@lartnestpasdemain]

This project is incredible, I will make sure to follow its progress, and share ideas if I can think of something useful :) Keep up the good work

[@TheBibitesDigitalLife]

Thanks so much 😍

[@Haat6]

Love this content. Getting into these dynamics in a digital setting like this really spikes my interest.

[@pabloh6369]

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!

[@TheBibitesDigitalLife]

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!

[@pspeasant6220]

@@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

[@TheBibitesDigitalLife]

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

[@vikingwalrus2253]

Really cool stuff, not just the simulation, but the animated presentation of all this information is well done too. Awesome!

[@TheBibitesDigitalLife]

Thanks a lot 😁

[@m1x0]

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.

[@risingSisyphus]

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!

[@williammills-curran2575]

I just discovered your channel a few weeks ago and was horrified to find your previous post was four months ago. It's really interesting and impressive to watch the progress of your wonderful project. I am delighted to find you posting again.

[@TheBibitesDigitalLife]

Thanks a lot for the subscription ! I know I'm pretty slow, but I'm getting better 😅

[@aaronspeedy7780]

The procedural textures look so good!

[@wintryhard]

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.

[@higginsisaac]

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!

[@TheBibitesDigitalLife]

Thanks for the kind words 😁 I sure hope too

[@kovi567]

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.

[@SKy_the_Thunder]

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?

[@TheBibitesDigitalLife]

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 🤔

[@AhNoWiC]

This is absolutely fascinating, easy subscribe, looking forward to the next chapter

[@TheBibitesDigitalLife]

🤟 Thanks a lot 🤟

[@classicmicroscopy9398]

Best explanation of ecological niches I've heard! Thanks!

[@TodePond]

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!

[@I_eat_fabric]

This is completely unrelated, but I have an idea: add viruses that can mess with the genes/brains of their hosts. It would also be cool if these viruses could evolve to target specific genes.

[@sebastiangomez3750]

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

[@TheBibitesDigitalLife]

Real recognize real 😎🤝😎

[@davawen9938]

It's been a while and I really loved this video! may the algorithm bless you :p

[@pedroheck3667]

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!

[@TheBibitesDigitalLife]

Thanks a lot! And I agree, I want to allow new modules and functions to evolve by themselves but I still think that's a little far off 😅

[@gamblingguy6879]

Great video as usual, can't wait for the others. My 2 cents: try evolving the whole world along with the creatures. I understand your pain, i've tried it myself multiple times, and always struggled to achieve balance. The main issue (imho) was the hardcoded nature of the environment, plants reproduction rate, health/energy gained by different food sources, initial population, starting number of predators/preys and so on. My solution was to let the hardcoded variables to evolve along with everything else. You don't just rate bibites, but you also have a fitness score for the environment itself, that will try to evolve those parameters too. It's extremely time consuming, since you have to wait for the ecosystem to collapse before giving it a score, but i found it much interesting than try to manually search for the perfect starting condition.

[@fluffy_tail4365]

fascinating endeavour! There are so many possible tidbits on what it might be going wrong, but the digestion seems a good lead

[@mateusnicolinibezerra9757]

This is so fucking amazing. I can't wait to see this channel grow and catch the attention of researchers around the world

[@AdolphusOfBlood]

I feel it should be pointed out detritivore will evolve before Predation

[@TheBibitesDigitalLife]

True! It's theorized that scavenging is an evolutive precursor of predation, and will be part of the evolutive pathway toward predation!

[@schabe6419]

And I thought I was the only one who cannot get real preditors alive... XD

[@TopNotch770]

Outstanding work friend! Looking forward to your future videos! I'd love to see you comment and speculate on the creatures from Avatar and how they've developed their unique biological features as well as their relationship with the environment!

[@franklinz8098]

In real life, there are many instances where carnivores died out. New carnivores almost always evolve from a small opportunistic creature that is mostly herbivore or scavenger. Your simulation only had meat and plants, with no variation in plant food, while in reality cellulose food like leaves and grass is very different from fruit, tubers, and seeds. The latter spawn more infrequently, requires good movement speed and senses to get, and offers a lot of energy. So in this sense, "spawning meat" is actually the right thing to do, as it creates middle ground. Another thing is that your simulation is much much smaller scale than an actual ecosystem. It is more akin to an ecosphere in a jar where keeping predation is also super hard because any fluctuation will make the predators go extinct, and only works with "hard-coded" predators. What you should do is increase the scale of the simulations, and also make species much harder to go extinct.

[@CalebTerryRED]

Maybe instead of spawning meat, some plants require more omnivorous tendencies to digest properly. In the simulation the optimal herbivore digestive tract is to focus exclusively on plants, maybe we change it so that some plants (like grass) require a heavy emphasis on digesting plants/cellulose, but digesting nuts will be more efficient if you also evolve protein digestion. Then some herbivores will naturally be closer to digesting meat than others

[@4984christian]

Wow, the production value of the video is insanely good :D

[@TheBibitesDigitalLife]

Thanks! I really did put a lot more effort in this one, and I'm starting to become better

[@gerritthompson2246]

I know I have commented a ton of comments but I have an idea: have trees where there are fruits which are better than pellets but also allow predators to hide in the trees

[@greatestgrasshopper9210]

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.

[@Dealve]

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!

[@TheBibitesDigitalLife]

You pinpointed my sentiment about all this exactly 😁

[@adriankowalczyk815]

Maby plants should be able to evolve too so they can avoid being eaten?

[@adriankowalczyk815]

O:

[@sharansiva7635]

I think you might need extra neural node possibilities to allow for more complex behaviour.

[@dude-jk2hn]

Oh my. So many variables to simulate! Best of luck.

[@phillyphus]

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.

[@TheBibitesDigitalLife]

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!

[@gangreneday]

I for one welcome our new bibite overlords

[@just_a_dustpan]

Plants are much harder to digest than meat, all of that cellulose if cell walls is so incredibly different to the protein of meat

[@awaski977]

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! :)

[@pierreboyer9277]

The video is great & fun. And your system seems very interesting. I'd like to know more about the code. Maybe a video presenting that? Cheers

[@LisaSimpsonLiberal]

I hope this is still being developed

[@TheBibitesDigitalLife]

Absolutely is

[@korigarim]

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.

[@nathanwi1147]

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.

[@FierinMahLazor]

THIS IS SO COOL HOW HAS KZbin 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!

[@TheBibitesDigitalLife]

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!

[@JaniMikaelOllenberg]

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.

[@teambellavsteamalice]

Very interesting, it seems you run into a similar problem as Thrive. There photosynthesis is just too good not to have. With all the energy you need you can just add more flagella and seek ammonia and phosphate at a leisure not having to worry about glucose. There is no real downside. Just like to getting huge. The best is a huge spikey killer algae... What either problems need is a cost of benefit dial you can tweak. For bibites digestion might be a perfect answer, and also relate to the real world. A herbivore doesn't need to always outrun the predator. Just outrun long enough to make it costly and outrun other herbivores. The cost of speed should be higher for herbivores, like less energy storage which they should need. You can't really simulate sprinting, sneaky or poisonous predators, but you can make them digest a long time on one prey. So very efficient metabolisms that don't add much unnecessary mass. They will be slow while digesting, but they won't need speed then. You could add omnivores that are a bit slow compared to predators but more efficient and faster than herbivores. They will have a cost of needing some meat in their diets, or being medium efficient in both types of food.

[@the_cheese_cultist]

the memes will be amazing

[@itsjustalf7747]

I love when you can look at a predator and a prey species and see a latter of the prey getting better then the predator then the predator trying to top that and so on and so forth, and I find that absolutely fascinating and seeing you atleast somewhat making that into the game is really cool to me

[@emiliowd]

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.

[@iamasalad9080]

Co-evolution also occurs because of herbivory. Would be interesting if plants could evolve too.

[@janthebest1121]

Lets goooo

[@omniparadox137]

I just wanted to say it was a nice touch that you added at 3:31

[@user-pk9qo1gd6r]

The explanation is simply wonderfully done!

[@TheBibitesDigitalLife]

Thanks !

[@PrimumQualia]

Maybe add that plants don't want to be eaten and maybe they can evolve along side the Bibites, toxins spikes or protection can create biodiversity as sum adapt to eat a specific type of plant, that might make carnivory more likely as well

[@aturos.4542]

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.

[@mikepagliaro2123]

Wow ... what an amazing project!!! I'm not in the field of science. I'm just enamored with the magnificence of natural world. Your digital experiment is amazing, fresh, and cutting edge science. You and your team show what the ingenuity and cooperation amongst us can accomplish. That being said, at what point have you consider the concept of "cooperation" might play a role for the carnivores??? It's possible that "pack hunting" may play a part somewhere down the line. Although, I'm sure you and your team have considered this strategy as a potential "life line" to preserve the predators. Anyway, phenomenal channel!! I'm looking forward to watching every video and becoming a subscribing patreon. Keep it up!!

[@TheBibitesDigitalLife]

Hey thanks a lot😁 It means so much ♥️ There's a few examples of cooperation that I have observed 🤔 one of them I presented in one of the shorts on my channel if that interests you 😁 but of course I'll look forward to adding more functionnalities that will open more doors in that direction!

[@MrJoosebawkz]

unrelated to the video but you really glowed up dude! The haircut and new style suit you really well 😎

[@TheBibitesDigitalLife]

Thanks 😏