I Programmed a Living World...It Evolved...
Вставка
- Опубліковано 13 жов 2024
- Join the channel and get special perks including source code and project files for the ELB_Charts and AI projects: / @eightlittlebears
Join the Discord for free: / discord
---
Evolution simulator! Simulating natural selection! Biomes! Artificial Life Simulation!
Hey everyone! Don’t forget to join the Discord channel! It’s free! Link in the description!
This is the percent chance of a seed not germinating after 100 days
36.23720178604972098813098663067922123278%
maybe with creatures in the simulation will change the plant behavior, longer lives would be less efficient because they would be eaten more
@@Karnivor-0us yeah it’s a good point. It might also be that getting to full maturity more quickly is more valuable with animals because plants could start to fruit and animals could help disperse the seeds, etc.
@@EightLittleBearstrampling by large animals, defenses against (specific) herbivores and the presence of nutrients as a result of dead animals and feces could lead to interesting effects too.
Not to mentions hurricane's tornados , ect
@@harrysarso I genuinely read “hurricanes and tomatoes” 😂
@@EightLittleBears 🤔 thats called seed dispersal i think ;p
I love the simulation, but my favorite part is always your art! It really adds a lot of life to everything outside of pure numbers and calculations.
@@t-money2558 thanks! I’m not as good with the numbers as some of the other sim gurus so I try to keep it as visually engaging as I can 😃
Honestly, what I liked most was seeing the plants completely cover the colder enges, hibernate when the winter came, and then just keep going, flashing between green and brown.
@@samueltrusik3251 the interesting part about that which I forgot to mention in the video is that going brown actually indicates that the plant is struggling to remain hydrated (as opposed to explicit hibernation). The fact that extreme latitudes which have volatile seasons caused this to happen en masse and in unison wasn’t explicitly designed.. it is closer to (although not exactly) an emergent behaviour in the overall environment (I.e. I didn’t design hibernation, but the plants evolved it anyway).
This is really cool! Do you have any plans of implementing this alongside your animal evolution algorithms?
@@jesperisborn9894 yep! They are built on the same core so the next stage(s) is probably going to be tying herbivore diets into the new plant system.
It's likely going to be very difficult to simulate properly the biodiversity of rainforests without having to zoom in to focus on a specific location & then creating 3D renditions to accommodate the various plants at ground-level, in the canopy, & everywhere in between.
While trees in rainforest biomes are numerous, their general strategies for success are quite similar.
What's interesting is that the simulation chose to prioritize growth in the desert, but that might have something to do with their proximity to water. Something I've noticed is that while plant growth near water, particularly rivers & river mouths, isn't necessarily diverse, you will see a slight uptick diversity in trees, algae, & moss, specifically.
I don't know if there's proper research to back any of what I'm saying up, but I am a nature nerd who grew up on documentaries, so it's just some observations I've noticed.
@@MildlyOCD interesting points. I have been thinking about how to visualise the diversity of ground-level plants in rainforests but haven’t come up with a solution yet.. I’m sure one will come! It’s a good point about the similar strategies of trees in rainforests as well. That isn’t something I thought of!
@@EightLittleBears Maybe you could do some sort of parasite plants like vines that abuse taller plants for easier access resources like sunlight in dense areas.
Just commenting for the algorithm and to wish you the best of luck on your future endeavors! Love your vids!
@@MattCaballes thanks! Glad you like them!
I heard someone say that comments don't affect the algorithm anymore, just watchtime and likes and subscriptions (and presumably super thanks and channel joins too). But I don't have any sources for that to back it up.
It's always fun to see thatQ even at such simple and small scales, complex interactions and chaos can arise!
It'd be interesting to see how different types of seeds would effect the germination system?
Like smaller seeds get fewer chances to germinate, but more can be made, and they can travel farther.
While large nut-like seeds get more chances to germinate but can't travel as far, and the parent plant must use more resources to produce them!
Yeah I agree! Plus birds and stuff can carry the smaller seeds around so it can introduce some interesting interactions when animals come back into it!
Not sure if you already considered it, but it might be an interesting idea to make the temperature tolerance a variable too, with higher tolerances allowing for viability in a larger range of environments (or in more extreme seasons). Not sure what the tradeoff for this generalism should be, perhaps an overall lowered germination chance so they have better chances in varied environments (since other plants wouldn't be able to germinate half of the time at all), but worse chances in stable environments.
Yeah that is something I am considering but, like you say, the trade-off is the tricky part.. It will probably be something along the lines of specialists being generally better at everything as long as they are in their environment.
@@EightLittleBears My first idea is to make them more sensitive to overcrowding than specialists, both when trying to germinate and while growing. The only problem is that if they can't get a foothold early, they may not be able to at all.
My other idea is to separate the temperature tolerances for germination and growth and allowing them to mutate separately. If they drift apart, these plants can germinate during seasons beyond their tolerance then grow when conditions are more favorable. However, they'll probably need shorter sapling stages and lifespans to capitalize on this in environments where this difference is big enough. They also risk their two tolerances drifting too far apart, making them have less viable environments due to there not being enough change.
At least, that's my theory.
@@strcmdrbookwyrm great theory. I think separating the seed’s evolutionary story from the plant itself is particularly interesting and the code could probably be re-used for things like chrysalis stages in insects!
@@EightLittleBears What about plants with less enviromental diversity get an overall improvement on many stats, but the more adaptable the plant becomes, it gets a gradually lower bonus until it may become an actual hinderance.
@@Lallunalapruna123 yeah I think something like that would probably work!
I've been watching for a while now, and this is amazing to see how far its come and where its going! I can't wait to see how the creatures adapt to this!
Thanks so much for watching!
Amazing video as always! I am excited to see what adaptations develop when the plants evolve alongside creatures.
I'm also excited for your shift in focus towards UA-cam content. Your videos are always educational, entertaining, and visually stunning.
@@Briscanator thanks so much!
I am curious how the build would be affected by different seed distribution methods. For example how some trees only spread seeds after a fire where nutrients the the plant life used get added back into the soil
@@zane_sadauskis I am hoping to get this kind of detail in there at some point!
Another really strong video! This is one of my favorite projects being documented right now. Keep up the great work and can't wait to see how creature dynamics are bidirectionally influenced with plant dynamics.
Thanks for comment! Glad you enjoyed it!
soil quality should be added as a factor
Agreed!
Could be calculated by the amount of available biomass died on the tile over time
Loving the new stuff and how much care you put into calculating each factor as thoroughly as you can reasonably expect. Very excited to see what's next!
Thanks so much! ❤️
I think this just became my favorite video of yours. Really cool deep dive on plants. Can't wait to see more!
Thanks! Glad you liked it!
great video, glad youtube recommended me it!
Glad to see you were recommended it! 😃
I love this sm, evolution is always interesting just to watch but your va really adds a lot. I do wish there was a Key next to the map for the biomes, i quickly forget what means what but that might just be a me problem lol
@@aaronforpie thanks! I’ll bear the key in mind for next time 😃
I loved the frustrated acorn when you started going overboard with describing how plants get energy. So cute
@@magnolia1253 haha I had fun making it
nice, new video on the simulation
@@Choompar glad you like them!
@@EightLittleBears quick question, do you remember me?
I do! Though I’ll admit I had to look back at what your comments were… sorry I only have a mediocre at best memory 😅. Natural disasters are still in the plan!
@@EightLittleBears nice, next add disease
An amazing video and simulation. I cant wait to see what interactions with creatures does to both the plants and the creatures themselves!
Thanks! And I agree creatures and plants evolving together is gonna be wild!
the plants evolve longer lifespans (and size) because the additional time to mature isn't enough to dissuade them. Throw in periodic fires (idk maybe once every few decades) that have a probability to kill each plant in the tile, with a bias so that larger and older plants have a higher chance to burn. That should give you enough selection pressure against huge ancient trees.
you'll also want to implement fire resistant seeds
That’s an interesting concept!
Plant evo! Excited to see this alongside the animal evolution sim. You might want to do some sort of layering option so you can have rainforests
Two things I think could be interesting, but probably wouldn't change much (They are kind linked too):
One, I feel like there should be a penalty for having too much above ground mass verses root mass, especially as a plant gets bigger. I don't think it should bother short plants much, but if a plant gets taller it should need more root mass to keep it from falling over. It's a bit of a balancing act though, as if it's too severe it will stop tall plants from forming.
Two, I think that the amount of water in the ground should effect how hard it is for plants to grow their roots. My idea is that it costs plants more energy for plants to grow bigger roots in areas where there's less water because the soil is harder. It could also be linked for how long it takes a plant to grow, with slower speeds being effected by this penalty less. I think it would lead to more plant diversity based on environment, but at the same time I'm not sure if it would make much of a difference overall.
Both interesting ideas. To some extent, larger plants already require larger roots because, without them, they would struggle to collect enough water, but I haven’t thought about it from the “falling over” perspective. That might be even more interesting when large animals are added to the equation!
Suggestion: Animals that help spread plant seeds.
@@Average_NerdIII next phase is adding animals back in and I’m gonna try and implement this 😀
Your making a game like oxygen not included
Fungu could be an addition, either feeding off the dead or helping out plants
@@shilo-j5v agreed!
make some plant that can grow on shallow water. It would make more diverse plants. BTW, I love your content. Keep it up.
Thanks so much! And yeah I am going to try and tackle some aquatic themes after merging creatures with plants.
A bit late, but I want to try and give my 2 cents about why lifespan increased so drastically despite the lower time to reproduce. I think it boild down to 2 factors:
1. Resources being too bountiful and
2. Inability to compete.
Point 1 refers to that fact that it doesn't seem like any resource other than space actually depletes. Water is infinite (although the amount differs based on the tile), CO2 never decreases, and soil quality/nutrients aren't a factor in this simulation. On the last idea, actually, I do think they should be included since, although the process of photosynthesis does lead to the creation of carbohydrates, it doesn't lead to ATP; there's one more process inbetween where molecules break and recombine where ATP is actually produced (I think 5-6 times more than in photosynthesis), and this cycle as it's called uses other nutrients that you can only obtain from the soil, like Magnesium or Nitrogen etc. Having a soil quality factor that slowly depletes over time as it's used but replenishes once dead plants "decompose" (could just be once they die) would inherently lead to a lot more diversity. You should look into succession for even more factors/how soil quality could be balanced!
Back on topic of infinite resources though, since there's nothing to compete for, there's very little competition between individuals, meaning that there's nothing to actually kill off the plants once the simulation more or less hits equilibrium. It's like humans in real life: we have so many easily accessible resources and medications now that there's little to no reason to compete, so our lifespans continue to increase.
Now onto Point 2. I might be wrong since you did mention you gave the plants ways to compete, but since you weren't very clear on what they were, I can only assume those ways to compete are root size and mass. The issue is, however, like I said above: there's no reason to compete over water because, as far as I can tell, there's no limit to it. The only actual resource they're competing for is space. And here is where things get interesting. Since, as far as I can tell, there's no way for plants to kill each other, that means that they are incapable of competing for space once that space is already occupied. Since being overfilled on a tile is damaging to ALL plants on that tile, they try to avoid it because there's no incentive. But if you are incapable of competing for space once that space is taken, how do you get the space?
Simple: you never let go of that space. By growing to live longer lives, the plants are effectively killing any and all other plants before they can even germinate by occupying the space for longer and longer. All a plant needs is 1 descendant that will keep passing down its genes, so the slow reproduction rate isn't an issue here. And I think those "random trees" are the culmination of this process: the seedling stage takes so long because you don't want other plants to take it once you die, and the adult stage is so short because you want the best chance for your seeds to take your space once you die. And then the cycle repeats itself.
There's 3 ways you could solve this issue:
1. Make more finite resources (like water, CO2, and soil quality/nutrients)
2. Add more ways for plants to compete (maybe some release chemicals that kills other plants around them, or they strangle and sap the nutrients from a different plant etc.)
3. Add more ways for plants to die or factors to worry about (like wind resistance, soil hardness/softness, rain and intense weather etc.)
The Sapling is a really in-depth evolution sim with very advanced plant evolution. Your simulation doesn't have to have all of these factors at once, of course, and I'd argue looking at only a few factors at a time can be even more interesting than 30 factors you can hardly keep track of at times, but if you want to see how others have dealth with plant evolution, this is the game I'd highly recommend!
@@ceawithac4095 wow thanks for the detailed comment! So, water is finite on the tiles but it replenishes (around 10% of total tile capacity) every turn. Bigger roots allow plants to take a bigger share of the available water every turn. That said, I think that, because so many of the plant became small and grew very slowly, the amount of water on tiles was easily enough to sustain them so it was sort of “infinite adjacent” lol.
Solar energy is also kind of limited in a different way. There is a max amount plants can get from a given tile based on their overground mass (and indirectly through other stats) but this doesn’t deplete at all. That said plants compete through height and overground mass because bigger/taller plants have better access to light in a congested environment.
Soil/nutrient quality and atmospheric effects are all things I want to look into! In terms of “more ways to die”, there should be some of this in the next update when animals come into the equation (and natural disasters at some point).
Also, I’ve not played the Sapling but I’ve been watching the videos for years and love them!
@@ceawithac4095 also please feel free to join the discord if you haven’t already 😀! Great place to foster these kinds of ideas!
Maybe this is overkill, but you could create 2 levels of underground water (in addition with the surface water, so a total of 3 levels) and only the plants with the biggest roots can access the deeper level?
At the same time, maximum root depth should follow a ratio with respect to plant height, to allow trees to access the deepest waters and coexist with shrubs without competing to the death
@@carlosroura7581 yeah I agree!
I think you should make fruits that fruit eating animals can transport further. You should also make nuts that specialist animals have better access to. Maybe even an arms race where the plant makes harder nuts and the animals has harder teeth for example. You should make arboreal animals that get tree resources like fruits leave and nuts faster. Also smaller herbivores might favor small plants and larger ones large plants. You could even make a smaller animal that needs very few resources and can sustain a population on a single tile like incects. You could also do a flying animal. It could be like those that favor fruits in trees become arboreal, then if they become even more arboreal they become a parrot or some bird. Then maybe that opens a whole new pathway where they can pre predators. Also wetlands would be cool. Like wetland plants and animals to live there.
All good ideas! I think in the next phase where I bring animals and plants together I will be trying to implement some similar things (or at least seeds that animals can transport around and different seeds/nuts etc. thanks for the comment!
2:01 i hate it when people do this, because i was listening to everything they said beforehand and they just handwave it away like not a single soul cared, i came here to learn god dammit!
@@EnderCats8 lol sorry! Maybe one day I can simulate the actual chemistry! 😀
Nice love your videos. Something that most genetic simulations don't consider, is that our DNA evolves to be redundant as a way to be more resistant to mutation. I'm curious how a mutation resistance gene, would affect the simulation.
@@EbonyWolf. that is an interesting idea and a few people have mentioned similar things. I can set mutation levels manually so it would probably be quite easy to allow “mutation chance” and “mutation amount” to mutate itself (and lowering would be roughly equivalent to developing a resistance to mutation). I’m probably going to do a specific video on that!
Glad to see you back
@@Misto_deVito6009 good to be back! 😀
Longer lifespan means more time between reproductions, meaning less mutations per unit of time. In real life less specialized r-strategists are often the ones to first swoop in after a big dieoff, being faster to mutate to fill newly-open niches than the longer-lived K-strategists and becoming the ancestors of dominant taxa. Perhaps in a more dynamic environment, where conditions change and more factors are in play, meaning there's more adaptation to do, faster mutations will give an edge?
@@ozirmokion7358 that’s a great observation, and actually the desert specialists were in an area where temperature fluctuations were severe across seasons, so that is one piece of evidence which supports your theory!
First the Bibites and now this? We're getting spoiled!
just seen that the sapling has one coming out tonight as well!
I had a thought about using the world model of earth as a control value, same Continental locations, same regional environmental factors such as heat and cold, how much water, how much it rains and more, even similar solar rotations around sun that makes up year is similar, idea is to see how close you can get to earths ecosystem while do similar natural random state of growth. Maybe even environmental disasters such as hurricanes, Earth quakes and volcano eruptions and etc?
@@demian_csomic_winters9484 it’s an interesting concept. Think I still have quite a way to go before I could attempt something like that! That said atmosphere and natural disasters and so on are all things I want to add at some stage (essentially an evolving world rather than just evolving things in a world). Thanks for the comment, and please feel free to join the discord! 😀
@@EightLittleBears for now could make flat map of earth and general idea for different regions and so on, like where median line and artic poles are, more complex stuff like spherical world, disasters and so could come later, something like this could probably make in similar manner to game making but with more sandbox style, and you could "update" or make "dlcs" to alter your scenario game of sorts. Like making a program where you can edit the solar system such as how many stars, what type of stars, how many and types planets how do they all move and so on, down to smallest details though how much depth you can control can be improved over time, and as its more of tool then genuine game balance for players isn't really required to worry about as for testing but who knows maybe in future others could use such useful tool too.
Even adding ability to design both real and fake plans and test them in scenarios, like what if world tree of elf's as really and had genuine capabilities as they are shown what would real world look like with such a tree and with right settings and programming you can see how they could naturally work in a IRL world if real.
And the great thing about this idea is you have thousands of real plants and variants you could model for templates even long dead Species of plans and animals could be reacted in this manner.
this comment was meant to be short then I kept thinking and typing my thoughts I had them. I hope it's clean enough to read and understand lol
So now begs the question do you plan on adding aquatic plants to the build?
@@zane_sadauskis haha - this made me smile.
On a realistic note, hopefully next year, but the next goal is to merge the animal and plant systems together so it depends how long that takes, along with a couple of environmental changes to get lakes and stuff.
@EightLittleBears Awesome, I love seeing the progress you are making for a more comprehensive simulation. Really amazing stuff you are doing!
I think you're missing some of the competitive advantages large plants has and that means we don't really see the same progression we see in the real world. In the real world large plants have significant advantages that let them outcompete small ones like blocking out sunlight (which you don't really accurately model since you you just have a congestion penality that applies to all plants but in the real world large plants would be completely unaffected by the light the they block from other plants), they are also able to produce much larger seeds that thus have more energy and better chances to germinate and they can produce many more seeds, plus they time their seed dispersal so some years they produce a massive abundance of seeds to ensure they don't all get eaten by animals. Also another aspect that might need simulating is the light spectrum because one of the big differences between tall plants and underbrush is what wavelengths of light they absorb, tall plants will absorb the most energetic parts of the visible spectrum but underbrush will focus more on efficiently absorbing the parts that haven't been absorbed by the large plants, which is why their leaves are different shades of green. And if you're gonna introduce animals you should probably also simulate seed dispersal so there's some opportunity for mutualistic symbiosis between plants and animals by letting plants evolve animal based seed dispersal.
Also if you worry that this would make large plants too competitive I could suggest also implementing disease, since large plants are just naturally more vulnerable to disease due to their long life span.
@@hedgehog3180 thanks for the detailed comment! Some of these features (like speed dispersal) are hopefully going to be in the next update. Also, the light spectrum was particularly interesting as that isn’t something I’d heard before!
@@EightLittleBears It's one of the reasons why you don't see grass in forrests since grass is adapted to use the same wavelengths of light as trees, another is that grass can't develop its complex root system in a forrest since the underground is already taken up by the roots of trees. On the topic of roots another thing that might be worth modelling is the density of nutrients in the soil since that's one of the most important abiotic factors for plants. You can model nutrient levels in a pretty similar way to how you modelled ground water, but have the nutrients get washed out by overflowing water, and in turn the rate at which it gets washed out is based on how crowded the underground is since roots stabilize soil and prevent run-off in the real world, especially large roots like the ones trees have. Of course if you combine this simulation with your animal simulation it'd make sense to make the nutrient replenishment scale in accordance with how many animals a tile has.
If you would add atmosfear or carbon silicate cycle you could simulet global extinctions. For exapmle when a plant gets to got at it's job and spreads into regions where are no animals to eat the dead biomass. This will make a thiner CO2 layer, because all the carbon is now in a different form, and this will cause a global climate change an ice age. But afrte a while the CO2 layer will be restored, because the cycle is self-balancing.
What a cool concept! An atmosphere is in the plan, but I haven't considered how to actually implement it yet..
i always make sure to watch the vid all the way through
@@vinestick amazing! Thank you! Supposedly it’s what the algorithm cares most about!
@@EightLittleBears of course, hopefully your channel picks up traction
Hm. Naturally evolving plants will learn to basically cover the earth even in harsh conditions? Not entirely true. However also not that far off. Even arric and desert conditions have plants that have evolved to exist at certain times. Of course our earth does have some hellscapes in which nothing lives for long. However life gets closer than one might assume. Maybe some of the conditions need to be more extreme, to make the simulation more accurate.
@@jarjab2games agreed. I think the extremes will be more apparent in a bigger world, but I need to do some code optimisation for that!
Good stuff.
Hope you're having a good day, cus you made mine. Hahah!
Good luck.
@@snowmanthegamer4983 thanks so much!
I need water next
now what will evolve faster digital panda or digital bamboo
Maybe they will evolve in perfect harmony?
Can I ask what you used to make this? I’m building my own evo sim in Unity!
@@AmyClerk996 awesome! I also use Unity
cool anyways last week
Can plants evolve to consume less water or sun?
@@Lallunalapruna123 yeah, being smaller and reducing rate of growth will lower a plant’s water and energy consumption
Comment
@@nightmarereaper5601 👊