Еvolution simulation, algorithms, swarm intelligence, neural networks, AI.... foo52ru in English. You can support the project on Patreon: www.patreon.com/SimulifeHub
If anyone is experimenting with this & has access to a classroom, hav3 each student design an organism with several attributes & rules of behavior… Attributes: fat reserves, quickness, likelihood of mutation, ability to recognize food & danger, if it senses that it’s being eaten, release a spore containing information about The organism that’s attacking it… + ?
I'm making a simulation like this too. I was originally inspired by conways game of life, Particle Life and Life Engine. Yours is the one that most closely resembles what I'm trying to build, though. I would really appreciate more insight into how you built this, wich programming language you used and such. Do you have a Discord or somehwere where I can learn more in depth?
Love it! Sexual reproduction might slow the genetic drift and allow for more experimental genes, would be super interesting how these worlds would look like.
A lot of these suffer from seed piling… it would be interesting to introduce a 'rot' concept that removes seeds not utilized in two growth periods eg. or even better, if part of their energy could be passed instead to the seed above? 🤓🧐
What youre missing are Jurisdictions. All sprouts of the same jurisdiction follow the same rules. These rules are toggles like 'sprouts cannot kill other sprouts' or 'sprouts can not take food if a sprout with more need for food is in the area' or 'sprouts can only crossover with min/max generations apart'
Would it be possible to figure out the few most prominent species of a given ecosystem by clustering the genes of all organisms and averaging out the resulting clusters?
I think adding some consistent cyclicality into the environment could go a long way in evolving organisms that are more adaptable to instability. Like day/night cycles or seasons or some equivalent of tides.
Omg finally a simulation where both the agents and the environment change. So many creators keep their agents inside a static environment then wonder why nothing interesting happens or get diminishing returns on complexity. This is so fascinating and refreshing.
Here's an idea: Destructive cells, which resist and absorb excesses of energy and organic matter and cause nearby cells to drain more energy, but are themselves highly energy intensive and self-destruct over time, thus giving a way for organisms that prefer energy poor environments to create one for themselves. Also arguably we humans are causing a new mass extinction and rebuilding of ecosystems...
these simulation reveal so many critical and facinating concepts that we see in our nature. The way single cellular life becomes predominant but then when stability has emerged multicellular organisms naturally evolves and takes over. How we can see circular cell like structures at 13:45 and a structure that grows in a similar way to crystals or mold branches at the same time age. I find it facinating how there is so much that is unknown about nature, that cannot be described with a single math equation, but when plotting simple rules like this over millions of steps eventually a nature like system seems to always emerge. In a sense nature in real life is also a set of rules played out over time.
Honestly I'm amazed by how often you get stable ecosystems. I've seen lots of other attempts at evolution simulations that always eventually collapse no matter what you do. The only other ones I can think of don't because the organisms have pre-programmed ai made by the developer, which limits their behavioral complexity as a consequence
Great video, that simulation is crazy. Could you please post the genome pool accross the generation, would love to make a probabilistic phylogenetic graph out of it.
I first found this channel when the 3rd tree evolution video was posted but I couldn’t stand the ai voice so I never subscribed. 1 year later this shows up in my recommended, and the ai voice barely bothers me at all. The way short-form content has changed me 😢
Blue cells can extract energy from the soil. Only red cells can extract organic matter. They then convert it into energy. This was done to introduce diversity in the structure of organisms. Perhaps not very successful. We need to come up with some kind of naive “chemistry”. Combining different chemical elements gives different possibilities.
The source code is available to paid subscribers on Patreon. The project is written in Processing. Once you install this environment, you can run the simulation on your computer. But it works very slowly.
