Evolution Simulator (Part 4/4)

[@luckyoma0]

Creature: OH NO A PREDATOR TIME TO RUN! Creature: *runs* Creature: *accidently runs backwards into the predator and dies*

[@disrespectthemwomensubjuga5471]

Lmao 😂😂😂

[@kristianskarklins1245]

Bro! Literally before i liked the comment,the like count for ur comment was 666. :D coincidence

[@yellowdude6311]

Glad to be your 1000th like...

[@lololoshka-zh2gz]

no need to worry, predator runs backwards too

[@TheGreatMarbler]

YellowFellow1234 Hi

[@moth.monster]

Don't you hate it when you turn inside out and start going backwards?

[@MitruMesre]

+Kinkzoz 2:23 for anyone who wants to watch it again (I did)

[@brassicac]

#relateble

[@jackrandy3494]

alan null. coolbeans

[@pensive955]

Lyndon Choong #sotru

[@MegaMGstudios]

post #876218 #relateble

[@SolarFlairYT]

"Tringle"

[@willbeard5196]

Solar Flair ugh

[@wildewudubearn1028]

Solar Flair I

[@glenthemann]

lmao scrolled down to post this

[@BananaSlayer_]

Tringles = Pringles.

[@Danube-TV]

Solar Flair triangles!

[@arch758]

i find it interesting how the worst is the same as the best. only facing the wrong way. deep.

[@ioulios12]

Damn :O

[@Ryanthebrobdingnagian]

You'd expect that With randomness.

[@RiceBoy7thst]

there is a fine line between crazy and genius

[@TheDslide]

xD true that's deep

[@Llortnerof]

@@phoenixrio9273 This. At high generations, the bad ones are basically those with birth defects, which is randomly generated. So their numbers wildly fluctuate.

[@oooChickenatorXooo]

Do an asteroid. Wipe out 90-95% of all creatures randomly every once in a (long) while. compare generation 10,000 with an asteroids with generation 10,000 with no asteroids.

[@Kunumbah1]

Honestly this is essentially what happened in the great dying. You had a dominant groups of organism on earth (Proto-mammals) who died off which allowed the dinosaurs which were a much smaller group to thrive. Using this logic the S46 might all die off and it'll give the S33 (Triangles) a chance to become the new dominant species.

[@rightpurpose2076]

On steroids??? Lololol jk

[@Kalbr0shorts]

YES, BRILLIANT

[@butthead1535]

@@Kunumbah1 when Dinosaurs went extinct, mammals and birds did the the same thing as well

[@Noname-67]

Not only just asteroid but an environment shift, otherwise nothing would change much

[@villanelo1987]

Being honest, I think I was more attached to a group of dots and lines than I have been to any character developed by Bioware in the last decade. :/

[@rockagold3819]

When I played this I had like a whole mini backstory of how species began taking control and falling like mini empires and I started rooting for some and others and cried so much and praise the doritos and tetrahedrons oh god what is happening

[@Foodude]

villanelo1987 That's funny because I was just watching Mass Effect Andromeda cringe compilations before these 4 videos

[@villanelo1987]

@Foodude Watches That would be funny, if it wasn't so sad. =(

[@castiel7330]

villanelo1987 comment of the year

[@skydaz3r]

Right?

[@britjohnson4478]

i ran this for 32 generations, one actually evolved a wheel-type structure

[@triplea657aaa]

That's amazing

[@colinjava8447]

@@triplea657aaa you don't see it in animals for some reason, but it would be an efficient way to move. But I suppose the wheel would have to sort of separate to the body cause veins would get stretched by the rotation. So in the animal kingdom you would need something like a bone that can freely rotate

[@jamessinclair2898]

@@colinjava8447 Prehaps some kind of semi-symbiotic relationship? Like how male Anglerfish merge with the larger females to give her both sets of genatalia?

[@Derpynewb]

@@colinjava8447 you dont see it in animals because physcis. maybe if a bone acted as a crankshaft you could have rotation. But then there are issues there. genetic mutations. There would need to be tight tolerances and the muscles would proabably pinch the fuck out of themselves and harm themselves. The design is not simple enough to be reproduced efficiently. Unlike actual crankshafts in an engine, we cant build things naturally with tight tolerances and safety margins. forget quality control lol.

[@upsydaysy3042]

Wheels are only an advantage if you can build and maintain smooth roads. That's one of the reasons why animals don't have wheels. Then there is the problem of developing a wheel that must be separated from your body.

[@mickeymoose636]

Amazing how in 300 generations, a randomly generated group now looks almost identical.

[@magnusanderson6681]

yup, I'm assuming the better behemoths are just too unlikely to happen (I mean come on tho, imagine a s9_13, that's the same except has four legs in conjunction that work like this but with constant power. It would be slightly better.

[@LukeOrionMarble]

+Magnus Anderson that would be called an S93

[@magnusanderson6681]

Yup, just make 1 environment where they have small bumps on the ground and make it so going right and total time spent on the ground are both factors, that count equally (the worst at one and best at another gets put at .5, so add percentage better and divide by 2) and youll end up with slow, but steady ones along with doritos and i dont know what in the bumpy environment.

[@magnusanderson6681]

And maybe the bumps get bigger over time while, the flat one slowly turns downhill, and speed starts out more important but becomes less important over time.

[@docdaneeka3424]

That's not such an unusual dynamic, lots of processes even 'apparently' complicated/random ones, will converge to a particular point or one of a set of points when you iterate them.

[@ThylineTheGay]

One thing i LOVE about cary, is that he keeps most of his projects open even after over FIVE YEARS!!

[@ezrakornfeld8436]

It’s 7 now

[@alvinalpha_seven5330]

bfdi

[@sissylei5955]

And guess what returned after a decade long hiatus?

[@ThylineTheGay]

@@sissylei5955 idk, what?

[@skoupidia2057]

>tetra becomes sentient >tetra makes a religion denying tetra evolution

[@5C2WMedia]

Skoupidia tetralution

[@Photosounder]

**tips tetrafedora**

[@rizzopizzo0127]

Yeah just like us :D

[@martinferrand4711]

@@rizzopizzo0127 you got it x) Bravo ! :D

[@osamahafez5476]

Was there not a person who created the simulation in order for the evolution to develop. Ur own logic has flaws.

[@ilikecats830]

I dont know why but I really like watching these

[@SomeFreakingCactus]

It's interesting while not being mentally challenging, and combined with the music, is very relaxing.

[@bingobongo131]

Kyle Hart that was surprisingly well explained o.o

[@littlellama400]

Kyle Hart huh, neat

[@TS111WASD]

It’s because you also like men

[@kriss3d]

I showed this to my little daughter. She was quite upset that he chose to kill 500 in each generation. She found the creatures to be cute.

[@sliver170]

It is simply impossible that a dorito would _transform_ into a tetrahedron ALL BY ITSELF. Just ask yourself, fellow tetras, have YOU ever witnessed a dorito transform into a tetrahedron? No! This is an insult to the TetraGod! We were all _created_ in the TetraGod's perfect image, and we were given dominion over all other inferior nodelings. PRAISE TETRAGOD

[@magnusanderson6681]

well you know what! once dorito's evolve a 1-turn greater lifespan, our masses will be able to destroy you out of sheer luck! Edited to actually make sense.

[@metachirality]

Best comment 2016

[@Aweseb64]

Tetrahedron (46) mutated from 45, gaining another muscle. The Dorito (33) went extinct a long time ago. (I find this so fascinating)

[@WilliW64]

Even the most simpler bacteria has so much complex parts that it's not possible they just appear. This simulator works well because the creatures are really simple, and the "kill" and "reproduce" procedures shown are determinant, they are perfect for the situation. It looks more like that the whole thing was developed for the creatures to evolve than they just randomly evolved.

[@sliver170]

WilliW64 Now let it run for a few billion years in a simulation with pentillions of particles (that's a grand understatement) and countless variables with many different forces and laws at work in a 3D planet sized environment (not a line of a few meters) where specimens can interact in countless ways with each other and the environment. Don't forget to add a 10^30 kg entropy sink in the sky. Oh, what massive, complex, and bizarre systems one would find in such a simulation.

[@feefeee]

What if you ran the simulation for a year 8 hours a day?

[@manospondylus]

They'll discover fire and build pyramids.

[@theoreticalknot4012]

And then a year later they'd begin to wonder if a superior race had created them in a simulation to see how they themselves evolved. And they'd be wrong- so hilariously wrong...

[@АсенДоцински]

Judging by the results later on - almost nothing.

[@tdobac55]

I think if there were more selection pressures and more room for variation something quite interesting would occur, but these would be limited by the computer's processing power

[@R3tr0v1ru5]

That's where the environmental changes come in. Without changing evolutionary pressures there comes a point where not much can be improved upon (as shown in this simulator).

[@ElTitoAlberto]

What happens if you re-run the whole process again from the beginning with exactly the same settings? Does the same creature appear at gen 300?

[@jjt171]

Yes/No since it's all random, theres a chance this creature's common ancestor/common ancestor mutation will never spawn in the beginning, and instead allow for a different, but equally as efficient creature. But, yes, this creature has great "traits" that allow it to move right. given enough time, this creature could emerge again

[@RiceBoy7thst]

@@jjt171 relative to it's "birth" environment.

[@Llortnerof]

@@jjt171 It might even result in a "better" creature. Or much worse. Or we might get convergent evolution and get the same end result from a different starting point.

[@GotYourWallet]

It depends how the randomness is programmed. It's kind of hard to get a random value from a computer so usually you use an arbitrary "seed value" like the current date/time and all the calculations involving chance will use this seed value as a starting point. If the seed value doesn't change then the same creature will appear at gen 300. This is known as a deterministic system.

[@Llortnerof]

@@GotYourWallet In one of the later videos it's discovered that the simulator is not deterministic.

[@mettatonex7221]

2:22 I like this one. He's taking the road less travelled by. Everybody else developed the fundamental understanding that right is good and left is bad, but this guy is brave enough to explorer the unknown. What lies on the other side of the diverged path?

[@TerraeChannel]

How about inducing a "biological crisis" by putting harsh stress on those little creatures? Like putting an obstacle at some point during the course. The bigger and sooner the obstacle, the greater the biological extinction?

[@odinlindeberg4624]

or you could make a bump in the ground that might just launch it into the air.

[@herbertbert3147]

There wouldn't be an extinktion because the deaths are allways 500 . This would change if you set a minimum in Range and than the 500 .

[@TerraeChannel]

Herbert Bert I misused the word extinction. I meant the disparition of one type of form and the emergence of new ones.

[@firesong7825]

If you don't make it so that the obstacles are randomized then the organisms will only be optimized for that specific obstacle and not many others. Removing that obstacle or changing it would make it very difficult for them. I'd rather see one that could teach it how to work with any type of terrain.

[@GarryDumblowski]

You could, in theory, create actual extinction by introducing two genera: one predatory, one prey, and have them pair up randomly. if the member of the prey genus catches up to the prey, it survives and the prey dies. But if the prey outruns the predator, the predator dies of starvation and the prey, unsurprisingly, survives. Then you could trace the evolution of both genera to see how a predatory relationship affects how they evolve, or even tweak the conditions to see which favors the predator and which favors the prey. Of course, each genera would have to run under slightly different rules from the other, otherwise you'd still get pretty much the same results as you did here.

[@nrxpaa8e6uml38]

"How would random mutations ever create such perfect tetrahedrons as we are? Just look at how smoothly we walk this earth! If you changed just a tiny bit, you are crippled! We are therefore *irreducibly complex*! You just want to believe in evolution because you don't want to accept TETRAGOD (Peace be upon him)!"

[@joefortescue9096]

All hail our secret Tetrahedron overlords

[@martijnbouman8874]

''You guys think that we came from simple Doritos! We are not animals!''

[@gabrielabrahao4383]

Praise Tetragod

[@logicreason3231]

Doritos Lives Matter

[@kamel3d]

nRXpAa8E6UML creatures on earth are more complex than the ones in this simulation, the fact you get somewhere from this algorithm don't prove that evolution is correct, because if you accept that god created all the animals and humans at once also is a good explanation for this world and yet you don't believe in it

[@unpronouncable2442]

I think what you call "Anchor puller 2" is a counter balance that allows the creature to synch up all its movement and stabilise itself Also have you ever thought about just leaving your simulator running for the night? (I assume that generating one generation of creatures takes a very short time)

[@williamcastro4440]

It also seems "Anchor puller 2" forces the anchor node to the ground earlier than if it wasn't present in the system. Without Anchor-puller-2 Anchor-puller-1 would pick the anchor off the ground and then, because of it's angle with the anchor, expand it backward only slightly in front of it's starting position. Anchor-puller-2 improves on this by forcing the anchor to touch the ground sooner and also farther from the starting position, allowing that more efficient movement.

[@RebelZhouYuWu]

imgur.com/a/wpqtk2D It doesn't really get much faster. As you can see from the chart, it hits a wall around generation 929 before it somehow managed to improve in generation 2474, and then spends the next 4000 generations at the new wall.

[@ruynobrega6918]

@@RebelZhouYuWu Hi, have you thought to increase mutation rate when the fitness curve stagnates? Just a random thought here.

[@rohansonkusare7025]

A few thousand generations later they will evolve into a civilization of stick figures.

[@camromantic]

It’s the Generation 8756... the Great War is happening. Racism for white nodes.. and population decrease for the tetra’s. HOW.. HOW DID THIS HAPPEN?

[@Zamolxes77]

the goal "move to the right as far as you can in 15 seconds" is not enough to evolve into stick figures.

[@devinslurry365]

@@Zamolxes77 We are already all stick figures. Some are more buffed due to less grease on the wheels and some are more bloated due to allocation of things and food inputs.

[@devinslurry365]

@@camromantic Everything is racist. You got the memo.

[@devinslurry365]

Stick figures will win on a long enough evolutionary timeline.

[@ludiwang1172]

This makes me feel nostalgic for the creatures of Gen. 1. I miss them! ...Wow, just realized how little of a social life I have.

[@OuterSpaceandTimelapses4]

lol 💀

[@RavenFlight413]

If you threw a limiter for food in there instead that'd be interesting, Like, instead of a timer, They have a limited amount of energy until they cant move anymore, and to get more, they have to reach certain distances or milestones that replenish the energy? The ones that run out of energy die off, the ones that can reach the next re-fuel/food point live on to go further, but the distance between the points for food get further and further apart?

[@Vista-hv6ej]

That's a great idea

[@anedge6706]

Fuck two years and nobody have done this.

[@matth7800]

This is just Hill Climb 🤷

[@Kunumbah1]

an edge Check out Primer on KZbin he's done something very similar.

[@MachineThatCreates]

I hear what your saying but it's not necessary. Their "genetic instruction" is motivating them just fine. Genetic Imperative is probably a better term.🌴

[@zoomi-]

"Oh my gosh, it's so homogenous, it looks like a wallpaper." My thoughts exactly.

[@ToonEugen]

Found you! :D I hear about you on somewhat BFDI fan video. So yeah..

[@CriticallyGnart]

it just seems to be a T R I N G L E

[@DanDAlittleMan]

Same

[@DanDAlittleMan]

The entire thing is just a triangle at gen 500

[@tonyazrael8223]

This is now my favorite anime.

[@mariafe7050]

This isn't an anime

[@HelloHello-vk5ob]

Tony Azrael same dude

[@Robiness]

@@mariafe7050 WHO ARE TO TO TELL WHAT IS OR IS NOT AN ANIME SIR?!!?!

[@yunicraft2698]

Attack on Evolution!

[@Aladato]

5:25 was so smooth and nice to watch lol

[@minefilms1122]

can you release this publicly please?

[@minefilms1122]

Pedro A >posted june 27 2015

[@HardcoreMontages]

+Aaron Gosselin Where has it been released?

[@huehuehuehuehuehuehuehuehuehue]

www.openprocessing.org/sketch/205807

[@plotylty]

kek u wot m9 HUEHUEHUEHUEHUEHUEHUEHUEHUEHUEHUEHUE txh

[@jackthepenguin3474]

+kek u wot m9 HUEHUEHUEHUEHUEHUE brain tumor

[@lauroneto3360]

If he continue to generation 1.000.000 he would get Trevor from GTA 4. Really, I've tried here.

[@spoopyduck4651]

Nah it would be big smoke

[@autumn7]

if you put a decimal point instead of a comma its still 1

[@spoopyduck4651]

The7SuperStars maybe that was the intention! :P

[@BaalTomekk]

@@autumn7 There are far more countries in the world using a decimal comma instead of a decimal point.

[@shakya00]

@@BaalTomekk Americans think that they are the center of the world. So...

[@QuantumBraced]

Lo and behold, when the selective pressure was to get something that can run as fast as possible, we essentially got something that looks like a cheetah, albeit much simpler, but equivalent when you consider the constraints of the simulation! Amazing! Great job! We really need something this in schools, so kids can see how evolution actually works in the real world as it's otherwise impossible to observe.

[@syborg64]

yeas I totally agree, it would so much cooler and intuitive. I wish schools didn't actively try to be boring

[@EmptyHouseGuy]

You're forgetting that a program was built specifically for the purpose of assembling these species in the first place, and a mechanism to introduce and implement change was artificially constructed. In other words, this demonstrates little more than natural selection, which results in less species existing at the end than at the beginning of the simulation. Note that no new species were actually created through this process, which actually results in a massive purge of existing species. Only slight configuration-based improvements were made to one of the pre-existing species, which basically means this produced no new species at all.

[@azentra7560]

You are missing 2 key points here. In this case, there was only one single objective and the conditions never changed, causing species to eventually be more efficient at it, which does indeed result in survival of the fittest. However, there were definitely new species introduced. Species S69 did not exist at first, but later died out, but there is also species S58, which came in later and did manage to survive, even though there were only a few left. And most important of all, the most effective species, S46, was actually introduced after the first generation. So yeah, your statement of no new species being produced isn't entirely true, even in this simple case...

[@dansegal4644]

so metaphorically, if you add a bunch of physics, let the creatures control their muscles independently, make your creatures millions of cells big, and extend the timer from 15 seconds to about... 5 billion years, metaphorically, the creatures would start a civilization. They will research technologies, build cities, build rockets, all in order to reach to the right as much as they can. And all you need for that is a super computer, which will probably be built in the near future, and someone very commited.

[@TheIlidius]

well technically we are only muscles on bones, controlled by a cluster of booleans. If we take out the energy consumption of the whole process we don't really need our other organs at all.

[@DavidSantiagoPoloSilveraBio]

The problem is that in the similuation, the only selective force is "run to the right the fastest you can". In the real world, the selective forces are much more complex than that.

[@deadpirateroberts9937]

​@@DavidSantiagoPoloSilveraBio You dont need to make a selective force at all. Just make a bunch of laws governing this simulation/universe and watch as the atoms interact and form complex things like galaxies and creatures using those laws, such as adding a speed limit of 299 792 458 m / s.

[@devinslurry365]

Metaphorically was the word I was looking for. Thank you. You negated to mention farming and water allocation in warring tribes when they did not in fight and extra-fight as per the old testament.

[@devinslurry365]

We would just devolve into Apple colored Jello.

[@alexjones3035]

This is incredible. I tried it on my computer, but rather than eventually getting a triangle that uses those bands to lift the black wheel off the ground, my (best) creatures evolved to have a sort of wheel in the front, composed of a high and low friction ball. The way it works is that the muscle on top expands and the bottom one contracts, pushing the low friction ball in the front into the air; then the top muscle contracts and the bottom one expands, which quickly spin the two balls in front around and kicks off the black one in the back. It's a more complex design than the one in these videos, and much less efficient; my top creature only managed around 14m. Still, it's fascinating how purely random chance still managed to optimize. I'll run it through a few more times to 100, and see what happens. Thank you so much for making your code available!!!

[@TheSteve94]

S46 works better than S33 because the extra node shifted the creature's center of gravity towards the front of the creature instead of the center, which aids the same forward motion required for movement like you explained for the creature with 2 nodes. What also happened in S46 is that the additional muscles constrict the anchor node's freedom of movement, causing the anchor node to become a definite limb, while the other nodes become a definite body. Tl;dr, the additional node moves the frictionless node forward more consistently, requiring the creature to recover less ground after each muscle contraction.

[@gullijons9135]

@Kernels Sprinters actually stand up straight once they've reached maximum speed and kick their feet forward instead of backward. Try looking at a video of 100m sprinters taken from the side during the first 30-40m, that's when the shift happens.

[@amazinc754]

the best one I've gotten in my testing was S59. It was similar to S46 but had a few extra nodes in the center and high tension connections which made it even faster

[@RandomTermite426]

Wait, that floppy straight line that goes nowhere is still alive?

[@jayit6851]

ITS A SNEK

[@Stubbari]

Basically that's better than those who go backwards..

[@GrunOne]

It might have been a mutation just in that generation that lost a bunch of parts

[@Lol-fo2zq]

Randomtermite426 Because there are always creatures that go backwards

[@hellohowareyou9745]

Randomtermite426 no it's a random mutation

[@loganrussell48]

I think one reason the tetrahedron is better than the "dorito" is because of your scoring. The fact that the average distance of nodes is used causes the tetrahedron to have a higher average even if the 3 corners are in the same place. Also, the node in the middle of the tetrahedron exerts a more rightward(which is good) force on the anchor, rather than more upward(like the top node). The farther rightward the anchor is pulled between its pushes, the farther it will push the right corner node.

[@-yttrium-1187]

Yeah, the middle node seems to cause a more smooth circular motion of the friction node.

[@Kasokz]

you also mustn't forget, that it has to have a controlled forward pull, otherwise it will flip the anchor side and starts to move backwards. the middle node is probably there to influence the average pull so it can move with the highest viable distance.

[@Markd315]

Logan Russell Hey you went to my high school!

[@yanxiangmeng4967]

It seems the evolution is nearly coming to an equilibrium by 300 generations. Next time you should try changing the selection pressure and observe how different species can evolve. For example, instead of flat ground, make the ground at a slope and make them climb upwards.

[@norealmagicever3292]

the lil triangle is really cute, actually. im proud of him

[@FooBarBash]

Whoa, this is just incredible. Such a cool, simplistic model for evolution. Really interesting!

[@anynamewilldo7776]

I think you should stop, if you keep it going in this manner it will eventually become sentient. We must not tempt evolution, lest we suffer like the Quarians.

[@GalaxyDominoes46]

Terncote lip

[@WilTheNyanCharmander]

ALL HAIL THE TRINGLE

[@azuregriffin1116]

Any Name Will Do is that Elfen Lied in your profile pic?

[@reginalddelafontane4428]

Fuck that shit when you have a kid do you say "oh fuck we just made a potentially better version of ourselves KILL IT WITH FIRE" no you don't so if we made a sentient being that could over throw us... Fuck it we win

[@mete3254]

He needs to change the context and tasks. Just moving forward is too limited. Imagine tasks like, climbing, running on an unstable surface, escaping or catching something, with minimum movement. there is an Incredible potential here.

[@teh201d]

this is the best reality show Ive ever seen.

[@gumarks_]

This must've taken a lot of work. Amazing job!! Totally loved it, it's really interesting.

[@wolvie90]

This was weirdly elegant and beautiful. Now I'm imagining a planet full of spastic ball-and-muscle creatures, lets call them Bams.

[@ethanwagner6418]

It's like after millions of generations, you get a creature that literally walks like an animal.

[@potato-hj9nm]

Sadly that doesn't happen in this because the task is so simple.

[@ethanwagner6418]

Yeah

[@Midaspl]

TBH it looks like running animal a bit. Compare this movement to a running rabbit.

[@JaceITG]

In this simulation there is no need for a more complex creature than the final one seen here. If you had a random or varied terrain then it would be possible to get an animal-like creature

[@Stubbari]

Random terrain and much more simulations.

[@cirvis240]

One of the coolest series i've seen on youtube! :D

[@beastbum]

What's it like with uneven/random terrain?

[@Toukan1]

It's mostly just a guess but I suppose it would evolve much slower, but eventually get it (though it may be after several thousands of generations).

[@beastbum]

I would have thought that, with unpredictable terrain, multi-purpose creatures would eventually evolve, with multiple lifting legs becoming advantageous. Therefore, cooler and more wacky stuff would evolve. But you're right, it will take much longer to evolve.

[@RandomTermite426]

They would probably become 2 tetrahedrons joined together that work like legs.

[@redstonewatcher6651]

2:22 instinct: run from the predator else you die Mutant S46: THIS IS SPARTA

[@Blubb5000]

THIS is how evolution works. The bad die, the rest mutates and tries to survive, the best live on.

[@annalihuang1601]

Is it meant to say 'the next one is coming in less than 24 hours in the description, cuz i thought this was last one

[@annalihuang1601]

I mean 48 hours

[@carykh]

Anna Li Huang Oops, you are right. There is no next part... for now! Also, we have the same last name \('O')/

[@shatterdpixel]

carykh what is she is your secret sister :O

[@kkirschkk]

carykh could you keep updateing this pls!

[@shatterdpixel]

you mean triplet?

[@StrydarXtheXGrim]

What a great little experiment. I mean it's not a game but a smashing little program and it could be the basis of something truly awesome. I hope you keep this kinda thing up. You don't see creativity on this level very much outside of academia

[@Bodgie7878]

Strydar SMASHING

[@R4V3-0N]

I sorta want to manually create my own creature and to see if it'll fail or succeed against the rest.

[@whatisthis2809]

5:24 whoaaa smooth

[@NinjaMamut]

I strongly identify to the ones going backward.

[@cristig243]

Or maybe the champions go in the wrong direction to begin with :)

[@ufzletryk285]

cristi g no he said they r supposed to be going right watch the first part before you watch the last

[@cristig243]

@@ufzletryk285 I was aiming very high, to the champions of human society, world leaders. Thank you anyway :)

[@ufzletryk285]

cristi g your still wrong tho either way it doesn’t matter

[@bryanchu5379]

I know this is asking a lot but what if these creatures all existed in a 2d space and had to compete for resources?

[@StrydarXtheXGrim]

Bryan Chu a good question but I'm assuming that this simulation wouldn't be adequate for that. You'd need a lot more parts than just nodes and muscles for something like that. I guess this is more a utopian world where all that governs who stays and who goes is which genes are best for movement alone and hunting and gathering are not even a concept

[@AvocadosDiaboli]

Its already in the the game The creatures Wich are the fastest stand for the one wich is going to get the Best Chances to survive and reproduce just like they hat to compete

[@galesx95]

Simon, that's not what he meant, he meant what would they look like graphically, and seeing in the end how the different parts evolved and why they were the more efficient.

[@odinlindeberg4624]

I second that.

[@kaiufkdlsmf]

I have a not-dissimilar idea that might be easy enough to implement on this sim: Every muscle costs energy to complete a cycle of movement (expand and contract) The energy is proportional to the strength of the muscle Every creature has a finite amount of energy Energy can be replenished by collecting blocks of 'food' Increase the time limit Place two creatures at opposite ends of a map, of size say, 500 meters Place food on the first 250m of the map, they should be placed further and further apart. Copy the pattern with a line of symmetry at the 250m point The 250m point is the finish line Let the creatures loose upon the world If a creature runs out of energy, it dies The winner is decided based on a function of which expends the least energy and how far it manages to travel in x amount of time Please feel free to expand upon this idea if you'd like

[@zachburke8906]

I believe this system has a fatal flaw. Instead of time, energy should the limiter. Because an animal might spam a useless muscle but will feel no negativity.

[@Geckuno]

I wouldn't call it a flaw. It is the way the coder chose to program his world.

[@bjarke7886]

Not a flav but it wouls be an improvent

[@Santiago_Nyczka]

Why choose? Both time and energy could be limited.

[@Cd5ssmffan]

you can simply replace the word "time" with "energy" however what about those that end up never moving (first few generations)?

[@Photosounder]

It's not a flaw, you chose your selection pressures. He selected for speed just like cheetahs select for speed in gazelles.

[@J7Handle]

I ran a simulation up to 130 generations and s33 became a hypermajority with average 10 m distance. I then replaced the value of SEED (line 2 in the code) with 321 instead of 0 and got a two legged s45 race that retook the hypermajority after about 50 early generations of an s33 monopoly. Think of the s46 final monopoly shown in this video, but with an s45 than has two pushers being cycled with one lifter and one front post. We saw this in one of the early generations in this video but its problem that caused its extinction in this series and almost in the 321 seed is the critical struggle with getting it right and mutations in the s45. With seed 321 the s45s had the best creature with the two legged version as early as gen 13 with a gap and then maintained best creature from gen 31 onward, yet lost more and more numbers to s33 until the 60th gen. This is very fun.

[@UltimateDuck97]

I think the tetrahedron is better than the tringle because the anchor puller 2 is pulling the anchor to the right, gaining horizontal distance as well as height. And so when the anchor is back on the ground it is slightly further forward than before and the creature is now pushing from a point further than before, gaining extra distance. Btw great video, really awesome software and a great visual demonstration of basic evolution.

[@daniellugosi]

Great video! I would have loved to quickly see another evolution process starting from 0, to see if the same general shape is found. This looks quite optimal, but it would still be interesting i think, for comparisons sake.

[@godoverlordquacken4003]

Triangles would take over

[@IferMasterofFire]

Pretty cool! Next you should see which creatures are more adaptive to different terrain - slopes, viscous air (weakening all muscle strength)

[@Ethernet3]

I think there is still room for improvement. If the creature at 6:40 would have had 2 black nodes on the left, those could work as a wheel.

[@MattiaPiola]

This was... amazing! And inspiring! And enlightening! Thank you :)

[@karldilkington8587]

Why not leave this program running for days/months/years? It would be interesting to see the result of a 10 year long trial and error evolution.

[@zionhayes3030]

Until you check it and find out the computer died 30 minutes in

[@squidsword0]

Diminishing returns. It looks like the entire evolutionary process follows a logistic curve. If you run it for 3000 generations instead of 300, you would arguably not see as much change from the from 0 generations to 300 generations.

[@cletusawreetus-awrightus2799]

Well our fundamental locomotion hasn't really changed at all in the last few million years, so...

[@darkduckpl9620]

Untill battery dies again and the world ends

[@marcodiamanti202]

Thanks for a stimulating half an hour with my sons. Peace out!

[@tauheximus]

There are 3 types of commenters on this video: Those who just like the simulator, those who want the simulator to be released, and those who are begging for BFDIA 6.

[@DownwardsRising]

Really interesting simulation! But I bet intelligent design could do way better at making a fast creature ;) Look at #889 at 6:31, if the two nodes at the back were black, it would essentially have a two driving pistons going on back there that would make it faster than a single piston by being able to reset one piston while the other still drives the creature forward i.e. "bipedal motion", as some have been wondering about. Imagine what quadripedal motion could do! That being said, it's possible that 889 could have been randomly generated as a double piston anyway. Also, I noticed that the simulation became dominated by a 4-node species with a plateau speed of just over 20m in 15s. A creature bigger than 4-nodes needs exponentially more iterations to evolve because it has so much more nodes, muscles, and the interactions between all of them to tinker with. The simpler creatures don't allow the more complex creatures the room to breathe, and so the bigger creatures of the likes of s57 became extinct in part two, and never had the opportunity to develop their potential because they were drowned out by a plethora of cheap-and-easy 4-node creatures that had perfected the art of being cheap-and-easy while the "mega-creatures" were still in their experimental phase.

[@magstau4177]

evolution can do this too, but it will take a lot of time.

[@alexandermcmiller6175]

I noticed that S58 appeared and went extinct several times lasting only a few generations each time.

[@samueleproiettimicozzi8134]

"You are a king, not a warrior. While the other heroic spirits mastered their Noble Phantasm, you've mastered none!" ok sorry for this

[@devinslurry365]

The best creatures follow Benoit Mandelbrot . "the art of roughness" of physical phenomena and "the uncontrolled element in life."

[@devinslurry365]

This is another criticism that is relevant. There are more ways to criticize this than there are ways to code it. I hope the OG reads the posts.

[@MandolinSunrise]

Nice one, and well explained, thanks! I guess if you automate the system and run it for thousands of generations then all of the creatures would end up almost identical. It would be interesting if they made you change their environment for them -you could domesticate them :)

[@handupandu6000]

This is one of my favorite projects of yours lol I tried this myself and the majority evolved to be S57s, and they moved in a wheel motion! There was one frictiony node at the back, and four at the front that spun rapidly. Pretty cool

[@Т1000-м1и]

Perspectives music really reminds me of like 4 years ago, even 5, old life. Just sitting there, watching random Russian gameplay videos.

[@DumbdogsWin]

I want a download

[@rayedhamayun]

Description

[@MertensHelbelga]

I though your comment was "What a download!"

[@gailraby1722]

+Martin no excuse for stupid..

[@godoverlordquacken4003]

@@gailraby1722 No excuse for bad grammar

[@failedfishermanBC]

Imagine if this guy just laid the groundwork for a computer program that will go on to one day evolve sentient virtual beings. :O

[@cortster12]

His simulation is based off of other simulations. Biologists have been simulating evolution for years.

[@failedfishermanBC]

Cortster Bummer.

[@cortster12]

Gagan Baath Why is that a bummer? Just goes to show that scientists are smart people who have a handle on things.

[@failedfishermanBC]

Cortster Good point.

[@sciencepower608]

Are we a simulation?

[@Rin8Kin]

Marvelous demonstration of how occasionally junctioned singlecelled creatures could evolve into more advanced ones.

[@doublemcbilly6175]

You should release this to the public, let people run it on their pc's, and have a periodical leaderboard that breeds the top 2 creatures. I can imagine people finding the "perfect specimen" and trading creatures with unique characteristics.

[@tanjiro_tsukki7912]

0:24 Cary: Let’s push this Simulator to it’s limits Me: The *REAL* limit is 2763

[@owenplays3770]

Only bfdi fans will understand 😂

[@Bazinga9000]

There should be an option to create your own creatures and test them against the population. Also download link?

[@wyattbuckbee7766]

Download link pls?

[@Gukslaven]

It's in the description?

[@negatiff54]

dead link

[@annacatton5929]

nope

[@2FuNnY4uDude]

More mutations! More generations! Obstacles? Great project!

[@GeFlixes]

This is Evolution Simulator thingie seems to be a very good way of explaining how evolution works in principle. Also, the histogramm looks a lot like a plot of for example, economical power or technological advantage, while the left plot looks like how either a competitive Civ game or a innovative-competetive market works - seems like innovation in business really is just trying out lots and lots of new ideas and sticking with the ones that work.

[@an_annoying_cat]

I think the middle anchor might be a double anchor, both an anchor holding the top anchor in place (and maybe pulling the black node up a little) and pushing down the white node, by being so close to it and being balanced by having small and strong muscles. The middle node holds together the other nodes by not letting them get that far apart with strong muscles and keeping them to what their supposed to do. That's where this species triumphs over S45, by a node that simultaneously acts as the white and red nodes, and also pulling the rest of the structure back to the black node so the muscles can expand quicker. It's beautiful. Magestic.

[@royalshadow1177]

Plz let it be downloadable. I love evolution simulators. Plus, have you tried boxcar2d yet?

[@Karataron]

this is a very interesting video series, thank you for this great content.

[@danieljohnson987]

I wonder if he started the simulator all over with a new random set of species -- does it always eventually lead to the same triangle + node as best evolution? -- I'm running it now to find out -- first person I have ever subscribed to-- damn this is cool.

[@A-Milkdromeda-Laniakea-Hominid]

I watched all 4 and it was highly enjoyable and super interesting. I really like the ingenious idea of simplified creatures, great rules. I'm tempted to download your generator and tweak gravity; make me some Moon Creatures lol.

[@adamleblanc5294]

This is the coolest pet project I've seen in a while...

[@jaredspence3020]

tldr: The second pulling node keeps the anchor node close to the ground, making it more efficient. I'm a runner and the question of the second pulling node got me thinking. As humans we have the hamstring to pull our foot off the ground, hipflexor to bring the whole leg forward, and quads and glutes to push the body forwards. The ankle seems obsolete. However the ankle is important, not just as a shock absorber, but it keeps the toes on the ground a little longer and gives a last little push at the end of the stride, hence keeping the ankle. I don't think the second node and muscle work to pull the black anchor node forward while keeping it close to the ground. It is on the ground for a longer period of time and is more efficient than a creature without a second pulling muscle. It is possible the red pulling nodes have the same function as human arms, moving to balance the creature as it moves forward.

[@cizma27]

Wow! This is amazing, I thinks of myself as a good mathematician and physician and potentially good programmer, plus I love evolution and biology so this really hooked me, even though I am total noob for creating simulations. I tried running yours, for few times. And every time even after 400 generations in each go, 99% of my creatures would stay at origin not moving more than a milimeter. I wonder why is that!

[@darkmater4tm]

The top creature looks like an abstract running cheetah. The back legs give a strong push forward and a little upwards. The front feet keep balance while the back feet contract and fall to make contact with the ground again.

[@rockagold3819]

Whenever I played I thought it was similar to a running rabbit. But by the time you get to gen 300, it looks like a mix between a super-worm and yes a cheetah. For me, playing at about 200 gave it some consistent hops in the air, making it like a rabbit.

[@RaminSakhtemani]

Wow, This was amazing, you should publish your work in a science journal if you haven't already!

[@Back7sword]

Found this quite interesting thank you. Brings home a couple of real world truths to me. Firstly is that things don’t have to be perfect, they just have to work. Secondly, evolution seems to have difficulty taking a design concept backwards.

[@Mrcometo]

I think the central node "tunes" the rithm

[@odinlindeberg4624]

is there any chance that letting this run long enough would produce bipedal locomotion?

[@gumball1216]

odin lindeberg That's what I was wondering

[@CrimsonTheOriginal]

youre making a false assumption that bi-pedal movement is an efficient and effective method of movement

[@odinlindeberg4624]

CrimsonTheOriginal I don't remember the merits of triangular locomotion.

[@pmangano]

A chance? yes. is it a big chance? no idea

[@Stryfe564]

I doubt that bipedal locomotion would have any major advantage over triangular when it simply comes to getting the highest average speed over a perfectly flat terrain.

[@appsenence9244]

What if we allowed for it to be more complex structures and we have a different goal? Let's say, instead of nodes we would use electrical forces and chemistry in a 3D environment, could we create some awesome things? AI???

[@crazyhayden]

In this evolution game, it has been clearly tested and tried (according to my tests) that any creature will inevitably evolve into a triangle like being, and that triangle will slowly evolve to have the move downward, and rightward. It has been treid and tested. The ultimate being in a simulation of evolution, is a triangle. Thank you for listening to my theory. I call it "Triangle Evolutionary Supremacy"

[@adamhellerup]

Loving the little "bunny"-runner that wins it all :)

[@teytreet7358]

If you replicate the simulation, do you get the same result?

[@RingxWorld]

nah you can play this with cars box2d. you get different results but similar

[@velocita84]

no, it's totally random

[@erlendpowell7446]

Actually, assuming the developer is using pseudo random numbers in the simulation (which is most likely the case), then the simulation would run exactly the same each time as long as the same initial seed is used in the random number generator. Many games rely on this for replays, or multiplayer mode, to keep both clients in sync.

[@ChaosFonja]

yes and no. i have run this test quite a few times now and it always does the same thing. i have never reproduced the results in this vid. instead what happens for me is at about 40-50 generations the s33 creatures begin to overwhelm the population and by the time it gets to generation 100 almost all creatures created are s33's. once that point hits it is useless to continue running the simulation, the furthest i have taken it out to is about 650 generations and by that point s33 creatures made up 997 of the 1000 creatures. the average distance these creatures go usually peaks somewhere 12-15 meters.

[@cruelpulse]

Might be to do with the physics engine in the sim. I'd imagine RAM / CPU Clockspeed effects the physics in a very very slight way. Which makes no visible difference in the beginning of the simulation, but after many generations has a "butterfly effect". Chaos theory kinda stuff. If you can tweak your clockspeed or how much RAM you have, I predict you'll get a different result.

[@jessicabevan4657]

This is amazing, I ran the simulator for about 5000 generations, multiple times and never got past 15 metres.

[@ThomasOrlita]

the first time i ran it i got 20 m after 300 generations

[@SanteLTTT]

first simulation 200 gens, 21.6m S33 beast

[@Lukehunter990]

How can I try this?

[@thefeelstrain9920]

@@Lukehunter990 i know its late, but here you go www.openprocessing.org/sketch/377698

[@skullcrushers1000]

Will we ever get a download?

[@crusatyr1452]

Hey, you're picture is Rhett too!

[@rockagold3819]

XD It looks like Guardian X's Rhett is screaming at Rhett Having a Vietnam Flashback's Rhett

[@crusatyr1452]

Peter Lin FYI, mine is from a recent Good Mythical More, where they play the new Bop-It. Here, Rhett is doing "Sing It"

[@rockagold3819]

The GuardianX Niiiiice.

[@OuterSpaceandTimelapses4]

My guess for why S46 is better than S33 is because while S33 has one node per one muscle, S46 has one node per one and a half muscles. Each muscle has significantly less work to do, so even if, by pure chance, an S46 creature’s muscles aren’t as good at contracting or expanding, it can still do as good (if not better than) an S33 creature.

[@PetkoBossakov]

I like how this shows evolution working in jumps. You get radical changes over a few hundred generations, then it slows down to an almost complete halt, then it can potentially shoot up again if you change the outside conditions...

[@TheDiamondSwordsmen]

Could you possibly try changing this so that it tests how high the creatures can jump? EDIT: Sorry, didn't notice you already did this.

[@Wetbboi]

2:12 the creature literally started to fly💀

[@maydayzone8798]

Love the simulator but wtf is a tringle is it a triangle?

[@mikotosuoh5752]

Mayday Zone probably a gazzel since its so fast

[@Yolandi.]

Great series dude. Interesting to see that none of them just made a pentagon and started rolling. That would have been very cool.

[@joshsblee]

Dude this series of videos was sooooooo COOOOL to watch!! THANK YOU