Evolution Simulator of Creatures Jumping Vertically
Вставка
- Опубліковано 21 лис 2024
- This is the longest video I've ever uploaded. I think if I ever upload another video of similar length, I'll split it into multiple parts!
Also, ignore all the dumb things I say, because this video's mostly un-edited and unscripted. I also realized I said "you know" and "like" a lot, but it might just be because I recorded this around 11 PM.
Music is "Ev'lution" (not "Evo") and it's by DemiPixel.
"It's a very unhealthy guinea pig if it's like internal organs are like moving around at the speed of light all over the place."
*- carykh, 2016*
Totally
The number of insane quotes this guy makes is almost uncountable. I love listening to his narrative.
when did he say that
"Thats not healthy"
Dávaj pozor na učiteľku "very calm and, spastic."
Just imagine if Cary was god....
Triangle- Look Great One I can jump really high, does this please your grand design?
GodCary- Booooooring. *resets existence and bans triangles*
theblasblas Square: I cant jump that far, please let me go a few more generations
Cary Jesus: Okay
Square: *turns rigid and does seizures*
Cary Jesus: *B A N*
theblasblas that’s why humans have four limbs
Fastest runner: A coordinated creature using its nodes in perfect synchronization to achieve speed
Biggest jumper: POGOSTICK SEIZURES
Now you should do a simple platformer and have creatures jump platforms until they fall down and die
Yah imagine, we would all be crawling around for a billion years. OOPS ! That is what happened for real.
poor s34, banned for no reason...
you should add energy/stamina so you can't win with constant seizures
Brillant idea !
yea, like a muscle can only be used a few times, like for example, 10 times and it will go limp
do this
Also, make it so each species has a population cap. When progress stops in these videos is when the median creature becomes very similar to the best creature. Make it so it kills the worst of any species over 500 of that species.
OR
If one morphology takes over, cull the heard and introduce a round of extreme mutations with a radiation button or something.
I JUST GAVE YOU 1 MORE SUBSCRIBER FOR TWOW ITSELF
LET ME KNOW WHEN YOU HAVE TO BE IN SEASON 2,OKAY?
22:30 "I feel like all the bugs are fixed now" -every programmer before he finds the next one
sometimes it's true tho
and oh god does it feel good when it's true
"I was expecting a really elegant creature frolicking in the filed but instead we got..."
Seizure-tringles.
Also, the median rising even as the best stagnates is explained by more and more creatures performing above the median. So the worst performers are eliminated and replaced with better performing creatures, though ones that might still not beat the current best.
owo
If you were expecting creatures in the “filed”, I’m worried.
owo
@Micheal Drake bruh who cares
"I'm one dimension ahead of everyone else." -- Cary Huang, 2016
If he is 2D then that means cary is 1D
I'm 4 parallel universes ahead of you
I’m 4 Generations ahead of you.
Am i the only one who wants to see generation 1 million or something huge? Processing power and shit might become a limiting factor here but evolution is a mighty slow process (if not speeded up exponentially here) but could be interesting to see this scaled up massively
mikeyBIA write a program like this one with similar rules and save up for some time on a supercomputer
shiny penney i know that logistically its not feasible but i think it could be interesting to see the results
mikeyBIA I'd suggest emailing a professor at a university near you and positing your idea. An evolutionary biologist will have the funds to answer this kind of question using whatever supercomputer they have access to. This could help more than one PhD student publish a paper
You see how it plateaus? That's essentially as good as the most prominent design can get. If you leave it going for longer, it will only get higher if a new design turns up.
If you're insterested in that, run several instances in parallel, so you have different evolutionary lines.
@@Asdayasman finally a smart answer lol
30:50 is Cary discussing the sibling history of the triangle and the tetrahedron, and how they were separated by an evil god called MATHS to different dimensions, 2d and 3d.
Summary of what happened, and why:
In a free for all, triangles won after 200 odd generations. Once he eliminated triangles, the four node creatures won with HIGHER results.
Additionally, the S45 creatures dominated first, and then later the S46 creatures began to rise.
You can see this around 30:00 where the orange has overtaken the purple, seemingly out of no where.
I hypothesis that if he eliminated triangles and 4 nodes, then 5 nodes would win, and with a higher jump than the S46, and so on. I don't know if there is an upper limit, but I wouldn't be surprised if you eliminate everything under 10 nodes, then 10 nodes would win with a very high jump height (higher than all previous ones).
Why?
To maximize jump height, you need these type of effects:
1) as high "push off" force as possible
2) minimum energy loss while going up (not too much spazzing out)
3) maximum clearance height (no legs dangling down at peak height)
1) is the most important. 2) and 3) are just optimizing the interactions between the nodes, (eg a setup which encourages pulling all legs up at maximum height).
In order to get a high push force, you need maximum muscle force.
Having more nodes, means you can increase the max number of muscles, and therefore the push off force.
However, having more nodes means there are more ways it can stuff up and be 0, so in the short term, lower nodes such as triangles will out compete more complex creatures such as 4 nodes. This is because a random triangle has a lower chance of getting 0 than a random 4 node creature.
This means the gene pool in the population will become dominated by the creature which is the fastest to gain a critical mass in the population. Not because they have a higher long term prospects, but because less complex creatures can be optimized the fastest.
Once you artificially remove the creatures with the fastest optimization, more complex creatures with more nodes have a chance to optimize.
This also happened with the S45 and S46. (4 nodes with 5 muscles, and 4 nodes with 6 muscles.) At first, the S45 dominated, but then was eventually overtaken by S46 when S46 happened to get a foot in the door and was able to start spreading. Again you can clearly see that at the 30:00 mark, where the orange overtook the purple.
This is in effect the same dynamic of what happened with triangles and 4 nodes. In theory, if you ran the triangle inclusive simulation long enough, the 4 nodes should eventually optimize by chance and eventually take over. If you ran trillions of generations, then in theory, you would get higher and higher node outcomes with higher and higher jumps.
At 14:48 he says "S45 is just hanging on there, but is blinking in and out." Run it long enough and it'll take over.
Light And Shadow did you just write an essay?
Interesting, but 10-20 nodes will take too much time/generations to get their perfect jump. We should increase the population size to 10000-100000 and decrease the death rate to 20% so these complex creatures (5-10 nodes) may have more chance to evolve. Like in the video the S5x cannot evolve within 400 generations because their population is too small (
Helium iodine Graphite uranium yellow silly
@@veqium I'd love to run these simulation for fun my computer is running 24/24 anyway so there is no problem. Anyone know any software like in the video please tell me.
Thanks for this, was a good summary and I like the hypothesis. However, what you have suggested is to me like saying that if humans were to be replaced with omniscient, omnipotent, all but perfect creatures, we would be outperformed. Yes, of course that is the case, with more potential, more can be achieved, that’s just a fact, it just takes too long to happen, which is why these creatures do not exist. However, you make a good point, and I’m in no way trying to insult you, just making my own suggestion.
If you ever come back to this I have an idea for improvement:
Instead of allowing creatures to jump many times, give them one jump and end their simulation the moment they hit the ground again. If they jump before the two second mark, scale their fitness back proportional to how soon they jumped. This should decrease the ability of spastic creatures to reproduce without and reward creatures that settle down before making one strong jump.
If these videos have taught me anything, it's that someday triangles will take over the world.
Edit: Also, if you wanted an elegant jumper instead of just a spazzer, maybe weight them based on an average of ALL their jumps in the time limit? So one that jumps lots of little times will be killed off as opposed to the ones that make one or two good jumps? Also maybe have them "land" and stabilize for a second before their muscles start activating?
Ndoki Hasaki So these videos are basically a proof that the illuminati dominate the world?
Spazzers prevailed because of the artificial limitation enforced by the frames-counted bug.
Ndoki Hasaki How about making them consume energy when they move, then killing off those who use too much energy? Spazzers move so much that maybe this would calm them down?
Ndoki Hasaki triangles have already taken over, it's called the illuminati
triangles have already taken over, it's called Engineering. :P
I bet that if you waited long enough, the creatures will begin spazzing so much they fly.
have you considered having bones? instead of muscles being able to push, have rigid structures like bones and muscles can only exert force in the form of pulling, just like real life. The creatures would be a lot more interesting.
I agree!
carykhy read this now
CARYKH DO THIS!!!!!
also maybe tendons (muscles that cant stretch but have slack. idk sounds cool though.
*****
Reasons?
*****
Wot
Have 1000 creatures
_KILLS 500, half of the population_
*this does put a smile on my face*
Like Endgame:
*Thanos* I am inevitable
**NOPE**
*Iron Man* I am…Iron Man
No I am cary
XD vooo
I'm honestly more curious about why S58s and S45s never died in spite of the S46's ascension to victory
The Poubel S45 is very close to S46, so I think it dies often, then get created bcuz mutations happen. S58 IDK
Like Light and Shadow said, I think S58 has the potential to outjump S46 in the long term. However its population is too small for a mutation to appear (30/1000) and the more complex a creature is the less chance it has for getting perfect jump in 15s. Imagine if its population is 3000/100000 sooner or later S59 will appear and take the throne of S46.
i'm curious about why there was never a creature with more than 9 muscles or nodes. maybe it's just because carykh didn't think of a way to name the species, like S-5-10, S13-78... or S5N10M, S13N78M...
@@LeKhang98 I agree with all of that. To add, when triangles ("tringle" is a hate slur) are allowed, they outcompete the hell out of flimsy S44s, so the unsupported boxes never get a chance to evolve additional supporting muscles to get up to a S45 or tetrahedron. I really wanted to see more complex species, at least to see an S58 jumping.
Also you probably didn't get to see any malformed mutants at the bottom because of the racist no-triangles eugenics rule. They simply couldn't be created because they didn't fit the parameters.
If you wanted to continue eugenics to keep it interesting without hamfisting the course of evolution, I'd suggest that rule be something like "# of muscles must not equal # of nodes, and randomly adding (where possible) or removing a muscle from violators to comply. Also to start it down an interesting path, I'd seed it with creatures from 3 to like 8 or 9 nodes, and have occasional periods of increased likelihood of morphological mutations instead of just strength size and friction.
@@slambo2001 it also seems that mutations in this simulation are made by changing 1 variable, and so the triangles (where 1 mutation can mean a doubling in performance) are able to sprint ahead at the beginning and wipe out other competition before those random movements can be coordinated by chance
It's interesting how the exclusion of the triangle creatures eventually led to better-adapted creatures overall. It's kind of an analogy for invasive species wiping out native species before they are able to evolve a defence, even though those native species might be better suited to survival in the environment in question.
Another example might be the first Native Americans hunting horses. Horses were ultimately an incredibly powerful resource for growing civilisations, but were unluckily used for food in North America thousands of years before anyone thought of domesticating them.
The problem is that with the triangles, he had the bug where only video frames heights were counted, not the simulated frames. This means the first generation was heavily crippled, and only favourited the first ones to be simulated, while there was a big chance the second half of the population had a member that jumped pretty high on his first jump.
Furthermore, if the initial generation of mutations is not random, this discards all the later mutations of generation 0.
He should have redone the simulation again with the triangles, without that bug.
This isn't true, horses are an old world animal
That isn't true, horses existed and went extinct in North America thousands of years before they were reintroduced by conquerors.
Super late, but there is a saying that evolution is a blind watchmaker. That is- evolution isn't designing anything, it fumbles around until it gets a solution that works. The triangles ended up being dominant the first time around because they were simple, and it was much easier to get them to work than the overall more effective 'pogo stick' method. There are plenty of examples in nature where creatures have sub-optimal features because of evolutionary quirks. Mammals have much poorer color vision than other vertabrates because all of our ancestors spent millions of years being nocturnal.
If humans had to survive by jumping high, I'd be long dead. XD
Up Down Center yeah, that makes sense.
Doctor: PUSH! PUSH! Nurse! Get the net!
Nurse: Yessir!
(A net is placed over the bed)
Doctor: HERE IT COMES!
(Baby is born, and jumps several times immediatley)
Doctor: Oh, this one isn't as good as the median. Kill it, nurse!
Nurse: Right away, sir!
I’d survive 30 Generations tops
give the limbs air friction.... then you might see a very elegant creature that relies on catching the air to go up
Looks at a triangle having a seizur: "This is how you jump." Ive been it wrong my whole life
I want a pet tetrahedron now. I would name it S46.
Peyton Peterson What about S34?
LimeCyanizer I’ve got a pet S57.
@Peyton Peterson S33s are Tringles.
@@Rainb0wMapp1ng I got a pet S45.
I want a pet tringle
i think that the maximum number of nodes should be increased to like 20 or 30 or even more because the designs would be more complicated, intricate and interesting.
Or just more balls of spazzing muscles.
crazyguywithasword 👍🏻
I agree with all of these
Computer will blow. ..
Yeah you'd get that, but you'd also need a Supercomputer or some code rework to allow each generation to take minutes. You'd also need to give up on the step-by-step mode without a supercomputer but you could see it after each generation is complete.
What if you added a sort of energy consumption consideration? Like, the more muscles and more movement per muscle, the faster you'll run out of energy, similar to real life.
That way, you can naturally get rid of spazzy organisms.
I have noticed that these creatures make their evolutions rather quickly and then it starts to slow down until almost no progress is made. what you should do if at all possible is change what determines if a creature is successful for not. ie. start with jump height and then change it to how far it can run.
without restarting of course. in case I didn't make that clear
It is possible, you just have to program it.
i would say check to see if the high jump was in the first 5-7 secodsn and that gives for 0.7 of its score thif it was after that then it gets 0.25 score,,, so when they cull the bad oens the oens that jump sooner have less chace to be remvoed
This is because, at the start there's a lot of progress to be made, at the end, there's only a few possible changes
#EndDiscriminationAgainstTriangles
xLuCk3Y #trianglelivesmatter
*tringles
#StopCuttingRatesAndBringBackSimplicity
He did. Can't discriminate against something which no longer exists.
No thanks
You could penalize the amount of twitches, and that way it'd probably evolve to more controlled jumps. Another more realistic way could be adding energy levels, they couldn't be twitching like that because they'd lose muscle contract power as they do that.
I'm thinking it'd have to go many more generations to see high results (like running 20 meters, for example) because a charged up, maximized jump would require more nodes and specialized muscles to plateau on jump height. I'd love to see advanced, mutated creatures!
You should add weight as a factor in the sim. I think more complex creatures could win over spaz kids if they could get balancing nodes.
Maybe make the max pop for a spiecies be 400? Forcing more variety?
Than nothing can take over.
Redrar i once read a paper on how to prevent a species to take over without compromising it. if i can find it ill link it here ;)
TheAwesomeGamingClub I saw the fourth line and I was like "somethings up, ain't it?" So I checked the initials and I read it before the last line.
Patrick Winter But this makes triangles more likely to dominate, which cary hates
Please do more videos on Evolution Simulator!
Make this a phone app and I'd love you.
yes
lol. your phone sadly doesn't have the power to calculate those simulations
thats the cool thing about simulations... they dont have to be in realt time -.-
This isn't very demaning... Plus, the smartphones nowadays has about as much power as a 10 year old top of the line gaming PC, they ain't so weak.
ian justiz lol no
What you were saying about Eras - The reason that doesn't happen is because this isn't real evolution. Evolution is adapting to a changing environment, and this is a constant environment.
they are adapting to a changing environment. They used to live in warm fuzzy chemical soup and suddenly there's a predator that consume you if you don't jump out of the way. Luckily they only eat 500 people a week and you breed at exactly that speed.
"A matter of two or three months." - Cary, 2.5 years ago
TWOW is still not finished...
Is it
He finished it today
Great video! Have you considered varying the "selection pressures" to trial the evolution of different strategies?
For example, if you changed the measurement of jump height to be cumulative rather than looking at the maximum so that the total distance jumped in 15 seconds was measured. I think it would be interesting to see if strategies where lots of little jumps were favoured over strategies that produce the occasional big jump.
Another aspect I think would be really interesting would be to somehow quantify the "energy" cost of either the movements or having different qualities of nodes (or both), and factoring that in. This way you could optimise the efficiency of the behaviour. It could also be cool to make it so that the higher a creature jumped in it's allotted time (or the cumulative height jumped in that time), the more "energy" it's offspring had to work with.
There are a lot of really cool ways to make these simulations truer to the evolution of living things on earth, I don't imagine the coding would be easy though!
I think a good fitness function that could favor little jumps is the sum of the height (height of the lowest node) over all the frames counted.
You could also subtract the distance (absolute value of difference) between the currant height to the previous ones multiplied by a small coefficient (or just add the reciprocal) in order to make them less spastic.
ארד קרן
Yeah that's what I meant when I said the measurement of height could look at the cumulative height rather than the maximum.
I just love your commentary :'D "It's a very unhealthy gunea-pig if it's just moving around with the speed of light" ROFL
"He's off the charts. To the left."
"This is more than twitching, this is more like a seizure." LOL I died at that part.
It might be a good plan to write something that ends the simulation and moves onto the next creature if all nodes are touching the ground simultaneously
If he wants to keep the rules the way they are, I agree.
But why would a creature not be able to raise a node? I frequently end up completely horizontal, but I manage to raise a limb and jump out of bed again, usually^^
Benjamin Philipp Benjamin Philipp we have joints with pairs of muscles working together. Also bones. If he added non-contracting muscles as 'bones' then that would be possible.
Pablo NoEsEspañol
I don't quite follow the mere logic here, there must be something weird about the set-up (you really only need contracting muscles)
I'll have to look into it further...
Thanks for the info
It requires bones. Try holding a piece of string between your hands and pulling it with your right hand in a way that doesn't move your left hand straight towards it. Now imagine a rigid object the same length as the string between your hands as well. When you pull the string, your hands will collide with the rigid object and move laterally instead of through it. This is how bones and muscles work. That's why it's called the musculoskeletal system; because they work together and are both integral to movement.
Edit: And to further clarify, once all the joints are on the ground, there is nothing angular without bones that would be able to bring them back up again.
I sure hope that explanation made sense :P
Megabobster
Hey thanks!
Wait... did I interpret Pablo NoEsEspañol wrong when he said "we have [...] bones"?
I thought he meant "in the set up" - I take it now that he meant "we as actual organisms do, but the ones in this simulation don't"?
That would explain the "non-contracting muscles" thing - the _bones_ the set-up is missing :D
s46:*having seizures 24/7* cary: wow,wow, this is like such a calm peaceful gentle creature, you know?
more nodes = higher jumps
because there will be more actual technique instead of just blind luck as with 3 and 4 nodes
CaryHatesKary banned Tringles! I love Tringles! (Sorry for saying a shapeist hate slur ;_;)
Is it just me, or is everything funnier at 12 AM?
i think it would have been more realistic if it were creatures learning to swim
he would have to programme in buoyancy and water physics and that's a lot of work
dat would be kewl to see
+Jason The Ranga nah, thats not very hard, just no gravity and the friction effects its movement speed, it would be easier to program
I think it'd be more complex than that
programming drag to propel a target is hard
I noticed that you took interest in a few select species, but you weren't always able to find examples of them when you wanted to. Have you considered adding a feature that would allow you to see the worst, median and best creatures of a selected species?
It wouldn't be as useful once certain species become more rare and the dominant ones become the majority, but it'd be interesting. ESPECIALLY if you could eventually force certain circumstances that would keep the species diverse.
"In about three months, TWOW season one should be over!" Oh Carry, ever the optimist :)
I think the reason why the median height increased in Gen 1 when you removed 3-node creatures is because I think 3-node creatures (I'll call 'em 3n) are both the best and worst jumpers. 3n are more like to be stable and spaz out to leap, but since there are fewer nodes it is also more likely that all nodes will end up on the floor and preventing a jump. Say 3 nodes of a 4n creature are on the ground, but the 4th node is still in the air: there's still opportunity for the 4n creature to lift off the ground, and even potentially bring other nodes off the ground as well.
Progress could only be made when the triangles were killed off. Face it people, eugenics work.
/jk
"Mens Right" is a fitting name
New View ?
Mens Rights Edinburgh Just kidding ;)
Fkin triangles
No, if we had stayed with triangles it would've improved
You should combine the running and jumping simulations so that the creatures can either run fast or jump high to survive, and the most successful would be those who both run and jump.
Also, are darker nodes heavier than lighter ones? Because they should be.
HoujuuGaming dark nodes only have more friction.
HoujuuGaming its like a sandpaper covered ball ( dark ) vs a ice ball ( light ) nothing to do with weight
HoujuuGaming more friction
Like Nathan King said, next time give the nodes air friction
Then you'll get an elegant creature
soup guy but then a creature would be able to jump upward infinitely, good idea for a flight simulator, air friction, and then it goes until the creature hits the ground again!
nice profile pic
theAnarch thanks, my sister drew it on my white board while I was at a camp, I just added the name and the 'fright' marks around the creeper
i meant soup guy but you too lol
At the end of the four node generations, the design is some what interesting. There's one dark node to make a firm grip on the ground during the main red piston's spasm, it effectively pushes off the ground, with red's spasm and pulling towards a lighter node to point the mass skyward then firing itself upwards, then while in the air, reversing itself to cheat the rule. I love it.
"Evolution Simulator"
14 minutes in, divine intervention
Retconned 🔺 s out of existence. Cary hates 🔺 s!
Idea! What would happen if you did this simulator, but every gen you switch between distance and height.
that would create a jumping-sprint thing.
Id watch that tho XD
Great idea!
I'd be interested in seeing how this would work at different frequencies--1 in 2, 1 in 10, 1 in 100, 1 in 1,000, etc. I've speculated that one reason the European coral reefs survived the K/Pg mass extinction better than most was that they were near land, and endured periodic nutrient overload due to flooding on land. This should have hardened the reef systems against nutrient overload (a major factor in reef die-off in the K/Pg mass extinction). It would be interesting to see a computer model demonstrate that this concept is viable.
+1 do that, sounds great or maybe stairs...
yeah!
I think making fitness the vector magnitude of the jump height and horizontal movement "jump" to get an amazing long jumper would be cool!
#TriangleLivesMatter
#MakeEvolutionSimulatorGreatAgain
EZIC Name XD
#AllShapeLivesMatter
#IlluminatiLivesMatter
Awesome Gaming omg XD
I love hearing about the programming/code side of these. Thanks for giving incite, honestly I'd watch a complete video where you just break down the video and each part of the code and how you did it and whatever. Would be very interesting!
every time he says "Breakthrough" take a shot
Haha you tryin to kill us?xD
And for every "I really like"/"I like" as well
Imagine how drunk you could get from doing that everytime a BFDI character says like. Most of the time, it is Match.
@@nicefloweytheoverseer7632what's that
This video was off the charts... to the left.
So it was backwards.
Hey! Cool little program! I think you should modify the fitness function to take into account the number of high jumps too. That way we'd get creatures that keep jumping instead of creatures that just happened to jump high once by chance. I think they would be more elegant too. I would suggest a fitness function that would be based on the average altitude at which the creature is over the whole 15 sec period, that way it would force them to keep jumping.
yknow the simple creatures thriving makes sense
whats one of the most common creatures on earth? bacteria.
Though i agree that the whole seizure winning thing shouldn't happen, but you haven't implemented damage to nodes or muscles(which does make perfect sense) from hard impacts
I find it funny that you can actually see the bug he noticed at 3:45 all the way back at 2:11 when he was talking about fixing the camera; the fitness is only 2.56 but the creature gets up to 9.23 cm.
evolution stimulator episode 57- inbreeding creatures for hundreds of generations so they’re seizure prone and therefore jump, like, two feet
I was skipping through and then I heard Guinnea Pig organs flying around at the speed of light. 😂
I wonder if you could link the jumping with the distance to find the ultimate organism...
Adam Wolowczyk a 45° angle goal line
Yeah. He should try defining fitness like this:
fitness = (distance moved right) * 4 + (maximum jump height)
The “* 4” is so they move significantly to the right, not just randomly wobble right while jumping high like some of the creatures in this video do.
44:09 here you go
What was the mutation chance?
Jamesthe1 Στ {[SigmaTau Entertainment]} 50 precent
imagine that if you let it simulating *over 9000* times, creatures learns to fly... so exciting!
i like the new series cause he's more chill and involving us with the bugs/coding
unlike people where if something goes wrong then they cut it out, this dude includes it
This simulation is absolutely amazing. What if we add more gradients?
28:34 No Man's Land? I see an uncanny parallel to my spreadsheet.
Getting dunked on?
Someone should make their TWOW book a Tringle. Tringle's never lose.
Except when they're forbidden.
I think this video was exactly what i needed to destress. Thanks for the fun content Cary!
Footnotes: Slowest creatures at the bottom. (going backward = slow)
Carykh: *I put more work into this than you thought*
By the way, I decided to name it "Ev'lution" :)
Fix twowbot
How did you make the music? Did you use a certain program?
Garageband.
You are such a good troll 😂
JohnSanity As much as I love to make jokes, that's actually the program I used.
"Except I shouldn't say funsies because that's a dumb word." -- Cary Huang, 2016
GOD DAMN PUT IT IN FULL SCREEN AND TUEN MAGNIFIER OFF!
And don't do grammar mistakes either
Honestly, I was expecting a roughly 10 to 20 minute video when I clicked on this (because I didn't check the time before clicking), but when you said you were going to do another generation, I checked the time. Seeing it was almost 50 minutes, I was pleasantly surprised. I love this kinda stuff and I'm glad you work on it :P
I recently learned a bit about deep learning and always found programming interesting so this is very fun and cool to watch, just found your chanel and love it
I think you should always describe your species definition and the rules thoroughly in case someone stumbles upon one random video
42:02 interesting how the winner is still collapsing after 37 seconds. Overbred!😂
I subscribed for the Evolution Simulator videos, but also enjoy TWOW.
Maybe the thing is that triangles converge to their best solution quicker and dominate. Other forms need more time, which also proves that it's not only important to find a good solution, but sometimes is better to find an acceptable solution quicker.
I love how clearly your enthusiasm and curiosity comes through. This is really cool.
Wow. Has anyone realized that when all the creatures are shown, at the bottom, it says that going backwards is slow?
Edit: I meant to say that going backwards doesn't change the jump height.
He forgot to change the text from his last evolution simulator
Ya dude i subbed like 20 minutes ago cuz these evolution videos are awesome
That awkward moment when you hit [Sort] and nothing moves because they were randomly generated in an already sorted order ;P
(can anyone calculate the possibility for this?)
1/100.000
I think it's 1/1000! , which is waaaay smaller
no, he said 1/100,000. it is the european system where periods are commas to us and vice versa
I know how that works, that's not the problem. What i am saying is that i believe 1 / 1000! is the possibility of having 1000 elements already sorted, since factorial 1000 should be equivalent to the number of the total existing permutations, and only one of them is the correct one.
1 / 100,000 is enormous compared to 1 / 402,387,260,077,093,773,543,702,433,923,003,985,719,374,864, 210,714,632,543,799,910,429,938,512,398,629,020,592,044,208, 486,969,404,800,479,988,610,197,196,058,631,666,872,994,808, 558,901,323,829,669,944,590,997,424,504,087,073,759,918,823, 627,727,188,732,519,779,505,950,995,276,120,874,975,462,497, 043,601,418,278,094,646,496,291,056,393,887,437,886,487,337, 119,181,045,825,783,647,849,977,012,476,632,889,835,955,735, 432,513,185,323,958,463,075,557,409,114,262,417,474,349,347, 553,428,646,576,611,667,797,396,668,820,291,207,379,143,853, 719,588,249,808,126,867,838,374,559,731,746,136,085,379,534, 524,221,586,593,201,928,090,878,297,308,431,392,844,403,281, 231,558,611,036,976,801,357,304,216,168,747,609,675,871,348, 312,025,478,589,320,767,169,132,448,426,236,131,412,508,780, 208,000,261,683,151,027,341,827,977,704,784,635,868,170,164, 365,024,153,691,398,281,264,810,213,092,761,244,896,359,928, 705,114,964,975,419,909,342,221,566,832,572,080,821,333,186, 116,811,553,615,836,546,984,046,708,975,602,900,950,537,616, 475,847,728,421,889,679,646,244,945,160,765,353,408,198,901, 385,442,487,984,959,953,319,101,723,355,556,602,139,450,399, 736,280,750,137,837,615,307,127,761,926,849,034,352,625,200, 015,888,535,147,331,611,702,103,968,175,921,510,907,788,019, 393,178,114,194,545,257,223,865,541,461,062,892,187,960,223, 838,971,476,088,506,276,862,967,146,674,697,562,911,234,082, 439,208,160,153,780,889,893,964,518,263,243,671,616,762,179, 168,909,779,911,903,754,031,274,622,289,988,005,195,444,414, 282,012,187,361,745,992,642,956,581,746,628,302,955,570,299, 024,324,153,181,617,210,465,832,036,786,906,117,260,158,783, 520,751,516,284,225,540,265,170,483,304,226,143,974,286,933, 061,690,897,968,482,590,125,458,327,168,226,458,066,526,769, 958,652,682,272,807,075,781,391,858,178,889,652,208,164,348, 344,825,993,266,043,367,660,176,999,612,831,860,788,386,150, 279,465,955,131,156,552,036,093,988,180,612,138,558,600,301, 435,694,527,224,206,344,631,797,460,594,682,573,103,790,084, 024,432,438,465,657,245,014,402,821,885,252,470,935,190,620, 929,023,136,493,273,497,565,513,958,720,559,654,228,749,774, 011,413,346,962,715,422,845,862,377,387,538,230,483,865,688, 976,461,927,383,814,900,140,767,310,446,640,259,899,490,222, 221,765,904,339,901,886,018,566,526,485,061,799,702,356,193, 897,017,860,040,811,889,729,918,311,021,171,229,845,901,641, 921,068,884,387,121,855,646,124,960,798,722,908,519,296,819, 372,388,642,614,839,657,382,291,123,125,024,186,649,353,143, 970,137,428,531,926,649,875,337,218,940,694,281,434,118,520, 158,014,123,344,828,015,051,399,694,290,153,483,077,644,569, 099,073,152,433,278,288,269,864,602,789,864,321,139,083,506, 217,095,002,597,389,863,554,277,196,742,822,248,757,586,765, 752,344,220,207,573,630,569,498,825,087,968,928,162,753,848, 863,396,909,959,826,280,956,121,450,994,871,701,244,516,461, 260,379,029,309,120,889,086,942,028,510,640,182,154,399,457, 156,805,941,872,748,998,094,254,742,173,582,401,063,677,404, 595,741,785,160,829,230,135,358,081,840,096,996,372,524,230, 560,855,903,700,624,271,243,416,909,004,153,690,105,933,983, 835,777,939,410,970,027,753,472,000,000,000,000,000,000,000, 000,000,000,000,000,000,000,000,000,000,000,000,000,000,000, 000,000,000,000,000,000,000,000,000,000,000,000,000,000,000, 000,000,000,000,000,000,000,000,000,000,000,000,000,000,000, 000,000,000,000,000,000,000,000,000,000,000,000,000,000,000, 000,000,000,000,000,000,000,000,000,000,000,000,000,000,000, 000
1/10^1000
THNX SO MUCH FOR OPENING THE APP TO THE PUBLIC LOVE YOU SO MUCH AND I REALLY APPRECIATE THE WORK YOU HAVE DONEEEEE 😍😍😍
Thanks for the efforts! Really like all the programming behind this. One idea: you can track the evolution of one particular genome, right from the beginning. Each winner has an ancestor, would be cool to observe the path to glory 😊
@carykh For a creature to be considered elegant I'd say it should contain at least 20-30 nodes/muscles. Creatures with less than 10 parts simply can't be elegant IMO. Also, Cary, I think you should put some restrictions on species. For example, make it so that no species can make up more than half of the population. It is much more interesting to see different creatures than just single species.
The best jump never decreases, does that mean that:
a) The best creature is guaranteed to survive, and
b) The simulation is deterministic, and initial conditions are embedded in the creatures design?
a) Yes and it seems like the condition are the same.
b) if you roll the start condition of generation 1 you can have different results.
GreenFox X3 Yeah, wasn't referring to the seed but I can see it was ambiguous. More like any given creature will always behave the same no matter how many times you simulate that particular creature.
Joaquin Pirotto A creature will behave in the same way cause if a Node is at x y fixed, it cannot do in other ways. The behaviour will be different only if x and y are not fixed.
Rafał Stepulak I saw it too, so it means the best creature has a very low chance of dying or the program was changed (maybe unintentionally) for this test
If you look when he kills them, it appears as though the chance of them dying is heavily dependent on their fitness/ranking.
Why does it run so smooth on your machine? What did you use to run the software?
a good computer?
He speeds it up
it shouldn't be a problem if you have a good cpu
Wolfee True enough, but this simulation probably heavily relies on the cpu and not the gpu
well I actually made the mistake of running it in browser... after using processing 2.X it runs much smoother, even on an 1000€ gaming machine.
This has a long way to go, but I'll be damned if you're not on the right track. APPLAUSE sign is on. Keep it up!
S45 is Internet Explorer. It started off kind of in the background when it first came around, but it started to gain traction. It was largely dominant for a few years, with over 90% of that species making up the total population, but then S46 (Google Chrome) came around and is now the most widely seen species of web browser.
So, just since I'm not sure if anyone actually covered it in the comments; at around 23:40 you start questioning why more nodes had a higher median despite complexity, or what-have-you. But you literally forgot to account for the bug you fixed moments prior (in the context of the video at least). Every first generation was run step-by-step meaning that the prior first-gen (before the bug-fix) had significantly more zero centimetre results because it only 'snapshot' possible jump heights significantly less often as opposed to after the bug fix where each frame, regardless of playback speed, was fairly represented.
And yeah, I'm a little late to comment on the vid, but I am most of the time; so fuck it.
I was going to say the same thing. It's like he had sudden amnesia, after realizing and speaking about the fact that the bug fix accounted for the difference in sorting.
This.
Do a video with vertical and horizontal score maybe creatures will lean to walk!!!!¡
This should totally be funded research. The results give a good example of how AI could Work
“Going backwards = slow”
Lemme just PHASE THROUGH THE GROUND
Dude... I am 5 years late but this video ist one of the best videos I have ever watched... Also the music was awesome!
The best species 46 are spasming around like mad all the time.
It seems like the stretch-contract cycle is much shorter, right? Or maybe the muscles are contracting one after the other?
I think I get it... there are two muscles connected to one node and each is contracting-stretching in turn, making the node spasm all over...
How about an evolution simulator with pushing a block?
Wow, actually a really good idea!
Please tell me where to find this so I can play it
m
n
O
P
R
I'm getting so many ideas when I watch these videos. Like how cool would it be to have a long jump one? Measure both vertical and horizontal movement while all nodes are higher than 0.
Or have them start on a platform so if they move too far to the left or right, they'll fall down and get a negative number.
Or having a size and weight property for the nodes, instead of just friction, to see if smaller but denser nodes are better than larger and less dense.
“Twow ends in three months”
Cary, July 2016
Try this again with an average high rather that a highest point. That way consistent jumpers survive better, rather than twitchers.
at 26:50 in the bottom right corner the code says:
translate(-camX*scaleToFixBug,-camY*scaleToFixBug);
what does alap mean
Luca Capperucci as late as possible?
As Long As Possible
Essentially, this is the “keep going until I tell you to stop” button.
As long as possible
Some say if you run this simulator for long enough the creatures will start to gain conciousness and seem to... walk around... trying to find a way out...
These Evolution Simulator vid's are interesting as fuck. I had never heard of your channel before, but this simulator you created is most impressive and I appreciate all the work and time you put into this. Keep up the good work and thanks again.