Youtube comments of (@PrimerBlobs).

  1. Lots of comments relate to racism, so I think it's worth discussing that a little. Racism in humans is a very complex phenomenon that can't be simply traced to genes. To be honest, I don't think I really understand racism, but I'm confident saying a full understanding would take inputs from biology, psychology, sociology, and history, perhaps more. But as the video explains, the green beard phenomenon is very brittle and rarely shows up in nature, so it's not an explanation, and certainly not a justification for something as complex as racism. Thanks for reading, and be kind to each other. :)
    19000
  2. If you're about to leave a comment saying that faster creatures aren't actually less efficient, read this first. I presented that part a bit strangely. At 2:14, I say moving quickly is less efficient, giving the example of a creature moving a unit distance in half the time, using twice the energy. Then, at 4:53, I show a formula for the energy cost per unit time, which depends on the square of the creature's speed. I gave distance per time, energy per time, and distance per energy at separate parts of the video, and that was confusing. So here's a more explicit summary. If we double a creature's speed... - its distance per time is doubled (the definition of speed) - its energy per time is quadrupled (because it depends on the square of speed) - its distance per energy is halved: (2x distance per time) / (4x energy per unit time) That last bullet is the "efficiency" from the video. With its starting energy for a day, a 2x-speed creature can only travel half the distance.
    14000
  3. I honestly didn't even make the connection to Among Us until reading this comment. But I can't say I'm sad to discover a little extra meme fuel.
    3500
  4. A number of people have commented that at 5:22, the green population should be expected to beat out the blue population because the blue population suffers from mutations while the green population benefits from them. That effect is real, but it's too small to tip the scales, and I determined it was too complex for the main message of the video. But because some people are interested, I thought I'd lay it out in more detail here. Only 1% of replications by blue creatures produce a green creature (4:52), and another 1% produce orange creatures. This means blue's replication rate is effectively only 98% of the stated value. 9.8% instead of 10%. The green population gets an expected influx of creatures each time step equal to 0.001 (a tenth of a percent) times the number of blue creatures. Here's a desmos graph of the expected changes per time step for blue and green, from the equation in the video. (The green equation has an added term for bonus creatures from blue's mutations.) https://www.desmos.com/calculator/zhig0dcftg x is the number of blue creatures, and there are sliders for the other parameters if you want to play with them. N is the total, and N-x is the number of green creatures. We can see that when there are 49 or more blue creatures we expect to gain green creatures and lose blue creatures. And if there are 44 or more blues, we expect to lose more blues than greens. But for any number of blues less than this, we expect green to do worse. Green is never expected to outnumber blue. This all assumes no oranges. The fact that the functions add to less than zero for any value reflects the fact that orange is gaining in any of these scenarios. So it's true that in this setup blue suffers losses from mutations to different colors while green benefits, but the magnitude of this effect with the given parameters is too small to make up the difference in base replication chance between green and blue. Additional sims I ran reflect this fact, with green regularly losing, but I chose to animate the first sim I ran because it was a good reminder of how chaotic this system is. I appreciate folks commenting with a critical eye, and this will help me know what to explain more clearly vs what to gloss over in future videos. It was also kind of fun to make the desmos graph. If you like it and want to see more desmos graphs when relevant to other videos, please reply to this comment and tell me so!
    2500
  5. This video contains a mistake! At 7:44, I show a graphic of the Australian Senate, which uses the "single transferable vote" (STV) method. This has similarities to instant runoff, but it's a multi-winner system, which does a better job of making room for third parties. The Australian House is what uses instant runoff, and the House is actually more polarized (at time of writing). If you'd like to learn more about STV, CGP Grey has a nice video about it: https://youtu.be/l8XOZJkozfI Apologies to Australian friends.
    2200
  6. Super cool. I'm really going to have to start using Shapes for my graphs.
    2000
  7. I thought about this, but a 5-simplex plot is a little hard to draw. :(
    1500
  8. Gosh, I hope not
    1500
  9. We are an ancient civilization. Great video!
    1100
  10. :)
    1100
  11. Well stay tuned for the next video (actually fairly soon)
    913
  12. Me too, me too. There's an econ one cooking currently.
    848
  13. <3
    787
  14.  @Alisu-  I think that's true for many comments, but an older version of this video had some serious comments about it, so I almost put it in the video itself. In the end, I decided a pinned comment was best.
    646
  15. A clarification: The standard term for what I called a "Strong Nash" is actually "Strict Nash". "Strong Nash" means something else, but it's only relevant in more complicated situations, so I hope it doesn't cause confusion here. But strictly speaking, I should have used "Strict" in this video!
    632
  16. This is valuable feedback, thank you. In the future when I want to be silly/irreverent in these videos, I’ll try to do it in a way that doesn’t make it harder for teachers to use them.
    602
  17.  @umair5602  Expand?
    477
  18. This is part of what is sent to the server. I'm excited to see how this shakes out.
    462
  19. Hey, it's bi-yearly
    455
  20. Thanks :)
    439
  21. :)
    432
  22. On some Android devices, the app may not remember your personal top scores if you close and open the app. If you see this, please comment here or on the play store listing to let me know what device you're on to help me track down the issue. If you do see this, your scores and drawing entries are still stored in the database, though.
    411
  23. It's coming!
    384
  24. This one's for you
    362
  25. I'm streaming now on Twitch. Come by to chat about the video or anything else! https://www.twitch.tv/justin_helps
    248
  26. Never
    247
  27. A few notes ........ First, Punnett Squares was not the best recommendation for a path to answering the pop quiz question. Hardy-Weinberg equilibrium would have been a better recommendation for something to look up. A square diagram involving alleles is very useful for the question, but the diagram isn't a typical Punnett Square, since it requires assigning uneven probabilities to alleles, as if the two alleles were randomly selected from the population rather than from two specific parents. ........ Second, a common question: Isn't aging good for the species, since older creatures get out of the way? Yes, it could be. This pattern would increase genetic diversity by allowing more new mutations to exist in a gene pool. And that genetic diversity would make the population more likely to survive environmental changes. But! Natural selection usually doesn't work at the group/species level. Even if aging is good for a population in the long run, the individuals with less aging would still have an advantage during stable periods and be favored by natural selection. It's the same in the videos about cooperation. Cooperating is good for the species, but defecting can still be favored by natural selection in some circumstances. Group selection is only a thing for tightly knit groups that reproduce together. Multicellular organisms are the best example of this. Each cell could be viewed as an individual, and the body as a group. Cancer is when cells decide to stop cooperating with the body and instead work for their own individual reproduction, at the expense of the rest of the body. Since bodies reproduce as a group, bodies that have mechanisms to enforce cooperation among their cells do better under natural selection. Even so, individual cancer cells don't really care about this, so cancer is a never-ending battle between the selection happening at the body level and selection happening at the level of an individual cell. It would be the same with ageless individuals, except the blob population doesn't even reproduce as a whole, so group-level selection would be much weaker if it exists at all. (Also, I would argue that longer-lived individuals aren't necessarily destructive like cancer and could even benefit the population in social species, especially in humans.)
    222
  28. This is mistaken. You are more related to yourself than to any of your descendants. You are the best vessel for propagating your genes. This is why the "never" allele won out in the first sim with "aging" genes, and it's why mutation accumulation and antagonistic pleiotropy are necessary to explain aging. Maybe I should have underlined that more thoroughly, since I already knew it was a misconception people had.
    213
  29.  @majorgnu  We added some checks on the server side. It's still technically possible to cheat, but the method of editing client-side memory to send whatever score you want no longer works. (We didn't just knock scores that seemed too high.) There's more discussion of how people are doing so well on the discord, if you're interested.
    212
  30. Thanks! I'm still building out the software part and figuring out the most efficient workflow, but videos should come out progressively more frequently as time goes on.
    197
  31. If you want to play with the system yourself. Here's a desmos calculator that does the pure-strategy simulation and puts it on a ternary plot. It's a little different because it assumes an infinite population, and so you will see some subtle differences in the results. https://www.desmos.com/calculator/lmx7hy2dai
    196
  32. They now have subtle movements instead of just standing still, and of course, they now wave sometimes. Glad the extra animation work was worth it. :D
    185
  33. Yes, saving will be in future sims/videos.
    178
  34. Thanks for the kind words. Gotta start somewhere!
    169
  35. :)
    149
  36. <3
    137
  37. Lmao
    111
  38. I have read the newest edition and remember that comment in the new foreword. For this video, though, I wanted to take advantage of the popular meme.
    106
  39. Yooo congrats on the hit video and channel growth. It always confused me what Honey's business model must be.
    103
  40. Tbell or bust
    97
  41. Thank you!
    84
  42. There's also a strong instinct for visual beauty, a tendency toward warm communication, and memory/empathy for what it's like to not know the concepts yet, which are what make 3b1b really stand out, IMO
    74
  43. If the NSA wanted to get the high score, I'm sure they could do it, but I'll guess it's not trivial. :)
    69
  44. Thanks
    67
  45. It's true. My hope is that this small intro makes more people excited to look further into it.
    64
  46. Thanks so much!
    54
  47. Not twice the number, but you're right, and I should have mentioned that!
    49
  48. Perfect
    48
  49. I didn't add a label for the mutation rate, but I did say that orange and green each come out of 1% of blue's mutations. So blue is still 98% as good as the written stats suggest.
    46
  50. Shh
    45
  51. ++ Thanks for spelling it out. :)
    43
  52. I also cannot wait
    43
  53. I appreciate it. Onward!
    42
  54. I'm late :(
    40
  55. Yeah, confronting people with this is the main reason for the game. Not trying to teach anything in particular with it except that you can't trust your gut for probability judgements. (of course sometimes that's all you really have in which case I hope you're well calibrated)
    34
  56. Mmmmmaybe. :)
    33
  57. That actually happened on the first run. Should have included it.
    29
  58.  @bitonic589  You don't need to submit a name or email. It will automatically store anonymous play data though. I did that because I want to collect it to analyze for the main video.
    28
  59. Thanks!
    27
  60. I was really close to putting that in the video
    25
  61. Yeah, it was weird, but the long-run results were fine.
    24
  62. I was more nervous about the jokes than anything else, so thank you. :)
    23
  63.  @DrAndrewSteele  It's looking like this may indeed be necessary.
    21
  64. Congrats on the great score!
    20
  65. I showed sims with similar R0 values because I was trying to show the importance of R0 to the final statistics, not characterize the effect of different inputs.
    18
  66. Me too! :D
    18
  67. Well don't I look silly. Turns out it was somewhat trivial to edit the data before it was encrypted and sent. But now there are checks on the server side. So NOW I think my original response is valid. Then again, I've thought that before.
    17
  68. Yes! I was actually planning to put that in this video when I was originally outlining it, but I ended up deciding to split it up. It'll be in the next video or the one after.
    17
  69. Haha, thanks for the enthusiasm. Rest assured that more are coming. I'm getting impatient too. :P
    17
  70.  @bitonic589  No, the app pulls from the database every time. Has to work that way on web (afaik) and didn't want to restructure for local saves on android.
    15
  71. Hey hey. Bi yearly.
    15
  72. Thanks!
    15
  73. Thanks!
    14
  74. I will try. >_<
    13
  75.  @loopeater8338  Patreon early access
    12
  76. Thanks, I really appreciate the kind words.
    12
  77. More affluent countries tend to have stable populations, so fighting poverty happens to also be a way to fight future crowding, as if it weren’t already a win. Humans are way smarter than these blobs, and if you ask me, might have a shot at reaching a stable population by choice rather than having it forced upon us by fundamental carrying capacity limits.
    12
  78. Thanks for the sub! For now, I'm planning to keep using sims as a tool but not make any videos directly about the sim methods. It's an interesting idea to make a video about that, though, since building the sim is a great way to deeply understand the system. I'd be interested to see an implementation of this that uses a Markov chain. If you end up building something like this, it could be cool to discuss on the subreddit. If you want to look at my code, it's on github: https://github.com/Helpsypoo/primer. I warn you, though, it's a bit spaghetti and doesn't have a readme yet.
    11
  79. Yes, very enjoyable
    10
  80. I have no MnMs
    10
  81. Runaan!
    10
  82. I would mind of you didn't!
    10
  83. Lol, you're ruining the illusion that nobody notices little things like that.
    10
  84. I used a different process this time. The peaks are still zero db, but the perceived volume is a bit low. Appreciate you giving the feebback.
    9
  85. Thanks!
    9
  86.  @Aarush.A.S  Could you clarify what you mean?
    9
  87. hey
    8
  88. Yeah, that would have been good.
    8
  89. It's a public domain 3D model. I didn't make it, and I don't speak Portuguese. :(
    8
  90. Well that's the main thing, at least.
    8
  91. :)
    8
  92. Help support the channel by trying NordVPN free for 30 days at https://nordvpn.com/Primer
    7
  93. There is a (painfully disorganized) github repo in the description.
    7
  94. I chose a high mutation rate. In the sims shown, I made it so every trait mutates every time a creature reproduces. This isn't super realistic, and it's part of the cause of the fact that you don't see super stable clumps, but it since I was going to animate this, I wanted the population to be able to explore a wide variety of strategies in a shorter period. Natural selection doesn't affect the mutation rate in this sim. In the real world, natural selection would affect mutation rate. I don't know enough about the evolution of real-world mutation rates to comment about organisms. For a replicator, a mutation is as good as not replicating at all, since you produced something that's not actually a copy of yourself (said in more detail in my video on mutations). I think it's reasonable to say that DNA's relatively low mutation rate among possible self-replicating molecules is part of why it's at the core of all living things.
    7
  95. The style is very much inspired by 3b1b, and your saying it reminds you of that channel is a great compliment.
    7
  96. Thanks <3
    7
  97. Hey hey, I've been managing two!
    7
  98. Thanks for the thoughtful feedback. I do wish I had made equilibrium and the equation more concrete by utilizing some of the visual space to actually show creatures or remind folks of the meaning of the variables while discussing the equation. Lessons for the future. Thanks again!
    7
  99. Thanks!
    7
  100. A fair player has a about a 3% chance to go 5 in a row, but a cheater only has 24%, so about 1 in 9 times you see a 5:0, it's fair player.
    7
  101. Thanks ❤️
    6
  102. no
    6
  103. I'm jealous Hank got to say "A jet ski is very fun, Pete Buttigieg" to Pete Buttigieg
    6
  104. Hard to argue. For this particular video, there aren't direct sources, since the sim is original, and the illustrated principles are general knowledge gleaned from many sources through my education and broad reading. But I added a few links for further learning which I hope will be helpful. Thanks for the push.
    6
  105. Thanks for the thoughtful comment. You’re right to extend that dynamic to the real world, though there is some added nuance that the video didn’t touch on. Buyers and sellers can usually buy/sell more than one unit, so it’s not strictly about the number of buyers and sellers. What matters is that there is competition on both sides of the market. For example, even if there are millions of travelers and only, say, 10 airlines, as long as those airlines can produce enough travel seats to meet demand, and they aren’t able to cooperate with each other (whether by low or just because one of them will always defect), you’ll still see competition that benefits the numerous consumers.
    6
  106. Thank you!
    6
  107. Thanks! I hope you're doing well. :)
    5
  108. Thanks!
    5
  109. Feel free to tell your friends. :)
    5
  110. True, but I didn't go from the linear curve to the exponential curve with an antiderivative. The delta equation is a differential equation relating the growth rate to N, not an equation giving the growth rate as a function of time.
    5
  111. Thanks!
    5
  112. I wanted to do that but ran out of memory in the janky way I was making the graphs. :X
    5
  113. Thank you!
    5
  114. Sure I do. But there are too many to read them all.
    5
  115. Thanks, eh.
    4
  116. thanks
    4
  117. It's updating based on data for all the interactions between blobs. It converges to one value over time. R0 should be constant, though in real life it might change if things other than the virus change. Like the weather, or behavior.
    4
  118. It exists in the fuUUuuUUutuuure
    4
  119. You're right that I should have been more explicit that replication is a kind of "being born". And you're right that I didn't give a satisfactory answer to the question of why things exist, but I hope it was a useful mechanisms for framing a comparison between living and non-living things. Thanks for the honest feedback!
    4
  120. :D
    4
  121. All of the buyers are the exact same color, so nobody can tell them apart!
    4
  122. :) It already is. But we're not going to bust out any calculus. That's why I linked to Khan Academy instead of showing the details of how the exponential came about.
    4
  123. They reproduce asexually, so they just need the energy (and to not die)
    4
  124. :) I wouldn’t presume to usurp 3b1b!
    4
  125. Thanks! I have that hope as well. The simulations and related animations were a lot of work and accounted for much of the delay, and they'll be reused in the next few videos, so the next video shouldn't take quite so long.
    4
  126. That's a fair thing to want. It would mostly be relevant if we were modeling behavior in detail, though, which is super hard (beyond current human understanding). In this basic model, it would make no difference because the blobs don't react to the disease.
    4
  127. This video doesn't intend to replace any existing explanations. It's meant to lay some groundwork for a way of analyzing biology. I know the topic itself is touchy, but I hope to present things in a way that lets people engage with concepts without asking them to change their entire worldview. If I fell short here or anywhere else, I'd love to hear your feedback. Thanks for engaging.
    4
  128. Yes
    4
  129. I do want money, though
    4
  130. Thanks!
    4
  131. 100%
    4
  132. Nature has physical laws, which are rules
    4
  133. Yes, which is the basic reason reality hasn't played out quite like this. It's very hard to model those effects, so I focus on the base case and leave the actual predictive modeling to real epidemiologists.
    4
  134. I had one once
    4
  135. :(
    4
  136. You know, I think you might be right. I'll play around with it. Thanks for the feedback!
    4
  137. Also, I like your account picture. :)
    4
  138. I love letting the physics-enabled things hang around. First did it with mangoes, I think.
    3
  139. That is another name for instant runoff, though the exact implementation can vary.
    3
  140. I aged a full year
    3
  141. Genes don't change mid-life, so this would require a more complicated model with more conditional behavior, but those do exist, and there will be a video eventually.
    3
  142. I'm glad you clicked. :)
    3
  143. Thanks for sharing your story. <3
    3
  144. Haha, thank you. I found it encouraging that complaints in the comments were balanced between the left and right varieties of people who treat econ as a religion.
    3
  145. I guess don't buy it then. And yes Markiplier and I are in the same situation, yes.
    3
  146. Blender and python. There's a github repo in the description.
    3
  147. I would recommend Khan Academy.
    3
  148. You're like six months late where have you been.
    3
  149. Good idea
    3
  150. Did you watch all the way???
    3
  151. A really good question. You still need that first creature for reproduction to be possible. As the story goes, the first replicators were pretty simple. Simple enough to form by chance in the right circumstances. Something like simple membranes or protiens that could catalyze their own formation. Then, because the replication isn't perfect, it would sometimes result in a new kind of replicator that is too complex to have formed by spontaneously. Keep this going, and you eventually get things as complex as rabbits. I think the details of what those first replicators look like is up in the air (I should look this up).
    3
  152. Boo
    3
  153. I am, but I'm also addicted to the YouTube Studio app
    3
  154. I would like that too. :)
    3
  155. Not planning to directly come back to the negative number part in future videos, but here's an attempt at explaining it. The true equilibrium number can only be positive, but nothing about the setup of the mathematical model requires the number to be positive, so it is possible to get a negative number with the right parameters. In this case, the negative equilibrium number is a clear sign that our model is departing from reality, so we need to rethink it (and this re-thinking will be done in future videos). Though now that I type this out, maybe a more direct discussion of modeling would make sense in the next few videos. Thanks for the thought-provoking question!
    3
  156. Someday
    3
  157.  @themanthemyth4455  Thanks. The initial takedown of cheaters was mostly manual swatting. We added some validation on the back end that seems to have gotten yours too. :)
    3
  158. Things are slow right now, but the channel is just getting started.
    3
  159. Getting me excited for the bayes rule vid
    3
  160. The blob eyes? They're spherical meshes.
    3
  161. Simply put, you are not your genes. Also, see the next video on inclusive fitness
    3
  162.  @globaldigitaldirectsubsidi4493  Also true. It's a very important part of understanding ourselves, but it's not the only useful concept.
    3
  163. Clockwise if seen from above
    3
  164. You're not alone! Music in more recent and future videos is quieter.
    3
  165. Ah, I think I see what you mean. In this video, the “births” are just creatures popping into existence from nothing. This, of course, is not how biology works, but it was a simple place to start thinking about equilibrium. The next video talks about replication, where the overall birth rate then depends on the birth rate per creature.
    3
  166. Hey. It's been less than 12 minutes since it went up! But thank you!
    3
  167. A fair question. Any model has assumptions. This one assumes perfect competition. That's admittedly an idealization to simplify the initial analysis. Sooner or later (later, unfortunately), I intend to make one about monopolies.
    3
  168. It's a pretty yellowy orange, I'll give you that.
    3
  169. True, but the effect is small enough to not overpower the better underlying replication rate. Running more sims, green consistently loses, but to decided to animate the first sim for the lesson about surprising results.
    3
  170. :)
    3
  171. Glad you like them!
    3
  172. There are several others, amd there are viewers around the world. Merry Christmas to you.
    3
  173. True. And my savings financed it. 🤷‍♂️
    3
  174. You're right. The old video had a pretty serious mistake in it, so I decided to remake it. And the mistake was actually regarding Hamilton's rule, which will get its own video soon (soon in Primer time)
    3
  175. :0
    3
  176. I'm completely with you. I tried to communicate that this is an idealized model.
    3
  177. 2 max, then home
    3
  178. Oh dang. I can’t believe that slipped through. Thanks for pointing it out!
    3
  179. Thanks!
    3
  180. I appreciate this point of view. I'll stay coy about my own beliefs, but I very much dislike how communication around evolution often implicitly asks people to give up their whole worldview before they can understand a new idea.
    3
  181. Thanks, and I appreciate the willingness to give criticism. In this case, I meant the quick text as a slight joke about how I'm being pedantic about pokemon, but I can see how it could seem like something more important before you get a chance to read it. Thanks again!
    3
  182. Planning to get there over time, but laying the basics down first.
    2
  183. Blobs are weak and not very smart... so far
    2
  184. That is good feedback, thank you.
    2
  185. Its latin root is something like "first" and in English it has various meanings related to getting started.
    2
  186. Yeah, this one is more complex. Had to do it because of the previous mistake. Planning to get a little lighter next.
    2
  187. It's the Unity game engine
    2
  188. Tricked you!
    2
  189. I do that, but at the time that video was published, the was much smaller.
    2
  190. You're right that natural selection can lead to something good for the species, but the good of the species isn't the driver of natural selection. The more recent video on The Selfish Gene touches on this, and deeper sims of altruistic and cooperative behavior are coming!
    2
  191. This is a good question. I would like to make a video where new structures arise rather than just tweaking parameters. It will be a while, though. I believe the channel Stated Clearly is working on a video related to "major transitions" such as multicellularity, so could be worth subscribing to that and keeping an eye out.
    2
  192. The next one will be an evolution one, but there are more econ vids coming.
    2
  193. Godot
    2
  194. I did frame it as an overstatement. It was just more of an overstatement than I realized at the time.
    2
  195. "one of them will notice the predator" :)
    2
  196. It has only been up since Sunday.
    2
  197. You and me both
    2
  198. Yes. It's a built-in behavior that isn't allowed to vary (just like the desire to eat and go home afterward).
    2
  199. That is not what this shows.
    2
  200. What you described is natural selection. Certain traits give advantages depending on the environment. You're right, though, that it doesn't capture the full complexity of real organisms.
    2
  201. Blender and python. There is a github repo in the description.
    2
  202. See the next video
    2
  203. You may indeed suggest that, and I just may comply. (with time!)
    2
  204. The seller offers, and the buyer chooses whether to accept.
    2
  205. How many moneys are we talkin here
    2
  206. Custom made for the video. You can find the composer in the description.
    2
  207. The best bot I could make went infinite with a more generous reward structure, but couldn't do it with this one. So I think infinities are impossible here, but I haven't proven it. Nice job on the 146!
    2
  208. I did know that, yes
    2
  209. There's a version of instant runoff where you don't have to put a ranking for everyone. All systems turn out to have flaws. There are actually some math proofs about it. You could look up Arrow's theorem, as the most famous one.
    2
  210. Python and Blender. There's a github repo in the description.
    2
  211. I appreciate this.
    2
  212. You're right. That's why I say "*other* creatures are 90% doves". Correcting for the fact that you're a dove and can't face yourself complicates the expression quite a bit and requires you to know the total number of creatures. And for larger populations, the difference becomes negligible. I debated going through this in the video but in the end decided it would be more of a distraction from the main point. And you're definitely not dumb. Noticing that it doesn't make sense to you and facing your confusion instead of sweeping away the discomfort might be the biggest part to being "smart". Keep doing it!
    2
  213. And you might like conversations on the subreddit.
    2
  214. Yes
    2
  215. But the consequences of such a difference in terms of budget allocation is huge. The expected value in $ comes out to be decent, though I don't have the numbers at hand.
    2
  216. Definitions are valuable, and I did acknowledge that this is only one possible definition.
    2
  217. Thanks. I try not to ask people to question their entire worldview just to learn about something cool.
    2
  218. Blender and python. There is a github repo in the description.
    2
  219. They haven't invented pockets :(
    2
  220. It's on the list
    2
  221. I hear that
    2
  222. :)
    2
  223. Thanks for the thoughtful comment. Developing any structure has a cost. Having a cost is inescapable. The benefit depends on the environment, and isn't certain. Gills are very useful but are no help on land (but always have a cost), for example.
    2
  224. But then the way to win becomes to persuade the public of your ideas. Seems good to me, even though my views aren't really at the center of public opinion.
    2
  225. There is a github repo in the description
    2
  226. Ͳ They actually do stop eating at two food.
    2
  227. It's a public domain model that I didn't make.
    2
  228. The trees can only grow the mangoes so fast, so the environment has a finite carrying capacity. No deaths from aging would tend to increase the population size, but it doesn't go so fast that the trees get overwhelmed.
    2
  229. Blender and python. There's a github repo in the description.
    2
  230. But you don't need a majority
    2
  231. Unity
    2
  232.  @carolepetit6741  Not a worry. Explained it in two different ways in different parts of the video and didn't link them. It was just unclear, and others have had similar comments. I appreciate your thoughtfulness and willingness to communicate. Folks like you help make better videos possible. :)
    2
  233. I regret the snark. Just trying to make the point that more than one phenomenon can be at play.
    2
  234. Earlier versions actually had the predator number vary depending on how many blobs they successfully ate, but added another variable to the sims that wasn't relevant, so I turned it off this time. In some of the early test runs, they ate too many blobs and made them go extinct, then they went extinct themselves. :(
    2
  235. Those populations do seem to oscillate, but that's just a consequence of the randomness. I think if you double check the differential equation, you'll find that it always leads to an exponential. And when the death rate is higher, it's an exponential that decays toward a constant value (which is zero if there's no spontaneous birth). Perhaps it would have been better to show that instead of just giving the result when it starts at equilibrium, not sure. If it were a second-order equation, we'd see some oscillation.
    2
  236. My background is also in physics, so I'm with you that this is a great kind of question to ask. In the case of projectile motion, you end up realizing that Newton's laws work just as well if time is going backward. I'm not aware of a meaningful interpretation of a negative count of organisms, though. It's a bit like how we demand observables in QM be real numbers. Interestingly, though, negative temperature is a real in rare cases when adding energy reduces entropy, and it's mathematically similar to this case. Could make an interesting video itself at some point.
    2
  237. Yes, but would it be worth it?
    2
  238. I didn't mean to make a reference, but oh yeah, people did notice it
    2
  239. I don't believe you
    2
  240. This used to bother me a lot too, but affluent societies tend to have fewer children by choice, so if we solve poverty, it will go a long way toward solving overpopulation. That's not simple, nor will it solve everything, but it gives me hope.
    2
  241. A fair question. In science, our observation of nature is the ultimate arbiter of whether a theory is true or not. Simulations are useful because they help us understand the consequences of a theory when it's too complex to calculate everything precisely. And once we know the consequences of a theory, we can compare the theory to nature. If evolution works by natural selection, we should see examples of the phenomena from this video in nature. It turns out that we do, but this video doesn't show any of that, so by itself, this doesn't prove anything.
    2
  242. I love all your sims on Khan Academy!
    2
  243. I did mean to put it jokingly, but I was genuinely confused.
    2
  244. Someday
    2
  245. Good question. The frequency of a given action ends up being the same, even if single blobs have a mixed strategy.
    2
  246. I know, I sowwy
    2
  247. You can find the composer in the description!
    2
  248. Yesterday, the top nine scores were over 2 billion. We added some consistency checks on the server side and seem to have bonked all the cheaters. The top scores now are believable with a good strategy. Feel free to drop by the discord if you want to know more about what people are doing. :)
    2
  249.  @jebeda  Something you might want to know with this strategy is that payment processors typically take a minimum, which is in the tens of cents range. So smaller donations are hit harder by those fees, leaving the creators with less. I wish it weren't this way, since I would love to take your strategy of small donations to many creators. But I choose to support only my favorite creators with larger amounts because of this. But to each their own!
    2
  250. .... the coin is MAGIC
    2
  251. Yeah I'm surprised how many there are
    2
  252. There will be more econ videos, and those may be brought up as examples again, but I don't think I'll attempt to model any particular market in detail. Too easy to get wrong and oversimplify. I just hope to illustrate some of the powerful concepts at play.
    2
  253. My lack of 3D models skills guarantees that
    2
  254. You
    2
  255. 1/R0 is the expected fraction of the population that will be susceptible. To get the fraction that will be recovered or infectious, you can subtract the susceptibles from the whole population. 1 represents the whole population, like 100%, so the total fraction infected is 1-1/R0.
    2
  256. I was thinking of all the blobs as one species, but you're right that that's not the only way to define things. Especially since they reproduce asexually and don't mutate in this video.
    2
  257. Except blues
    2
  258. Oh stop
    2
  259. Check the latest video
    2
  260.  @NOL_bm  Sounds like you've been unlucky, but probably less unlucky than it seems. A fair blob only has about a 3% chance to go 5:0, but a cheater only has 24%. So about 1 in 9 times you see a 5:0, it's fair. So unlucky, but not crazy unlucky.
    2
  261. Python and Blender. There's a github repo in the description.
    2
  262. Great comment, great username
    2
  263. Yep. There is a github repo in the description.
    2
  264. It doesn't sound like you have either a robust model or data for your view. I think we'll be fine.
    2
  265. Thanks for the thoughtful feedback. I'll definitely use the visual space to keep the context of the equation/variables present in future videos.
    2
  266. The way reproduction works in this sim, negative numbers aren't actually worse that zero (which is already certain doom), so I didn't explore negative values. But in general, you're quite right. There might be future videos about war where the hawk-hawk payoff is sharply negative, though.
    2
  267. It is literally every second. Yes it's bad!
    2
  268. It kinda is, I guess
    2
  269. Python. There's a github repo in the description.
    2
  270. When the expected price is below their minimum, they don't put out a rocket.
    2
  271. Blender and python. There is a github repo in the description.
    2
  272. ok
    2
  273. My computer heats my apartment, if that answers your question.
    2
  274. I am pretty fond of the prototype version. :)
    2
  275. No
    2
  276. The simulation is performed in Unity. It's basically a video game with no inputs.
    2
  277. Okay. For you.
    2
  278. Right
    2
  279. I do intend to post the code to github. For now, it's so spaghetti that I don't want to inflict it on others, but over the next few weeks, I'll find time to clean it up slightly and share it.
    2
  280. Ok
    2
  281. Did you see the part where I say it's actually more complicated than that?
    2
  282. The main difference that matters is that you're sometimes tempted to be dishonest about your first choice.
    2
  283. The video says it's only partly genetic. Not a genetic determinist here.
    2
  284. You're right that this is far from the whole picture. It's an idealized case and only meant to help people understand one particular part of the larger debate about how we organize society. And that's why the last few minutes question whether the benefits of even this idealized model are the right goals. You're also right that power tends to become concentrated in the real world, which breaks the model.
    2
  285. Yes, that's exactly what I did.
    2
  286. Me too
    2
  287. Interesting question. Like many categorization systems, a "species" is more of a mental tool than something that's clear-cut in the real world. I did try to find parameters that would cause distinct groupings of creatures, such as by size. If I had been able to produce that, I would have mentioned speciation here and labeled the groups as separate species, but it seems this environment is just too simple to produce that. One possible way to produce a clustering like this would be to add safe zones that small creatures could fit into to "go home" at the end of the day (like climbing a tree), making it easier for them to avoid being eaten, and giving small creatures an advantage other than simply saving energy. For this video, though, I didn't think exploring that idea would be worthwhile. We'll talk about speciation in a future video, though.
    2
  288. The ages are spread out at the start. (The video says so)
    2
  289. In English, "primer" has many meanings involving getting started.
    2
  290. DRAAAAAAVeeeennnnnnn
    2
  291. If the system is in equilibrium, with a constant total population, that equation must be true. It's a starting point.
    2
  292. Close-minded fools, the lot!
    2
  293. I make them in Unity. There's a link in the description.
    2
  294. Over 70% are in the 18-34 range on this video so far, according to the analytics page.
    2
  295. Well in that case, the blobs will ship in time for any winter holidays you might celebrate. :) I want to defend myself a little by saying it's far from just my country that is experiencing winter right now, but I appreciate the reminder all the same.
    2
  296. Selection occurs at the individual level, but individuals don't change genetically during their lifetime. It's the population as a whole that changes (evolves) over time.
    2
  297. Great question. It’s a somewhat strange situation where the underlying value is speculated rather than a subjective fact (unless there are people who want to have shares of a company just because they like the company). Both sides can’t be right about the future value of a share, so the stock market is zero sum and doesn’t actually generate net surplus for traders. But there is value in the activity of evaluating companies and buying up newly issued shares, which is how companies get their initial funding or funding to expand. Perhaps this would be a good topic for a video!
    2
  298. Because THREEEE DEEEEEE
    2
  299. Ah, how foolish of me not to have done that already! Done.
    2
  300. Thanks!
    2
  301. This is the second
    2
  302. 👀
    2
  303. 3blue1brown is a channel that animates math. Very worth checking out.
    2
  304. I get excited when I see you comment
    2
  305. This video is limited to single-winner elections. Which, of course, are silly.
    2
  306. They chose to be ignored
    2
  307. Idk, bots probably
    2
  308. :)
    2
  309. I made it myself. It does not have a name.
    2
  310. Same to you, sir
    2
  311. They start at a random age
    2
  312. Good question. Regardless of the number of participants on each side, it's really about the ability for one side to cooperate (and stop competing), which is harder when the numbers are large. And these blobs are too simple to scheme together, so we didn't really see the requirement at play in the sim.
    2
  313. For simplicity. Even without it, creatures still have an expected lifespan due to probability, so it didn't affect the points made in the video. A more realistic sim would have this for sure, though.
    2
  314. It's true that I stopped the animation at an advantageous fluctuation. The qualitative result remains the same, though. I plead illustration-not-proof! But I do appreciate your critical eye.
    2
  315. Yep, that's mostly how it works, except the creatures head home when they're almost out of energy, and only if they've eaten at least one food. They have to get all the way to the edge before running out of energy, or they just stop and die. The value is a bit arbitrary and chosen to make creatures with stats all equal to able to explore some but not all of the map (whose size is also arbitrary). The starting number of energy units in the sims shown is 800, if you want to know.
    2
  316. Those are bugs in the animation! I'm surprised more people don't seem yo notice them.
    2
  317. There is no machine learning. Just simple, static behavioral algorithms with varying parameters, which are selected. There is a github repo in the description.
    2
  318. Sorry it took me so long to catch up
    2
  319. I majored in physics, which probably informed a lot of my thinking on modeling. Most of my programming was learned after college, though.
    2
  320. I'm glad!
    2
  321. That could be the case, yes.
    2
  322. Thanks, Sara!
    2
  323. True. I was careful to say "reliable" because of this.
    2
  324. Takes time and interaction to really learn this stuff. So don't worry if it's confusing the first time you hear about it.
    2
  325. Right. The narrow definition of surplus doesn’t capture the value we rightly put in the ability for everyone to eat, among other things. This was what I was getting at during the last part of the video that questions whether maximizing surplus, as defined, is actually the right goal in all cases.
    2
  326. Turn to the math side
    2
  327. Studied physics in college and materials in grad school.
    2
  328. Thanks!
    2
  329. No, it's magic
    2
  330. Preach
    2
  331. Oh, stay tuned!
    2
  332. Kurier
    2
  333.  @Bunnyquazel  No deep reason. My first approach was to consider the sim happening in discrete steps, which fit with the idea of day and night. This ended up working out, so I didn't change it. A continuous sim should work too.
    2
  334. goodbye
    2
  335. That's actually the reason I made the game. To help people experience how statistical tests have error.
    2
  336. No spoilers!
    2
  337. Hehehe
    2
  338. Mango
    2
  339. rip
    2
  340. You know, I just might eventually. I'd probably combine it with a quadratic pricing econ video. Several steps away from that for now, though.
    2
  341. Python. There's a github link in the description.
    2
  342. Lel
    2
  343. Genes can also be selfish at the expense of the group. That partial comment in this video foreshadows the more conplete discussion in the selfish gene video.
    2
  344. Noted. I'll make sure of it.
    2
  345. Hang in there. Nerds are happier in the long run.
    2
  346. Not sure this has much relevance to politics
    2
  347. Imagine replying to someone because it was 33 minutes ago
    2
  348. oof, my bad
    2
  349. In the description
    1
  350. No, but it does sound like that!
    1
  351. There was certainly some hand waving at parts. But do you see the simulations as hand waving? But in the end it's also true that this is a YouTube video, so I can't be fully rigorous about everything, which is why I try to stick to well established concepts (which these are).
    1
  352. There was a lot of work done by Nvidia chips in the making of this, if that's what you mean.
    1
  353. I animated just about the min number of days to see the effect. If this were an actual research project rather than an illustration, I would have made everything longer and shown more repetions, etc.
    1
  354. You could use the concepts from the video to come up with a prediction!
    1
  355. I know! I. am. Noob.
    1
  356. See the pinned comment
    1
  357. It will be similar but different
    1
  358. It's kinda like a video game with no player. So learning game development might be helpful. Not sure what else to tell you.
    1
  359. I will likely do both of those sooner or later. I had actually not ever heard of LBRY. Looks interesting, thanks.
    1
  360. See the pinned comment
    1
  361. Thanks! I love your channel.
    1
  362. Glad you enjoyed it!
    1
  363. See the pinned comment
    1
  364. Yeah, definitely a rough one for production quality. Future talking head videos will be better (and rare).
    1
  365. wdym it was sunday
    1
  366. Ah dang, you're right. Thanks.
    1
  367. Oh my god (maybe) this is an amazing comment
    1
  368. They still haven't reached out :((((
    1
  369. It's true, a more complex system like this would be better. But this is just to illustrate the statistics of the simple test.
    1
  370. Gotta start with something. Bayes is next in the series!
    1
  371. That actually happened. Maybe I should have included it.
    1
  372. Well, I hope the six year olds watching get off on a strong statistical foot!
    1
  373. Your delivery is the best so far, though.
    1
  374. My biggest inspiration for this series was "The Selfish Gene ". It's not super technical, and it doesn't focus on modeling, but it gives a good grounding focused on statistical reasoning. I hope that helps.
    1
  375. Intervention cam definitely have negative effects and produce a downward spiral, but that's not always the case, and I hope this video encourages people to think in a nuanced way case by case.
    1
  376. It's unfortunately not very sharable in its current state. Several others have asked for something similar, so I hope to make it shareable at some point.
    1
  377. OMG. That must be why this video is doing so well.
    1
  378. I used the same base simulation in a different video, and I've also mentioned Hamilton's rule before. This one follows from those.
    1
  379. There are several books mentioned across the different video descriptions. I don't remember them all off the top of my head, but The Selfish Gene is a big one. For the simulation part, I use Unity C# for the animations, but you could use pretty much any coding language. Python is pretty widely used, so it's a good first language if you're starting out.
    1
  380. And good luck on your journey!
    1
  381. Blender and python. There's a github repo in the description. It's ridiculously hard to get running, though.
    1
  382. It doesn't work out that way. Someone asked this on the subreddit, and I explained it there.
    1
  383. I know, it's spooky
    1
  384. There is Patreon if you feel so compelled. Thanks for the kind words!
    1
  385. thanks
    1
  386. I didn't talk about this, but the predators actually survive and reproduce depending on how many blobs they eat. That's actually why the first control sim has waves in the total blob population. So the equilibrium number of preds was probably higher in the 10% case. Should only affect the amount of selection pressure, not the direction.
    1
  387. Same mic, actually. I just had a better recording environment and more experience processing sound.
    1
  388. I'm just very distractable. :|
    1
  389. Not true
    1
  390. Yep, they're real simulations. Made with Unity.
    1
  391. Enjoy your hellscape of rampant cheating <3
    1
  392. Word
    1
  393. I just took it down today, though
    1
  394. :)
    1
  395. The sad thing is, that's a proven impossibility
    1
  396. Hey, thanks for the thoughtful comment. Racism is a complex human phenomenon which is not modeled well (or, really, at all) by this simulation. And while some people (historically but even, sadly, presently) might invoke natural selection to justify racist views, they are simply mistaken.
    1
  397. It still applies, but things get a lot more complex
    1
  398. Order of operations
    1
  399. Yeah, this one is too complicated.
    1
  400. There's a github repo in the description
    1
  401. I'm a Nintendo boi anyhow.
    1
  402. Oh, the animated sims don't run in realtime. I fix the game time per frame and just let them take a few hours when I'm rendering the animations.
    1
  403. Definitely going down this road
    1
  404. This distribution is useful for discrete situations, so not directly applicable to wave functions over space. But probability distributions in general are very relevant to quantum, and the binomial distribution does relate to the gaussian distribution in a way I don't precisely remember at the moment. :)
    1
  405. The coding, by far
    1
  406. Python. There's a github repo in the description.
    1
  407. There's a github repo in the description
    1
  408. Yeah I wish I could have worked in a less extreme version of altruism. The death penalty is so easy to explain, though. And reciprocal altruism is coming!
    1
  409. Yep, this is far from explaining human behavior
    1
  410. Certainly, but biology is also part of it, so it makes sense to understand that part even while acknowledging it's part of a larger whole.
    1
  411. I didn't want to get hungry for taco bell while making it, but here we are.
    1
  412. You're right that we are not our genes and that we can act "against" them. Contraception is a very direct example. But they do shape our development and in turn affect our traits and behavior greatly.
    1
  413. It doesn't repeat exactly, though it often finds a stable kind of state. Sometimes several stable points exist.
    1
  414. I'm glad someone noticed. :)
    1
  415. I only have 18 videos
    1
  416. There's more evolution coming!
    1
  417. Possibly. Certainly if it's misleading, but advertising can also give true information. Much of it is aimed at influencing emotions, though, leaning into the fact that people aren't actually rational consumers, which is another thing missing from this simple model.
    1
  418. I dream of this day also
    1
  419. The sim and animation are separate, but the animation is also handled with code.
    1
  420. It provides the max total surplus, but not the max for each individual. Some are left out of the market and get no surplus. So if a society wants everyone to be included, the system might not be best.
    1
  421. Yes, the point is to make it possible for an option C to enter the race.
    1
  422. They don't know. The temptation theoretically exists, but we'd need to see how it plays out in real systems. I share your skepticism that it would be a big deal, since people being too risky with their vote is the opposite of what we see with plurality.
    1
  423. They are just lil guys
    1
  424. I just came here to see what ugly and stupid people are into
    1
  425. It does sound fun.
    1
  426. Thanks. When you do, I'd request that you argue in a way that doesn't claim this is proof that atheists are right, but just something that might help them better understand a concept.
    1
  427. I actually agree with this. If you have any suggestions that have a Creative Commons license, I’d love to hear about them.
    1
  428. :0
    1
  429. It is skippable by using YT's innovative new seek feature. <3
    1
  430. It's just a spiral
    1
  431. Absolutely no pressure. Gotta look out for yourself first!
    1
  432. wut lol
    1
  433. I've been out for a long time. I just do some arams every now and again.
    1
  434. They'll only get stronger at fighting, which doesn't seem worth the sacrifice to me. 🤷‍♂️
    1
  435. There is a middle ground between full industry takeover and completely unfettered markets. To find the right policies, we need to think carefully about the trade-offs at play in each case, study it empirically, and reflect on the values of people.
    1
  436. Binary outcomes are so nicely simplified. ¯\_(ツ)_/¯
    1
  437. 42
    1
  438. Intuition is wonderful until it's not. :)
    1
  439. I do too, so I hope not
    1
  440. There is a github repo in the description. It's not super user-friendly, though. :/ (So far!)
    1
  441. There will be more econ videos, and this will be touched on. I’m not sure I’d agree that injustice is fundamentally built into our definition of economics, but I’ll wholeheartedly agree that the systems and institutions we’ve built up so far have often been quite unjust. And much of the reason I’m doing econ videos at all is to make it easier to face that and come up with constructive solutions. Thanks for the thoughtful comment!
    1
  442. There will be more evolution, but also more econ, and eventually other things. My goal now is to spark your interest in other topics. :)
    1
  443. There's one upcoming that will say "this might make someone conclude that Hamilton's rule is a lie!"
    1
  444. Blender and python. There's a github repo in the description.
    1
  445. Yeah I can't believe I missed that.
    1
  446. It's done in python, and there's a github repo in the description
    1
  447. Give me money
    1
  448. They do eat three in the simulation before altruism is introduced. But once altruism is introduced, one blob always notices the predator, and at least one of the three gets away. I could have made it so the blobs had a chance (not a guarantee) to see the predator. That would have been more realistic, but the goal was to focus on the difference between the altruistic and cowardly behaviors.
    1
  449. I always appreciate the willingness to give criticism. Energy per time is indeed quadrupled. But energy per distance is only doubled. I pointed out that relationship because it’s most relevant to the ability to find food. And I intentionally used “size” because it’s a commonly used word, and this isn’t a physics simulation. Does that satisfy you? Again, I appreciate the constructive criticism.
    1
  450.  @JayLikesLasers  I hear that peeve, and it did bother me slightly too when I was using "size" as length when talking about its relationship to volume. But keeping that seemed preferable to switching away from the more generally used "size" for the rest of the video. Rest assured that when geometry or physics is core to the concept being explained, I'll use more precise language. My background is actually in physics.
    1
  451. If I had to guess, it's because I'm just talking about this stuff because I think it's cool and not trying to stress you out. Or maybe just blebs.
    1
  452. Praise the algorithm!
    1
  453. Good feedback! It is ever a balancing act to strike the pace, so thanks for sharing your take.
    1
  454. Right, this phenomenon isn't at play in humans. But we'll get to some that are.
    1
  455. FINALLY
    1
  456. Python
    1
  457. They are simulated with code
    1
  458. Miss you. Q_Q
    1
  459. Blender and python. There's a github repo in the description.
    1
  460. That is what happens to energy per time. Energy per distance ends up only being doubled, though.
    1
  461. I said them in opposite order of the animation at the end there, yeah. :/
    1
  462. Funny how the simulation comes out the way it does without mentioning those ideas. It's as if they aren't related to the structure of elections. Those ideas might be at play in the real world, but I think you're making an error by trying to project them onto election science.
    1
  463. I appreciate the kind words!
    1
  464. Hey, the next video is out, and I wanted to let you know that your question here helped inspire the approach of that video. The new video plots expected change vs number of creatures, and equilibrium exists at the x-intercept (or N-intercept) of the curve. In one of the cases covered in the video, the intercept is negative if you include negative values in the domain. So there's a graphical way of looking at negative equilibrium, even if we haven't found it to correspond to anything in particular in the system being modeled. Thanks again for the thought-provoking question!
    1
  465. Got you
    1
  466. Well now you have something to look forward to
    1
  467. This channel is basically based on that book
    1
  468. It's open source on github
    1
  469. Huzzah
    1
  470.  @henrikstryhn  Right, the specifics of this simulation have nothing to do with real life. But the general principles listed at the end do relate to real life. This video's goal wasn't to prove that they're related to real life, but to show the principles in action in an easier-to-understand environment. And thanks for the serious question!
    1
  471. They take a while, my b.
    1
  472. Not sure it will be the next one, but there is one in the works.
    1
  473. There is a gitgub repo in the description. It's open source.
    1
  474. Thanks for giving in, and for this kind comment.
    1
  475. Thanks for the thoughtful comment. You're right that one shouldn't draw conclusions from this alone. It's an illustration of concepts, but I hope it's helpful to that end!
    1
  476. Very true, though it is the base model, from which other, more complex (and closer to reality) models are built
    1
  477. I had a similar thought while building the sim, but the lucky occasions of getting a food spot to yourself don't end up affecting the equilibrium. They affect both types identically and so don't change relative fitness.
    1
  478. Oh nice. I played SimAnt as a kid, but I didn't remember that plot.
    1
  479. It took seven years for someone to appreciate that part, thank you
    1
  480. Lol. Things are getting faster, don't worry. But I'll be a lot less likely to blunder out of impatience.
    1
  481. Working on it!
    1
  482. There is a github repo in the description
    1
  483. No, it doesn't. For very simplistic decision makers, maybe, but not for humans. We can understand this dynamic and take the long view of our actions.
    1
  484. The serif text is century and the sans serif text is kurier
    1
  485. That would be a mix between the first gb sim and the one with four types. The way DNA replication works, traits either go together almost all the time (if they are on the same "gene") or are randomly mixed.
    1
  486. LIKE AND SUBSCRIBE
    1
  487. What a great question
    1
  488. Thanks :)
    1
  489. Only one blob notices. The fate of the other blobs depends on the choice of the one that notices. Of course, in real life, it's very possible for two to notice at the same time, so I'm not trying to claim that this model captures everything.
    1
  490. They are fractions of the total population, so they have to add up to 1. h + d = 1, so h = 1 - d
    1
  491. It’s a little confusing because when there was just speed, I discussed it in terms of how it affected distance per time, but then when all three traits were together I wrote the equation for energy per time. The mechanics around speed are the same in both cases, and what was said is technically correct, but it was a failure of clarity on my part.
    1
  492. LIKE AND SUBSCRIBE
    1
  493. That's what the word means in biology, but I agree it's not what we normally mean by it among humans.
    1
  494. I'm always working on them, just a slo boi
    1
  495. Working on it
    1
  496. Feels like an overinterpretation
    1
  497. As long as sellers have to compete for well-informed buyers, that's not true. (I concede that that's often not the case in the real world.)
    1
  498. Out of curiosity, what was the other site?
    1
  499. It has a little axe.
    1
  500. It's multiplying 7 by 1/6. I did forget that smooshing them together in that way could be seen as a mixed number. :(
    1
  501. The computer did do what I told it, but I only told it the basic rules, not what the result of those basic rules should be. Simulations help us discover how different kinds of basic rules lead to different results in complex systems.
    1
  502. Oh, I have been.
    1
  503.  @beatoriche7301  I actually thought the comment was on point. It definitely gets more specific than the video, but I interpreted it as expanding on the "what is the goal" section near the end of the video.
    1
  504. ​ @beatoriche7301  I haven't read "Capital" (assuming you mean Piketty). It's been on my list for a while, though. Sounds like you're much better read than me in this realm. I'm just someone who likes blobs and felt like there was something missing from intro-level econ resources.
    1
  505.  @beatoriche7301  I'd be curious to get a recommended reading list from you.
    1
  506.  @beatoriche7301  How about your top rec from each list? I've read The Wealth of Nations.
    1
  507. The shake only works if they both do it.
    1
  508. Unity. There's a github repo in the description.
    1
  509. The energy per unit distance is doubled, which is the loss in efficiency. It works out to four times the energy per unit time.
    1
  510. Can you say more about why? I agree it wasn't guaranteed to eliminate all the creatures, but we saw in the simulation that the creatures did go extinct when the food was reduced suddenly.
    1
  511. MonkaS
    1
  512. New videos on the actual biology coming soon.
    1
  513. You bet!
    1
  514. Me too
    1
  515. Tracing the alleles gets a lot more complicated, but it still ends up working out. I actually simulated it for recessive, dominant, and codominant situations, and they all worked, though the variance was higher with less dominance.
    1
  516. I would disagree that this sim suggests that conclusion. It's a bit too simple, since each seller just sells one each day maximum, not expanding at all. There will be more videos, and at least one will focus on monopolies and how they might come about.
    1
  517. I can't tell whether I'm happy or sad for you.
    1
  518. That's what I'm counting on. :)
    1
  519. Blender for the animations and python for the sims. There is a github repo in the description.
    1
  520. Computer concrete
    1
  521. Blender and python. There's a github repo in the description.
    1
  522. Not next, but I think we'll get there
    1
  523. There are already plushies
    1
  524. Reproduction is asexual in this simulation
    1
  525. Yeh, tryna go faster
    1
  526. ... true
    1
  527. Lol, I deserve that
    1
  528. Could you say more?
    1
  529. I hope so too
    1
  530. It would probably make a difference, but the sim just has one day for each generation, so there's only one chance to run into a preadator.
    1
  531. It just goes toward the closest edible thing
    1
  532. It's true. There's a lot here.
    1
  533. Me too
    1
  534. Nah, it's just a spiral.
    1
  535. Wow, have you got an eye! I use Blender Cycles. Excited to try Eevee once 2.8 is more stable. (which maybe it is? I haven't looked in a while.)
    1
  536. lol, someday maybe
    1
  537. There is one already
    1
  538. No
    1
  539. Yes
    1
  540. Exploring that would take a much more complicated model
    1
  541. preach
    1
  542. I do appreciate the critical eye! I can't believe I swapped them, though. 🤦‍♂️
    1
  543. There will be more evolution!
    1
  544. Thanks for the thoughtful comment. There will be more econ videos that build on this simple model, rest assured!
    1
  545. And I do appreciate it!
    1
  546. If you like genetic drift, just wait until the next video.
    1
  547. YUGH!
    1
  548. The gambler's fallacy is about mistakenly thinking probabilities change over time, but in this case, we are using the results as evidence of one static probability or another. Hope that helps.
    1
  549. yes
    1
  550. As do I, and thanks!
    1
  551. The "might" was sarcastic. It's a good idea. And there are more econ videos coming!
    1
  552. You didn't watch to the end, did you?
    1
  553. It's actually a goal of mine to make videos that don't ask people to abandon their faith. It's true that science might prompt someone to think critically about their worldview, but I hate how so much science communication makes people feel unwelcome.
    1
  554. Very cool! I was actually planning to do a dove/hawk video, likely the video after next. I hadn't seen a dove/hawk sim that had creatures moving around in space, so it's cool to see that it works out. :)
    1
  555. Hey thanks
    1
  556. True, l should link to other sources, though it's not actually his research
    1
  557. Yep, there's a (painfully disorganized) github repo in the description
    1
  558. There's a github repo in the description
    1
  559. Nah, I just took an already slow process and took time trying to make it faster. Hopefully we speedin up.
    1
  560. It's true. But it was less effort since I didn't have to actually review a company or send a bunch of emails.
    1
  561. Python and Blender. There is a github repo in the description.
    1
  562. That's fair, but the issue is that it's a reason for some voters to not vote honestly, and it suppresses the development of third parties.
    1
  563. False. Just takes time and focus, maybe working on some earlier stuff, too.
    1
  564. Trust me, I want to. :)
    1
  565. Yo, I'm 30, and my room is upstairs
    1
  566. Thanks, and agreed
    1
  567. It's a good idea for a video. Some people hear the explanation and refuse to believe it. Maybe a simulation showing it happen a bunch of times would convince some fraction of them. And the animations could be fun.
    1
  568. Where did that happen?
    1
  569. It's still diploid because there are two chromosomes. The X is important too. It's just that you rarely see a YY because two XYs can't mate.
    1
  570. It’s a little confusing because I talk about speed as distance vs time and then the “efficiency” of speed as distance vs energy. Energy per time would indeed be 4x for 2x speed, but distance per time is also 2x (by definition), so energy use per distance comes out to 2x.
    1
  571. The biology series will keep going
    1
  572. That's what the strategic voting would predict in that case
    1
  573. Hello
    1
  574. See pinned comment
    1
  575. yes
    1
  576. That was harder to animate, so I just did the slice 🤷‍♂️
    1
  577. Because rock beats scissors (etc), not the other way around
    1
  578. There are! But not quickly enough.
    1
  579. No they just have little nub arms and are gentle souls.
    1
  580. Best compliment I could ask for
    1
  581. Just a blob with a lot of expendable friends
    1
  582. I swear on the blobs honor it will be less.
    1
  583. I always take forever
    1
  584. Me too
    1
  585. See the pinned comment
    1
  586. Nope, Blender. But I have been meaning to check out Unity.
    1
  587. Ok fine, it's processing the port now.
    1
  588. Thanks for the thoughtful comment. The energy function for speed actually didn't change, though I put it in a confusing way and several others had the same interpretation you did. At 2:16, I discussed the relationship between energy and distance traveled, while at 4:59, the equation is for the energy per unit time. If you go back and compare, I think you'll see they are equivalent. Lesson learned not to switch how I express something in the middle of a video without making it clear that that's what I'm doing!
    1
  589. Hello
    1
  590. For the simulations shown, the results are consistent. There is a good bit of fluctuation in any given sim, though. Larger ones would of course be more robust, but for this illustrative vid, smaller sims were easier to track visually.
    1
  591. They do have a mouth in videos where they eat.
    1
  592. You can be a nice person, even if your genes are selfish.
    1
  593. You might classify that as a service that just happens to produce a physical object. The same basic rules apply. (and the same caveats!)
    1
  594. People might be interested to hear about this in the discord
    1
  595. With the relatively small numbers, it continues to fluctuate, but the final states are qualitatively stable.
    1
  596. There is a github repo in the description
    1
  597. This is pretty different from Conway's game. Conway's game is just squares that are either alive or dead, and each round, they may change state depending on the squares around them.
    1
  598. The sim code is also on github, in natural_sim.py. It's probably a headache to unravel since I have done zero refactoring or working with others on it, but if you want to do something with it, let me know, and I can try to help.
    1
  599. I'm doing this full-time now, so the posting rate should speed up over time.
    1
  600. This is just a baseline. More long-term strategies coming!
    1
  601. Reproductive period is definitely relevant. But it turned out the core concepts of mutation accumulation and antagonistic pleiotropy could be demonstrated even if the blobs could reproduce their entire life. So I cut out the reproductive period stuff. One question that would be interesting to explore is: In social species, how much does inclusive fitness affect the survival of older creatures? (Grandma might still help her offspring)
    1
  602. Seems reasonable. I didn't try varying food requirements.
    1
  603. Not really. It's a very simple formula. The formula itself is irrelevant to the underlying concept of natural selection. You could even call it part of the environment. But I picked something related to physics in the hopes that it would feel somewhat right.
    1
  604. Soon
    1
  605. 🤔
    1
  606. Thanks! I'm a fan of yours as well.
    1
  607. I always appreciate thoughtful comments like this one. The fact that I was going to animate the simulation added some constraints to the number of creatures and the total duration. I did run the simulation multiple times and for longer. I found that with the relatively small number of creatures, there's a lot of fluctuation, and the simulations shown were enough to achieve the goal of the video. I agree it could be interesting to increase the number of creatures and the number of days to get a better sense of the true long-term distribution (and also that adding complexity could be interesting), but because the constructed environment isn't meant to model anything in particular, I decided to be satisfied that it illustrates selection.
    1
  608. Thanks for the kind words!
    1
  609. Check the latest video
    1
  610. Thanks. I still need to construct a channel trailer to cover some of this, but in the mean time, you can read more about it on Patreon (link in the description), whether you intend to become a patron or not!
    1
  611. Very valid criticism. Like you say, they’re very important, but in my drafts, I wasn’t able to come up with an explanation that was both good and brief enough to feel like it belonged in this video. I decided to exclude them in this short intro video and give them their own video down the road. It will realistically be several months before I do that video, but you’d best believe it’s coming!
    1
  612. The concept didn't require calc. Why use a tool you don't need? Plus I figured folks like you would see the connection anyway. :)
    1
  613. I don't actually use Manim. I use Unity.
    1
  614. Lel, tricked you
    1
  615. It would be interesting to explore that in a sim. Given that there's still extrinsic mortality, I wonder how much a lack of aging would would actually depress the genetic diversity. I still expect the no-aging gene would still lead to better individual fitness and do well in the population during stable periods, but it seems likely that some of the newer blobs that did appear would do better during an apocalypse. Would be interesting to build a sim and play with extrinsic mortality and environment volatility knobs.
    1
  616. On the polymerase point, I would love to do a "Simulating the evolution of mutation rates" video. There's some paper I vaguely remember seeing where they actually sequenced a bunch of bacteria and observed how mutation rates changed in various conditions.
    1
  617. I will be checking the comments here, but I think you're right that the subreddit is a better place for this kind of thing.
    1
  618. I honestly think approval is even simpler. You literally can't fill out your ballot incorrectly.
    1
  619. :)
    1
  620. Yeah, the blobs seem to be working out
    1
  621. False!
    1
  622. Sexual selection is a kind of natural selection. Here, it's especially relevant to realize that the other creatures are part of the environment.
    1
  623. Shh
    1
  624. Valid point. I wouldn't expect these few small sims shown to hold up in an academic paper, but I hope they are still useful as illustrations.
    1
  625. I meant listing out all the permutations, not doing the arithmetic.
    1
  626. yea
    1
  627. We'll be there in a couple hours!
    1
  628. Lol
    1
  629. That came from this channel, and they credit it. If you're trolling, you got me. :)
    1
  630. They just go to whichever edge they're closest to when they're close to running out of energy and have eaten at least once (which is what lets them survive). Pretty arbitrary, just something to signal the end of a round.
    1
  631. Who is dream
    1
  632. Not that far off
    1
  633. Impossible! Wait.
    1
  634. Hey, I just saw the patreon email. Thanks much!
    1
  635. The others are sad more often
    1
  636. It does plug a related book, but I didn't think of the video itself as an ad. Glad to know it came across that way, though.
    1
  637. The energy cost per time of a 2x speed creature is 4x, leading to 2x energy per distance. I stated it confusingly.
    1
  638. Too long
    1
  639. which limit
    1
  640. :(
    1
  641. Thanks! They always had a sense trait and a corresponding sensing distance. It just wasn’t allowed to vary until partway through the video.
    1
  642. Me too. They are coming.
    1
  643. Thanks for the thoughtful comment. I purposely used "some" and "pretty close" to make it a very mushy statement, not making a specific claim, hopefully avoiding much of the ire such a statement would draw. Most of the remaining videos in the series will discuss how reality departs from this model. I'm curious whether you'd agree that this model is a fair baseline from which to start understanding economics, if only for its simplicity. If not, I'd be interested to hear any suggestions.
    1
  644. The speed cost function stayed the same throughout. It's a little confusing because I first explained it in terms of the cost per unit distance, and the equation gives the cost per unit time.
    1
  645. No problem. Thanks for engaging!
    1
  646. See the pinned comment
    1
  647. Running from a fight is a method of surviving. But you're right that different payoff structures are plausible. This video only explores a subset of the possibilities.
    1
  648. The behavior is a static algorithm, and the attributes just live in arrays, copied with some randomized variation when creating a new creature. Nothing too fancy.
    1
  649. what
    1
  650. See pinned comment
    1
  651. Thanks!
    1
  652. Hello
    1
  653. It's altruistic under the biological definition, but you're right it's not how the word is often used in general.
    1
  654. The green beard video used the same base sim, but the behavior rules are different in this one.
    1
  655. Thanks!
    1
  656. Psychology is legitimate, but minds were molded by evolution. Complex topics should be understood through several lenses.
    1
  657. You didn't watch the whole thing, did you?
    1
  658. The intended message is that markets provide value but that not all are ideal, and so we must be thoughtful on a case by case basis. I'm certainly not advocating for planned economies, or even any particular program.
    1
  659. ​ @gabrielmello3293  ​ I appreciate the thoughtful comments. I'll agree that regulation is often ham-handed and can do more harm than good, but to claim that every free market is ideal is far outside the consensus of even right-leaning economists. Clear examples are monopolies and externalities (not touched on in this video).
    1
  660. Yes and yes. I have to believe there's a behavioral economics video in this channel's future, but it's a few away.
    1
  661. Python. There's a github repo in the description.
    1
  662. Yes, but natural selection doesn't look that far ahead. It's kinda dumb!
    1
  663. I appreciate your deep engagement. You could join the discord or reddit to discuss with others if you like.
    1
  664. I am devastated
    1
  665. From where?
    1
  666. Yeah I think this is right.
    1