Simulating the Evolution of Teamwork

I think I prefer the narrative that 1 mango lets the blob survive, and additional allow them to reproduce, because it’s so tragic that the blobs die every day

One thing I am interested in is what would happen if the team blobs learned to find each other and avoid the solos letting them fight each other

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!

fun fact: the Nash Equilibrium was made by John Nash, who was actually an extremely schizophrenic mathematician and the case study of how he dealt with it is taught in most introductory clinical psychology courses. He even has a movie about him that is great (albeit inaccurate).

How did this guy come up with such a perfect creature, the b l o b

Cooperative strategies almost always work better in repeated interactions, especially when a tit-for-tat strategy is being utilized. It is interesting to see how these equilibria form even under these conditions.

We have endless patience with you Mr. Primer, and it always pays off.

These concepts are very much over my head but I appreciate you putting them into blob form and explaining them

Next we are going to simulate the evolution of evolution

It would be interesting if the Blobs could only visit trees near their "home". (Edit I did that myself, see the bottom of the comment) I imagine a start with 50% solos with 2 Nash Equilibrii would result in a mix of "friendly neighbourhoods" where cooperation dominates and "battlefields" where fighting dominates. Edit: I made a simulation, with a grid of 32 by 32 worlds, each with 4 trees, and enough houses for 64 blobs. I used the same rules about blobs going to get food from trees and reproducing as in the video. Except each of my blobs can freely visit trees in their own world, and the 4 nearby worlds, and the blob's children can choose to move to a nearby world. I started with 1 friendly and 1 solo blob at far-apart worlds and looked at 256 turns. With the setup from 3:23 two "nations" formed, but (as some people below predicted before I tested it) the friendly nation eventually -- after almost all 256 turns -- was able to convert the entire world to its peaceful ways, by force. The friendly nation was able to sustain a much larger population, so enough friendly blobs "immigrated" to the unfriendly neighbourhood each turn, that they were able to cooperate with each other to get more resources than the "locals". Edit: My C++ code is on Github dot com slash nikolajRoager slash blobsOnLattice (I am apparently not allowed to include a link in a comment) The code is not tested on windows, and normal warnings about running random code from the internet apply.

And when the world needed him most He returned

I'd love to see a more complex simulation where, rather than the blobs dying after a single round of gathering, they instead do multiple rounds first and remember who the solo blobs are. A few other ideas to make it more interesting are: -Sharing with blobs who got nothing. -Solo blobs that will try to steal from teamwork groups. -Blobs who change what they do depending on how other blobs treat them. -Larger groups who get less each, but can block solo blobs. -Smart blobs who wait to act selfishly. -Traitor blobs who work with each other, but are selfish against team and solo blobs.

I'd love to see the outcome of a fission/fusion species much like coyotes. Unlike wolves they don't require a pack, but can group up if needed. It apparently worked well for them as they were the only predator to be almost unaffected by the predator war.

Every time Primer uploads it feels like a life checkpoint

I think we should all agree to appreciate how he uploads 3 times a year

Really interesting. I wonder what will happen if the blobs tried to find the right partner before shaking a tree. The blobs could have many more traits: -How long do they spend looking for a partner before just picking whoever? -How good are they at determining which traits their candidate has? -How well can blobs pretend to be another kind? -How many chances do they get before dying? Do they stick with an arrangement that was beneficial? Til death do them part?

An interesting simulation would be one of 2+ importing/exporting economies which trade currency for goods with each other, but neither/one/both can print money at varying rates.

Hey primer, i’ve been following your videos for years and I love how as I grow older, I understand more about your videos! When I was younger I never really understood what you were saying, and just liked the blobs and the numbers, but now I can truly follow what you’re saying and I think that’s fascinating.

Now, Here's what i would find super interesting to simulate: Introducing blobs that weigh their chances. I.e. a (perhaps purple) blob that knows it's chances and chooses either fight or cooperation based on what would be Most advantageous in regards to what their opponent is

Man I love your content! The way you do things seems so intuitive like it just makes sense the way you go about things and the results are always so interesting! I also love just how unique your type of videos like simulating natural selection or like social behavior through out human evolution it's something I never see! I really appreciate the hard work and dedication you have keep up the great work! been watching you since your natural selection video