Did Google Researchers Just Create a Self-Replicating Computer Life Form?

hello info person this is Anton and today we're going to discuss something slightly different or I guess maybe not so different in a sense that we're going to be discussing the idea behind the evolution of life but here based on a somewhat intriguing study by the researchers from Google and specifically a study that tries to mimic the evolution of Life by using an extremely simple computer code and in the process they actually do make some somewhat unexpected discoveries all of which you can discover Yourself by reading the study in the description below and so let's actually discuss this in a little bit more detail and talk about why this is maybe kind of important but also what it shows us in regards to the concept known as self replication but first I guess just a brief history so this idea of trying to simulate life like environment by using mathematical principles and computer simulations is obviously not new the famous Conway Game of Life is probably the best known example where a kind of a autom system can actually be deviced mathematically by introducing just a few simple rules and by then seeing how the system evolves you can actually try the examples of this in one of the links in the description but in essence this is known to produce quite a few interesting patterns with some of these patterns possessing somewhat interesting Behavior such as motion but the main ideas behind all of this essentially started with a kind of a mathematical SL philosophical exploration back in the 1940s and 1950s with essentially one of the main I guess fathers of this concept being a famous engineer mathematician and physicist John V Newman and you might know his name because he was also responsible for the concept known as the one new probes or basically self-replicating probes that could technically exist somewhere out there and represent a kind of a endgame for various alien civilizations that could help them colonize the rest of the universe and so he was actually super fascinated with the idea of self-replication with his research basically inspiring a lot of other mathematicians and a lot of other scientists to try to look into this a little bit further ironically though he was able to propose all of this before we even knew DNA existed and so in some sense his research became fundamental in order to help us understand how various biological units are able to reproduce and are also able to evolve over time and so in some sense this new study by Google is kind of a continuation of all of this work after several decades but in this case discover something that we've never really seen before which is why I thought it was kind of worth exploring so for example in that game of life I showed you previously one thing that it does not do is basically create its own copies or evolve in a way where it mimics life itself and that's despite being really complex otherwise but in this new study the scientists took a slightly different approach first of all they chose an extremely simple language okay this is going to be hard the name of the language is um right here on the screen and it's essentially is somewhat simple but somewhat eccentric language created by a computer science student back in 1993 that contains only eight simple commands but it's fully functional otherwise and though it's not really meant for practical use mostly because all of the programs end up being super long it is nevertheless quite functional and obviously can be used to write programs and Wikipedia article below does actually provide some examples such as the famous hello world program that in this case takes quite a few lines of code and so because this language is so minimalistic it's actually surprisingly useful for all kinds of evolutionary research here because there are only eight instructions everything becomes extremely simple and extremely easy to implement and so for this particular study the researchers basically did something super simple they created a kind of a digital primordial soup and to some extent you can see it visualized right here and what you're looking at right here are a bunch of random programs that don't really have anything in them other than a few really simple instructions but for the most part there is no General instruction there's no rule there's no regulation all of this is just completely random so in some sense it directly mimics the game of life but most importantly there's absolutely no pressure to do anything all of these individual programs and there are millions of them here are left to do whatever they want to do and this is really important there's absolutely no rule or any kind of Regulation to evolve or to self-replicate and so now all of these individual pieces of code start to interact randomly combining and mixing together and also executing additional instructions as they slowly change because of random interactions and in this case this is done for millions of generations and the Assumption here based on the idea that there's really no instruction no regulation no particular direction or force or pressure was that basically all of this is just going to randomly change over time possibly shift here and there but most likely not really affect anything too dramatically so for example unlike previous simulations these programs don't even have to do anything because there is no point system and they don't actually win or lose anything with the expectation being that because the population was kept at a fixed number they're going to do a bunch of random stuff but do nothing that's comprehensible or that has any direction and this is where the surprise starts in every single case they did something that was directional and they did evolve into something unexpected and you can actually see this visualized in the bottom left corner and it's going to start happening really soon in every single simulation some of these programs eventually emerged as self-replicating taking over the rest of the simulation and this emergence of self-replication seems to have happened multiple times and in many cases some of these self-replicators were much better than others and thus started to basically kind of compete for space even though this was never programmed or never intended so in this case you can see that one of these first replicators is now taking over everything but eventually after a few more Generations more and more competition started to appear and in some cases would completely overwhelm all of the other programs which is really bizarre because this mimics a biological system without any programming or any code to tell it to do so and because there was no explicit Fitness function yet Fitness still emerged and some programs were basically better than others this was a relatively big shock and you can see all of this in a couple of YouTube videos I posted in the description where it helps you visualize these programs after millions and millions of evolutionary processes and in many of these experiments it sometimes took up to millions of steps before this unusual behavior started to appear but interestingly enough some of these programs were basically just run on a laptop and after about half an hour in every single case you would see self application and this is of course really interesting even though we understand Evolution to a pretty good extent we still have no idea how various molecules early on eventually became self-replicating and even though this experiment doesn't really tell us exactly what happened what it does kind of show us is essentially some kind of an inherent mechanism that possibly creates complexity from absolutely nothing or basically the implication here is that complex biological behavior Can spontaneously appear if you wait long enough and so as the researchers say themselves nothing magical happened here and absolutely no Direction was given to anything other than a few initial instructions and so if they were able to create this after just half an hour on a laptop we can only imagine what could happen on a typical planet with a lot more complexity and a lot more chemical diversity after billions of years or I guess that's the basic conclusion from this study because of the limitations of the computer power here obviously nothing more complex arose from this and obviously none of these programs developed into some kind of a super complex intelligence or even went beyond simple self-replication and that's actually one of the biggest criticisms from various biological sciences so far here they don't actually think this would lead to more complex Behavior automatically or in other words even though self-replication is very impressive they don't think other things would follow afterwards and they do actually provide several examples where this might have occurred using other experiments even using things like RNA there are several experim that used various RNA molecules where various RNA strands were replicated in a test tube which resulted in RNA getting shorter and shorter and replication getting faster and faster and in those experiments even though I guess replication was achieved it was not rewarding complexity instead it was rewarding Simplicity and the speed of self-replication and biologist believe that this is maybe the opposite of what's needed to explain the complexity of biological life on Earth and so basically they don't think that having a bunch of self-replicators is going to guarantee further complexity mostly because here it's the simply licity and the speed that seems to guide everything on top of this because life actually involves DNA RNA proteins and a huge interplay of various complex systems just looking at self-replication is not really helping us enough to understand how all this could evolve nevertheless in the study there is maybe some evidence that this is a beginning of something more complex it just it would need more time and possibly more computational power on top of this the scientists actually try this using other Lang languages or other simple languages such as fourth z80 and a language known as suq and interestingly only suq was actually unable to produce self-replicators and here they believe it's because it required a much longer length of code with the longer code no longer being able to produce replication instead producing random programs with the main conclusion here being that the length plays a very important role in determining if self-replication is going to become possible but natur except for this simple replication there are still so many unanswered questions and so many things that we still don't know for example how much additional complexity can be produced given enough time also what exactly causes the self- replication to appear and is there any way for us to guide the evolution of these systems to produce even more complexity lastly is the evolution produced here similar to what we observe in real world or are there any specific notable differences that we can one day explore in future studies and so basically at least for now the only thing we know for sure is that self-replication seems to appear naturally but everything else is still kind of undetermined although because this is a fundamental property of Life technically this is a super important Discovery and one thing that kind of connects to all of this is once again researched by V Newman and his colleague from the Lo salus laboratory Stanis ulam in the early 1940s way before the discovery of DNA and way before computers became L widespread the first research on self-replication was actually done with Clay crystals and that's because surprisingly except for DNA and RNA crystals produced inside clay exhibit the only other known form of self-replication and that's because Clay is made out of large number of small crystals inside the environment that promotes crystal growth but it also promotes the growth of various irregularities especially when placed in water solution and so as these crystals grow and develop they actually end up producing irregularities that then break apart forming new crystals new irregularities and even undergoing a kind of a evolutionary change which actually surprisingly corresponds to all of the definitions of self-replication and mimics the biological life extremely well which of course raises a lot of questions for example does that actually mean that some kind of a crystalline life could exist Somewhere Out There produced by Clay on some other planet somewhere out there or maybe this is actually how life on Earth formed as well starting from these simple Clays early on with these crystals eventually replaced by various organic molecules all these questions have actually been tackled by various papers but we just don't have any answers yet nevertheless because of this new study hopefully this will lead to more interest additional simulations and possibly additional exciting discoveries but until those future studies or until other discoveries that's pretty much it all of the links should be in the description below thank you for watching subscribe share this with someone learning about space and Sciences come back tomorrow to learn something else support this channel patreon by joining Channel membership or by buying the wonderful person t-shirt you can find in the description stay wonderful I'll see you tomorrow and as always bye-bye [Music] [Music] n e

Channel: Anton Petrov

Views: 16,677

Keywords: google, self-replication, synthia, syn3b, anton petrov, science, physics, astrophysics, astronomy, universe, whatdamath, what da math, technology, JCVI-syn3.0, JCVI-syn1.0, synthetic living machines, Xenopus laevis, Artificial Life, Synthetic Genome, artificial cell, artificial genome, creating artificial life, scientists create artificial life, artificial life made, creation of artificial life, artificial cell created, artificial life, created life, new artificial life, mycoplasma mycoides

Id: L_IWVZPmc-E

Length: 14min 24sec (864 seconds)

Published: Tue Jul 16 2024