https://youtubetranscript.com/?v=RaApmHqiGzI
Hello there and welcome back to Navigating Patterns. What I’d like to go over today is the difference between complicated and complex. Now I’ve been teasing about this video for a while. I’ve also tried to make it once and failed because sometimes I fail. But I do so spectacularly. So in order to do this we’re going to have to define complicated, we’re going to have to define complex, and then we’re going to have to talk about some examples and see if we can sort of get to the bottom of the difference between complicated and complex, why it matters, and what we’re going to use this for, or what we have used this for in past videos, what we’re going to be using this for in future videos. So complicated things have lots of parts, lots of connections, and probably sequence, or maybe procedures that you can apply. They’re usually linear order in time, right? And a lot of times in complicated system it actually matters. So if you have something complicated that you’ve built, like a tree fort, it matters not only what order you build it in, but also what order you take it apart in. Take the floor off first, it’s not going to dismantle correctly. So that’s a complicated thing. Now why? Because when you build something complicated and you’re taking a bunch of parts and putting them together in a specific order, usually to a plan, but sometimes not quite, and what that’s creating is something that you expect. And then all of those parts have a value in the system that you can understand easily. And usually because the relationships tend to be linear in these complicated systems, not always, but almost always I would say, you can understand the implication of moving, changing, or withdrawing that from the system. So if you’re building like a tree fort or something and you’re using nails, you can probably get away with a certain number of nails and you can remove some and still be okay. For how long? I mean, you can get into all kinds of calculations. It might be up in the air. So in complicated systems, again, there’s parts, there’s connections, they can be taken apart very easily. The relationships between the parts in a complicated system are understandable because they’re more or less linear. One part connects to N other parts. It could be one other part, it could be 10 other parts, but it’s only those 10. And so when that one part is understood or removed or put in place, you know what the effect is, roughly speaking, and that effect is usually immediate. These complex systems, on the other hand, that’s different because there’s usually not a direct relationship between the parts. They’re intertwined somehow. There may be hidden reactions and actions going on. The systems may be interacting in a way that isn’t viewable by you or understandable by you. So all of these parts in a complex system are so multi-connected and so interdependent that changing one of them actually alters most of the others, or at least enough key pieces that it has an outsized impact. So for example, in a complicated system, you can take one board out of a treehouse and you know, you’re probably okay, maybe the railing will fall down if you lean on it. In a complex system, if you were to take one piece out, the whole thing might come crashing down or 80% of it might stop working the way it’s supposed to because very small inputs can have very large changes in output in a complex system. And what that means is that it’s overconnected, we’ll say. It’s not direct one-to-one linear relationships. Actually, the relationships are much more, we’ll say, much more important, right? The overconnection means that the connections aren’t merely complicated because a complicated system, you can slice it and dice it and understand it, but they’re so intertwined. There’s so much going on and maybe, again, things you can’t measure go on in complex systems or can’t measure, maybe can’t measure easily, but maybe can’t measure at all, ever, that there’s no way to tell, you know, what’s going on, right? It’s so intertwined that one small change has an outsized effect. Again, changing one thing in the system could affect 80% of it in a complex system, whereas in a complicated system, that’s generally not true. It can be true, but generally complicated systems, that’s not true. There’s always, you know, one part that can like throw out the works, but most of the parts aren’t like that. In a complex system, most of the parts have that quality, and so the relationships in complex systems can also change and switch while they’re in motion or while they’re interacting. Complex systems can, one minute this thing can be doing one thing and then the same part can be doing something else as the result of, who knows? And in a complicated system, you generally have first-order effects, and all the parts have first-order effects to one another. Again, they’re linear. In a complex system, it’s second-order effects are important. So you may say, well, what are second-order effects? Second-order effects are sort of the things that you can create with like a Rube Goldberg machine, where you set a ball in motion and then that does a bunch of things, right? And then so the outsized effects on the back end tend to be, those are second-order, right? They’re not a direct result of moving the ball. A bunch of other things happened in between. Now, and you can rightly argue, well, Mark, that’s actually just a complicated system. Yeah, generally speaking, but with a complex system, it’s almost all second-order effects. So you don’t know from impact to impact within the system, even if it’s measurable, where that’s going to show up, what that’s going to do. And I’m not saying you can’t make direct correlations in complex systems. You can, but you generally can’t trace and figure out why that action had that effect, because it’s too far removed. There’s other things going on. Usually complex systems are very well intertwined with things. So one example of, say, a complicated structure would be a building. A building’s a pretty complicated structure, but we know how to build buildings. We can use blueprints. We can build buildings. We can get parts lists for them, right? You know what you can’t do that with? A tree. I’m not saying you can’t grow a tree. You can, but how does that work? Like, I mean, I put a seed in the ground and bang, sure, kind of, but that’s not the same as a building. And then how is this little tiny seed having this emergence of this thing with roots, a trunk, branches, probably leaves. Not all trees have leaves, but most of them do. And doing photosynthesis, drawing water into itself, finding water in the river. How does that happen? This is a very complex thing, because you’re taking this little tiny seed and then up springs this tree. It’s a very complex thing. And you can’t simulate a tree by cutting it up and then rearranging the parts. In fact, we can’t simulate trees by cutting them up and building analogs to the parts and then making an artificial tree. Like, there are aspects of trees that we can make artificially. You can make an artificial tree that looks like a tree, but it doesn’t do any of the complex things that trees do. It only has an exterior that looks like a tree. It doesn’t find water with its roots automagically somehow. It doesn’t interact with the fungal system. It doesn’t talk to other trees. It doesn’t grow a trunk. It doesn’t grow branches in just the right way to get just the right amount of sun to keep it growing and maybe to shade other trees around it, because trees actually do that too, which is very mysterious. That’s a complex system. It’s coming out of this little thing. What’s that all about? So those are sort of some minor examples. And there is some way that complicated systems can be predicted. Their inputs can be known. Their outputs can be charted. The changes are relative to known relationships. And again, none of this is true for a complex system. Small inputs have large output changes, and those connections are not really clear. Another example is standard computing. If you’re going to just do programming, standard programming is pretty, you know, it can be complicated, but it’s not overly difficult. For some of us, other people really struggle. But when you get into the realm of AI, for example, and this is the big deal, right? It’s a big deal. Why? Because AI is complex, but most AI is also black box. It’s like, whoa, what do you mean black box? They don’t know how it works. In an AI, you may be able to train it, but you can’t describe how it works. You can’t describe how it’s doing what it’s doing. We just don’t know in most cases. And you could make arguments about that. And I’d be happy to show you the error of your ways. And there are papers about this. And some of the papers will claim that they know how some of these work. But actually, there’s a big push to in computing to make AI less black box so that people know what’s going on in it and how successful that will be. I don’t know, because there’s probably a tradeoff when you make AI complicated instead of complex for how much it can do. That’s my guess. Just throwing that out there. But that’s what makes AI interesting, is that they kind of train it, they guess a lot. And then when they get the guesses right, they go, yeah, it’s really successful. But they don’t know how it works. And so they can’t tweak it. They’re very specialized. I know they’re talking about artificial general intelligence. That’s not happening anytime soon. Because it’s a complex thing and we just don’t know how to understand the complexity well enough to manipulate it. And we may never. Complex systems may be beyond our collective intelligence as a species. I don’t know. Maybe not. Maybe. So some of the things that are some of the qualities, rather, that complicated things have is, you know, they tend to be fragile. They’re not very resilient. Whereas complex things tend to have a lot of bend in them, like trees bend in the wind and, you know, they can adjust to sunlight, they can adjust to the water amounts underneath the ground, they can adjust their height, right? They can do all that. And so they tend to be anti-fragile. And again, there’s a way in which complicated things can become complex. But I think that requires an emergence. And then when that emergence happens, the nature of the relationships change such that they become non-linear and things become complex because of over connection. That’s sort of the best charm I have for it. And I think that’s where we get confused, right? So you have a friendship with somebody and then something happens, it gets complicated, right? But then when you’re in a relationship with somebody, well, that’s complex. Like relationships are just like, because there’s so much going on between two people. There’s so much connectedness involved in that. And when you’re in a relationship with somebody, they start out simple, they start out, you know, sort of complicated, right? With lots of moving parts. But then if they, once whatever line is passed, and I don’t know what that line is, they become complex. They really do. And sometimes we misuse, we say, oh, it’s complicated. But what we really mean is it’s complex. And complexity is usually beyond our immediate understanding, right? And again, it may be beyond any possible understanding to some extent. And I know people don’t like to hear that because they like certainty in their world. But look, when somebody explains to me how a tree does what it does, until then, no, I’m not buying it. Complex things seem to be so complex that complexity itself seems to be the problem. We just don’t have enough understanding of how these things interact, right? Everybody thought the Human Genome Project, well, that was it. We were gonna have one number, we’d know everything we need to know about you, because that number was your genetic code. And then we discovered actually halfway through that project, because I knew somebody in that project who told me this, oh, yeah, mathematically speaking, that can’t be true. We already know this. There aren’t enough mathematical combinations in your DNA to explain all the things that are in a human. And I was like, oh, okay, and we spent how much money on this project? It was a good project, not complaining, just saying they thought they were going to get one thing and they didn’t get it. And then we discover epigenetics. It’s like, oh, and then epigenetics is ridiculously huge, and at least is some reasonable explanation for why everything in genetics is maybe 20 to 30% likelihood and not 80% likelihood for most things. So that’s an example of something that looks complicated. DNA is very complicated, but we can slice it, we can dice it, we can totally figure it out. And then it turns out that the system that DNA is embedded in may be complicated, but the rest of the system is complex. And a small change in the DNA may or may not have an effect based on your epigenetics. It’s like, oh, very depressing. Thought we were on the verge of something. It’s not there. And so you can see how something starts out complicated, because DNA and RNA are very complicated. And then, bang, complexity. Where does it come from? I don’t know. But there’s a difference there, and it’s a big difference. It’s a whopping difference. And I think that we get confused between complicated and complex. Complicated things are things we could understand, maybe given enough data. And most complicated things are understood. Like I said, building a skyscraper, very complicated, but we know how to do it. We can do it reliably over and over again. Complex systems, and there’s complexity theory. There’s all sorts of complexity collapses, all sorts of things around that. And again, they do tend to be more resilient, antifragile, if you will, to use a Nazim Taleb term. But the changes, small changes can have very large effects. Sometimes small changes have no effect in complex systems. It all sort of depends, but we don’t understand them. And maybe we can’t. And it’s a good lesson to sort of realize, well, there’s things that we can’t understand potentially in the world. And that’s good, because if we did, it would take all the mystery out of everything. And sometimes I don’t understand, for example, why people pay attention to the things they pay attention to, because some of them are clearly not helpful. But I can’t know somebody else’s mind, because it’s too complex. But what I do know, and what I can count on, although it is complex, is that I’m very grateful for your time and attention.