[00:00:00] Speaker A: Foreign.
[00:00:05] Speaker B: The Future, a podcast about evolution and intelligent Design.
Welcome to ID the Future. I'm Jonathan McClatchy and I'm joined today by a special guest, Jonathan Bartlett. Jonathan Bartlett is a software developer, educator and academic researcher. He's published books and papers in computer science, mathematics, theoretical biology, and philosophy of science. He's the author of many books including Calculus from the Ground Up, Electronics for Beginners, and Learn to Program with Assembly Language. He is a fellow of the center for Natural and Artificial Intelligence and writes for MindMatters AI. Welcome to the program, Jonathan.
[00:00:43] Speaker A: Well, thanks a lot for having me on. We can each refer to each other as Jonathan and we can confuse the audience the whole time.
[00:00:50] Speaker B: Absolutely. So why don't we begin. For those listeners who perhaps aren't familiar with, with you, tell us a little bit about your, your background.
[00:00:57] Speaker A: Yeah, so in general I tend to be just kind of a curious guy. I'm interested in a lot of different subjects and I write about them. That's, that's actually one of the ways that I kind of process my thoughts is by writing. And so kind of if you dig back into my, my past, my, my first book was actually on assembly language programming. And I wrote it because I thought that not enough people were coming out of computer science programs really knowing how the computer works. And if you know anything about computers with assembly language, that's really the language of the computer itself. And so even if you don't program in assembly language regularly, knowing assembly language helps you think like a computer. And so that's kind of why I wrote the book, is to help people really think like a computer. And it took off and actually it was required reading at Princeton for a while, so that was pretty cool.
So I also do a lot of education.
My main thing is teaching a homeschool co op, but I sometimes guest lecture at college classes and that sort of thing.
But I teach a calculus class. I actually hated all of the calculus books that people used because most of them I'm pretty sure were written specifically to make kids hate calculus. And so I decided to write my own. So I wrote calculus from the ground up. And while I was doing that, I actually discovered a couple of new things involving calculus. I'm not going to go into depth on that because it's probably outside the scope we want to cover. But I wrote some papers on it, got those published.
There was even a post publication review of that paper in Mathematics magazine. And so that was kind of fun. And then I've basically due to some illnesses in the family, I got really involved in biology and biochemistry and genetics.
And that kind of led me to have an interest in intelligent design.
I felt there were a lot of questions that were not addressed well the way that current biology was handling them.
[00:02:47] Speaker B: So you have a background in computer science and dabbled in calculus, and you got interested in mutations. What was it that first struck you about mutations?
[00:02:57] Speaker A: Well, here's what's interesting about mutations is that what we're taught in school is that mutations are basically haphazard. There's a lot of different words that you can use. Random mutations, haphazard mutations, spontaneous mutations.
And the idea that they try to get across is that mutations are not being caused by a system. There's not a control system that's managing your mutations. But what I found out is that that actually wasn't known for sure. It was sure that there are mutations that occur outside of a control system that are problematic and that sort of thing. But the idea that all mutations are like that, or even most mutations like that, is not something that was actually.
It wasn't determined. That's not something that was, you know, where people did experiments and said, hey, I wonder if mutations are engineered by the cell. Here's a set of experiments that we can do. Okay? We've determined that they're not. That is not how that worked. And so as I dig deeper, I found that there actually are a lot of known mechanisms that are generating mutations in a way that is not haphazard, that is geared towards the organism's benefit.
And so that led me to ask a whole bunch of questions like how do we measure this? How do we better understand biology in this light? Because even though biologists understand these facts individually, there hasn't been a lot of work to synthesize this into the way that we understand how biology works.
I have a philosophical bent and an educational bent, and I recognize that unless you make an actual effort to rethink your foundations, even though you know new information, you kind of go along with what you've previously been taught and learned and kind of the status quo.
[00:04:43] Speaker B: So you think there is evidence to suggest that mutations may in fact be directed in some sense?
[00:04:49] Speaker A: Yeah. What I think has happened is that I think people had a misunderstanding of what they should find if they found a directed mutation. You know, if I say that Fred built this house, there's a lot of different ways that Fred could build the house.
If I disprove one of them. I haven't disproven that Fred built the house. So that's kind of what happened. A lot with mutations is that, you know, somebody had in mind some way that they thought that mutations were directed and they would disprove that particular way and then use that to kind of imply that mutations weren't directed at all. So that. That led me to think, how could we calculate and how could we measure this in a more general way so we could more adequately understand what it means to be directed? How do we tell if something's directed, even if we don't know ahead of time the mechanism?
[00:05:42] Speaker B: So what findings did you come to as far as how to measure the extent to which mutations are directed or are stochastic?
[00:05:51] Speaker A: So the way to picture this is to imagine that you're in a building that's 1,000 floors tall, and each floor has a thousand rooms.
And in, let's say in 10 of those rooms, there's a million dollars. Okay. And you can open up one door. Okay? So let's say that you were there when they were constructing it, and you don't know exactly where they put the money, but you knew that they put it on the fifth floor. And so you said, hey, John, if you go on the fifth floor, one of those rooms has the money. Okay? So if you think about it, I still don't know where the money is, but I have a much better idea of where the money is than before you talk to me. So if you think about the abstract space of where the money could be stashed, there were a million possibilities.
But after you talk to me, I've narrowed it down to 1000.
So what I figured out is you can actually calculate the abstract possibility space and then figure out how often they're actually figuring out the answer in. In real life. And you can compare those, and you can come up with a measurement. It's called active information. That tells you how much information did they give you? Because what happens is someone will say, okay, well, well, this isn't directed because most of the mutations aren't giving you benefit. But again, if you go to the example that I just gave, even though you told me what floor it's on, most of the choices that I would make on that floor are still wrong. Right. The money's only in 10 of them, but there's a thousand to check. So 99 times out of a hundred, I'll be opening the wrong door. But I have a lot more direction than I did beforehand. So you can measure the amount of direction that that gives. There's a metric called active information, and it was originally developed by Bob Marx and William Demsky they had used it to measure the amount of information that programmers put in computer evolution. So you can say, you know, how much did the original programmers add to this algorithm to help them find their goal? And I realized that if we view organisms as directing their own mutation, we can apply that same idea to biology. You have to adapt it a little bit because you don't have quite the ability to manipulate things in biology as you do in software. But nonetheless, you can use the same basic idea of comparing the abstract search space and how frequently solutions are found there versus the actual searches that organisms do.
[00:08:32] Speaker B: So you basically are saying that there's a constraint or a bias on what mutations are going to take place in an organism. And, and that is a, as a design feature of living organisms.
[00:08:45] Speaker A: Yes. And I want to, I want to separate out two types of biases because one objection that a lot of people put forward is that like, well, of course we know that mutations are biased. Nobody thinks that they're uniform, random. And that's true. The question is, is the bias towards something? Right? So for example, if again, going back to our example of the money behind the doors, it's if you didn't know where the money was stashed and you told me that it was on the fifth floor, well, I might be searching in the entire wrong area.
And in fact there might not be any money where you told me if you didn't know something yourself about the search space. And so you can have biases. So there's a difference between just being a biochemical bias and being a bias that tends toward function, because there's nothing a priori that would tell us that a biochemical bias should have function. So if a biochemical bias is giving us function, then that's a result that requires a reason.
And so that's something that has to be explained and understood.
[00:09:50] Speaker B: What implications do you think this idea has for the plausibility of evolution to explain and give an account of the complexity in designoid features of living organisms without the need for interventions along the way by special divine fiat or something like that.
[00:10:10] Speaker A: So it kind of depends on what the definition of evolution you're using because to some extent it actually makes the possibility of some type of evolution more likely. You could say the organism might be pre coded with specific features that allow it to diversify in certain ways at certain times. And there's not really a limit to how much biological complexity could be outworked in that way. On the other hand, it greatly diminishes the evidence for Darwinian evolution. Because basically what we're saying is when you actually look at the mutations that we do see, the ones that give benefit are tending to be the ones that have an informational advantage already. And so once you recognize that, you recognize that, well, if you don't have an informational advantage, then even the things that most people will willingly attribute to Darwinism today, a lot of those are even unlikely to occur without having that informational advantage.
[00:11:11] Speaker B: So we hear a lot of theistic evolutionists talking about front loaded evolution. Do you think that this sort of scenario is compatible with that perspective?
[00:11:22] Speaker A: It depends on what they mean by front loaded. There are some people who say front loaded and they're indicating just the laws of physics. Front loading in the laws of physics I don't think is workable because the laws of physics don't generate informational complexity. I mean, we'd have to find radically different laws of physics to do that. If by front loading, there's actually a metaphor that I often use with people, and that is, if you imagine people don't do this a lot anymore, but back when I was growing up, we would install operating systems on computers. From a cd, you'd insert the CD and the CD would be onto an empty hard drive and it would generate an entire operating system with thousands of programs. In that case, what you have is, you have the CD is a single program. You could think of it as a common ancestor, and when you load it in, it generates thousands of different programs. But what it doesn't do, it doesn't do that by random mutation and natural selection.
It does that because it has specific information guiding where it goes.
So a front loaded evolution that was highly specific, that had lots of information at the beginning, I think is something that is at least much more plausibly workable than the standard Darwinian story.
Now, I think there's a lot of problems with that as well, but nonetheless, I think from an information standpoint, it doesn't violate information theory. And so that's largely what I look at academically is whether or not various scenarios are basically trying to get information for free.
[00:12:56] Speaker B: Makes sense. Do you have any particular examples of directed mutations in the literature?
[00:13:00] Speaker A: Yeah, there's several. Some of them are kind of technical. The one that's easiest to describe isn't evolutionary, but I hope it presents a good example.
So everybody knows that you have antibodies that will, when you get sick, they find the antigens, they latch onto them and they signal the immune system to come and take care of them. Well, your antibodies, if you get an antigen that your body's never seen before that doesn't have any antibodies for.
Well, what happens? Well, your body actually will mutate to generate a new one. But it doesn't just generally mutate and hope for the best.
It focuses on antibody genesis and not only the whole gene, because there's basically two parts of an antibody gene. There's the part that sticks onto the antigen and there's the part that signals the immune system. If you mutate the part that signals the immune system, that doesn't help you because then it can't tell the immune system that it found one. The only place where mutations make sense is in the part that sticks. And it turns out that almost all the mutations that occur are on the part that sticks. So it's not any gene, it's on the antibody gene and not any part of the antibody gene. It's the part of it that sticks to the antigen. Plus it has a mechanism to say, hey, I need mutations, and hey, I'm done looking.
So these are not random mutations in the sense that we're used to talking about. These are highly directed mutations. Now, within those, that small set of mutations, they are randomized, at least to some degree. It very much seems that way, at least. And I don't think that's a problem because you've, you've skipped over billions and billions of base pairs and you're focusing on like 500.
So you're decreasing your search space by gobs and gobs of exponents. So basically it's directing you where to go. So that's kind of the easy one. It's not evolutionary because those generated genes aren't passed on to your offspring. But it demonstrates the concept that your body actually does know how to mutate certain genes.
[00:15:06] Speaker B: Absolutely. And one objection I can hear people potentially saying is, well, if mutations are directed, why are so few of them beneficial? Right. 75% of all mutations are selectively neutral. They're neither here nor there in terms of the organism's fitness. And then if the remaining 25% of genetic mutations, the vast majority of those are deleterious, they're harmful to the organism's fitness and their sifted out by purifying selection. And there is only a very small fraction of genetic mutations are in fact beneficial. So how do you square that with the hypothesis that mutations are often directed?
[00:15:41] Speaker A: Yeah, so there's several different answers to that, because I think there's several different reasons, one of which is that your body is not omniscient. So there actually will be many Mutations that don't work out for that very reason, even within the example I just gave, most of the combinations that it tries won't be successful, but it's narrowed it down enough that it has a chance of hitting it. So that's one reason. But there's another, deeper reason that I think a lot of people have missed out on. And so for this, I'm going to do another kind of thought experiment for everybody, and that is, imagine that you have an organism that only has one gene, okay? So by some miracle, it only needs one gene. This gene can mutate to 10 different configurations, okay? And each of these configurations are designed ahead of time. So all ten options are designed. So it is only mutating in designed ways.
Now, for any given environment, one of those configurations is going to be better than the others. All right, so let's say in Environment A, mutation 3 is best, and environment B, mutation 7 is best, okay? So in any given environment, if the organism stays in the environment a long time, then eventually whatever configuration is best. Because organisms grow by exponential growth as it stays in the environment for a long time, that configuration is going to far dominate all of the configurations.
And so if you have mutations, since most of the organisms are in the optimal category already, most of the mutations are going to be to a lesser configuration. And so if you stay in the same environment, even if all of the mutations are designed, you can actually expect that most of the mutations are actually going to be at least slightly deleterious.
[00:17:39] Speaker B: Where are you taking this idea going forward?
[00:17:42] Speaker A: So one of them is just to kind of come up with a better model of mutations.
This particular idea of understanding the dynamics of mutations, I think will help us just understand the process better.
We've got this background assumption, it's called the origin fixation model. So we have the origin of a new mutant, and then it has a fitness effect and it carries on.
But what if that isn't how mutation actually happens? What if we need a new population dynamics model that takes into the account the idea that in large part, we're predisposed to mutations that mutate us in certain ways? In fact, there's a book called the Implicit Genome that kind of starts heading down this path, says, what if instead of just having a genome, you also have an implicit genome of what are the likely possibilities within a species or some grouping of organisms to. Of a possibility space that they're likely to mutate to? And how does that change our picture of biology? So I'm doing some simulations, population dynamic simulations. It's Kind of early stage work, but I've been working a lot on that. I'm also as far as measuring mutations and measuring the beneficial nature of mutations, I've been trying to get some biologists to work with to do more measurements in the field. So I came up with some ways to do some measurements for experiments that had already been done. But I would like to do some prospective experiments on can we look at these particular mutations? Because one thing that I think is interesting is that by having a general way of determining whether or not a mutation is beneficial, it takes a lot of experiments to determine if a mutation was directed. There's actually not an experiment that people do determine that something's not directed. They just say that it's not directed. But if you want to say that something's directed, you have to go through several experiments to show that it is directed in some way. And here's how it's directed, and here's all these sorts of things. And so by having a general method of finding out if a direction mechanism is there, you can then spend your resources better to find which mutations are directed. And so which mutations should we look for a mechanism on?
[00:19:55] Speaker B: That's fascinating.
Where does intelligent design fit into all of this?
[00:20:00] Speaker A: So if you have direction from the cell, the question is, where does that direction come from? On its face, you could follow this train of thought without necessarily following intelligent design, but there's kind of a deeper mathematics behind it. It's called a search for a search. So if you have a search that hits a target, you can measure, you know, how effective that search is at hitting a target. But if you have a search for a search that actually is harder to find than the search itself. And so mathematically, some people have said, well, we have these mutational mechanisms, but although these mutation mechanisms help us find better mutations now, well, in the past, those mutations came from random mutation and natural selection. But the fact is that as you push these in the past, they actually get harder. They get harder to accomplish, not easier.
It's not that it's easier to find a mutational mechanism that will give you results, and then you can use that easier mechanism to get the results.
It's actually a harder hill to climb. William Demsky did some mathematics research on this. It's known as the displacement theorem. And basically it says that you can push the design back and that that's fine, you can push the design back as. As far or as long as you want. And it still, it still makes sense, but it gets bigger, not smaller. Kind of the idea that a lot of people have is that if we push it in the past, it'll get smaller. And that's not the case at all.
[00:21:28] Speaker B: So do you have any publications or any resources that you'd recommend for people to go to to read out more on your work on the subject?
[00:21:36] Speaker A: Yeah, so I've published two papers with bio complexity.
So one of them is called evolutionary teleonomy as a unifying principle for the extended evolutionary synthesis. And that's kind of a mouthful, but of to break it down, what that means is evolutionary teleonomy. Teleonomy is when something works according to a purpose because of a code. So evolutionary teleonomy is understanding evolution as going towards a purpose because of a code. And basically I'm talking about how it would be useful. There's a new outlook on evolutionary theory called the extended evolutionary synthesis. The extended evolutionary synthesis is meant to be kind of a replacement for, for Darwinian evolution, which is known as the modern synthesis. But the problem is that nobody in the extended synthesis has really laid down what are the claims, what are the core principles of the extended synthesis. And what I did is I went through kind of what the extended synthesis talks about and I showed how this idea of directed evolution kind of speaks to most of the things that the extended synthesis is claiming. The other one is measuring active information in biological systems. And that's where I go through the mathematics of how you measure the directedness of mutations. I show some examples, both kind of theoretical examples and some actual practical examples in biology. I came up with one that was not previously known to be directed. There's still not been an experiment to show what the mechanism is, but I was able to show that mathematically. It sure looks like there's directional happening there. So those are my two major papers. I'm also working on a mutation dynamics simulator. It's called the implicit genome simulator and it's on GitHub.
So if somebody wants to, you can find that pretty easily on GitHub by searching for implicit genome simulator. And so those are kind of the main places to start with my work.
[00:23:37] Speaker B: And you also have a video lecture on this subject.
[00:23:41] Speaker A: I've posted a lot of videos. I. I have a kind of my own research I do under the auspices of the Blythe Institute. And We've got a YouTube channel, YouTube.comblithe institute. There's not an E on Blythe. Yeah. YouTube.comblytheinstitute and it has a lot of video lectures on a variety of topics actually, but a lot of it focused on directed mutations.
[00:24:05] Speaker B: For our listeners who aren't familiar with the Blythe Institute, could you give us a brief summary of what that's about and what its goals are?
[00:24:11] Speaker A: So our journal is Communications, the Blythe Institute, and we publish things that we find interesting. It's a peer reviewed journal and we've been cited in a number of different journals. We've been cited in Science, Robotics, been cited in BMC Biology, Bio Complexity, Journal of Mind and Behavior, and a couple other ones. So anyway, so it's a very small journal, but I think we have an outsized impact for the amount of man hours that have gone into it.
[00:24:41] Speaker B: Excellent. Well, check out the Blythe Institute and Jonathan's video output on YouTube as well as his academic papers on directed mutations. But thanks so much Jonathan for coming on ID the Future. It's great to have a conversation with you. I'm Jonathan McClatchy for ID the Future. Thanks again for listening.
Visit
[email protected] and intelligent design.org this program is copyright Discovery Institute and recorded by its center for Science and Culture.
