[00:00:01] Speaker A: Religious and philosophical tradition that takes teleology seriously as something real would be open to that.
[00:00:12] Speaker B: So in that conception, an irreducible intelligence is something that can basically solve problems.
[00:00:19] Speaker A: That, you know, if they had to be solved without intelligence, would be highly improbable.
[00:00:26] Speaker B: So, so that's you might say that.
[00:00:27] Speaker A: Intelligence is what they do, is they create.
[00:00:31] Speaker B: Create information that.
[00:00:35] Speaker A: You know, if it had to be accounted for in terms of chance, would be highly improbable.
[00:00:45] Speaker C: Idaho the Future, A podcast about Evolution and Intelligent Design.
[00:00:53] Speaker D: I think we can all agree that nothing comes for free in this life. There's a cost, even if we aren't the ones paying the bill.
And so it is with the fundamental building block of all life, and that is information.
Welcome to I Do the Future. I'm your host Andrew McDermott, and today I conclude my four part conversation with Dr. William Dembsky about his new monograph, Unpacking the Law of the Conservation of of information.
Dr. Dembsky is a founding and senior fellow with Discovery Institute's center for Science and Culture and a distinguished fellow with the Institute's Walter Bradley center for Natural and Artificial Intelligence. He's a graduate of the University of Illinois at Chicago, where he earned a bachelor's in psychology and a doctorate in philosophy. He also received a doctorate in mathematics from the University of Chicago in 1988 and a master of Divinity degree from Princeton theological seminary in 1996.
He has held National Science foundation graduate and postdoctoral fellowships. Dr. Demski has published in the peer reviewed mathematics, engineering, biology, philosophy and theology literature.
He is the author or editor of more than 25 books, most recently a brand new edition of the Design Inference, co authored with Winston Uret, a book you're most certainly going to want on your bookshelf.
Well, in this concluding segment, Dr. Dembski discusses the ultimate origin of information, saying there are only three frontloading, brute chance, and something he calls irreducible intelligence. Dembski explains which of those options seems the most scientifically and philosophically robust.
Well, without further ado, let's return to the conversation now. This is part four, the conclusion of our time together.
[00:02:46] Speaker E: Now, drawing on Douglas Axe's work, you mentioned that functional protein folds are as rare as 1 and 10 to the 77th. How does this law then turn this rarity into a lack of evolvability that material processes can't overcome?
[00:03:01] Speaker A: That number is one that Doug computed, you know, and it was he was looking at a beta lactamase protein that, as it changed it conferred antibiotic resistance on ticulobacterium and he argued that on.
[00:03:17] Speaker B: The basis of selective pressures, it was.
[00:03:21] Speaker A: Still that improbable to get from some protein that was, or some protein fold that was doing something different to that.
[00:03:30] Speaker B: So, you know, so you had.
[00:03:32] Speaker A: You have here a highly improbable event that's happened, and then how do you account for that?
[00:03:38] Speaker B: I think at this point, maybe it.
[00:03:41] Speaker A: Would be good to draw a distinction between conservation of information and the concept.
[00:03:47] Speaker B: That your listeners may be more familiar with, which is this earlier notion of mind, specified complexity.
And I think what Doug has identified is specified complexity. So the idea with specified complexity is it's a type of information that is at once improbable, but has a pattern that is easily characterized.
[00:04:12] Speaker A: So it combines what we might call Shannon and Kolmogorov information.
[00:04:17] Speaker B: But I think the way to see it is as follows.
[00:04:23] Speaker A: Just bear with me for a moment.
Imagine you got a SETI researcher search for extraterrestrial intelligence, and it's listening to a radio, Listening for radio signals along.
[00:04:36] Speaker B: Some channel, and he gets a sequence of eight characters.
[00:04:43] Speaker A: Bits, you know, everything is in bits, let's say.
[00:04:46] Speaker B: And it's the eight bits or the.
[00:04:50] Speaker A: Byte needed to encode in Unicode letter A. That could be the indefinite article transmitted by an alien intelligence that knows English.
[00:05:00] Speaker B: It might be, but it's 8 bits.
[00:05:04] Speaker A: 1 in 256 probability on a uniform distribution.
[00:05:09] Speaker B: If you're monitoring millions or billions of.
[00:05:11] Speaker A: Radio channels, you will witness that sequence not just once, but very often.
[00:05:16] Speaker B: Could you conclude by just receiving that.
[00:05:19] Speaker A: Sequence that you're dealing with an intelligence? No.
[00:05:22] Speaker B: Now, it could be that there was.
[00:05:24] Speaker A: An alien intelligence that was communicating and saying, you know, punching in on a keyboard. A great day has emerged in the history of our civilizations where we finally made contact with planet, the inhabitants of Earth, et cetera.
[00:05:39] Speaker B: Okay, but the thing is the.
[00:05:42] Speaker A: That alien intelligence had a cerebral accident right after punching in the letter A.
[00:05:50] Speaker B: So that letter A is here by intention. It was designed, but because it's so short, lacks complexity in terms of bits.
[00:05:59] Speaker A: We will not be able to detect that it is the result of intelligence.
[00:06:02] Speaker B: But if in Unicode you had that.
[00:06:06] Speaker A: Entire sentence or a paragraph, a great day has arrived in the history of our civilizations, etc.
[00:06:13] Speaker B: That would be enough complexity even at.
[00:06:15] Speaker A: The scale of the entire universe, you know, with 10 to the 80 atoms, 10 to the 90 elementary particles, you.
[00:06:23] Speaker B: Know, that would still be enough, you.
[00:06:26] Speaker A: Know, to reliably nail down that we're dealing with an intelligence.
[00:06:31] Speaker B: But there's more to it, right? Because, I mean, just having a Long sequence of random signals, random bits would.
[00:06:38] Speaker A: Also be highly improbable, but wouldn't tell us anything.
[00:06:41] Speaker B: So what's special about this? Well, it's in English. The vast majority of radio signals would.
[00:06:48] Speaker A: Not be in any natural language, to say nothing of being in English, to say nothing being encoded in Unicode.
[00:06:58] Speaker B: So there's this specification, this recognizable pattern.
[00:07:01] Speaker A: That can be, in complexity theoretic terms, briefly described, which would satisfy the compressibility in a Kolmogorov sense plus complexity in this probabilistic sense. There's a long sequence of bits.
[00:07:18] Speaker B: So it's kind of a tension.
It's high in probability, and that corresponds.
[00:07:25] Speaker A: To the narrowing of possibilities. But then also on the pattern end, it's recognizable. It's briefly described. Oh, that's an English sentence. Oh, that's Shakespeare or whatever.
[00:07:35] Speaker B: And so when you get that convergence.
[00:07:37] Speaker A: That tells you that you're dealing with an intelligence.
[00:07:40] Speaker B: So that's the gist of my book.
[00:07:43] Speaker A: My research, Design Inference and Design Inference two, which you were so kind to cite and which I hope your listeners will buy and read, if they haven't.
[00:07:55] Speaker B: Already now, specified complexity as I develop it. There is a reliable marker of intelligence.
[00:08:02] Speaker A: And if you've got it, you know, if you witness it, boom, you've got intelligence.
[00:08:07] Speaker B: And I would say what Doug Axe does is he gives us specified complexity.
[00:08:12] Speaker A: Shows that these systems, you know, they confer antibiotic resistance, they fold, these are folds, you know, so that itself is a specification. Now, most things are not going to fold in any sort of functional way.
[00:08:26] Speaker B: So that, So I would, what I.
[00:08:29] Speaker A: Would say is, and then he has.
[00:08:31] Speaker B: This improbability of 10 to the minus 77.
[00:08:34] Speaker A: It's maybe not at the level of the probability of, you know, probabilistic resources.
[00:08:39] Speaker B: Of the whole universe, but it's, it's, it's pretty good.
[00:08:42] Speaker A: I mean, you know, it's really good.
[00:08:45] Speaker B: So I would say, you know, he's got specified complexity, and with that he can draw design inference that this system, this, if you're getting this beta lactamase.
[00:08:56] Speaker A: You know, in the first place, you know that's going to be exhibit specified complexity and therefore design.
[00:09:06] Speaker B: So in a sense, conservation of information won't apply.
What conservation of information does is to say, you know, this is, this is also where the.
So this is where the Darwinist would come in and say, well, wait a second, Doug. You haven't really looked at all possible evolutionary paths.
[00:09:28] Speaker A: You're claiming that this is unevolvable. That's really not, you know, that it's that it's super improbable that it's a case of specified complexity, but it's really not. Because if you really knew what natural selection was doing here, if you knew all the underlying causal pathways, you'd realize that actually it's, it's quite probable.
[00:09:52] Speaker B: So what always happens in these cases is that the Darwinists see natural selection or some other mechanism as a probability amplifier.
[00:10:05] Speaker A: Okay?
[00:10:07] Speaker B: Now we've heard probability of amplifier before.
[00:10:09] Speaker A: In this discussion in the last one, but they really, they do see natural selection as a probability amplifier.
[00:10:16] Speaker B: And so what seemed to be specified complexity?
[00:10:19] Speaker A: Because specified complexity presupposes that you have.
[00:10:21] Speaker B: A very highly improbable event, right? I mean, remember the SETI example, it had to be improbable.
[00:10:27] Speaker A: That sequence that was being set, you couldn't just be the letter A had to be highly improbable.
[00:10:34] Speaker B: So once you got that, you know, if you've got that improbability working for.
[00:10:41] Speaker A: You, you've got specified complexity and you've got design.
[00:10:43] Speaker B: So the way Darwin has tried to upset the specified complexity apple card is by saying, well, you really don't have improbability.
[00:10:51] Speaker A: You don't really.
[00:10:52] Speaker B: It just looks like it's specified complexity, but it's really not.
[00:10:56] Speaker A: Okay?
[00:10:57] Speaker B: The probabilities are actually bigger, okay, because of natural selection, okay? And so that's where conservation of information kicks in. Because it says, okay, we thought that.
[00:11:11] Speaker A: This was highly improbable. It was specified complexity, this event.
[00:11:14] Speaker B: You're telling us that there's a probability amplifier, in this case natural selection, that raises the probability. Okay, but then there's a probability of that probability amplifier. What's the probability of that? How is it that the amplifier worked.
[00:11:30] Speaker A: To amplify the probability in that?
[00:11:32] Speaker B: I mean, it's this example we gave before. You know, you got an arrow that, that lands on a target far away, target highly improbable to land there. So there's a favorable wind that blows the target, blows the arrow to the target. So the probability of the arrow landing given the favorable wind now is suddenly very big, okay? But the probability of that favorable wind.
[00:11:56] Speaker A: Is now itself very improbable.
[00:11:58] Speaker B: And so likewise, getting a selection rich, capable environment that is able to overcome.
[00:12:08] Speaker A: These improbabilities in precisely that way is itself highly improbable.
[00:12:12] Speaker B: But I think people kind of see natural selection, the naturalistic biologists, they see natural selection as this kind of magic bullet that automatically just explains itself. But natural selection can take lots of different forms and lots of different contexts. Where, you know, often it can do nothing of interest.
You know, there's an example, an experiment.
[00:12:38] Speaker A: Of Saul Steelman in the, I think.
[00:12:40] Speaker B: It was the 60s in which he put, I think, was it some RNAs and a replicase environment.
And basically they evolved and they, but they didn't evolve to complexity, they evolved to simplicity. They became much more simple because that's. Simplicity had a selective advantage, you know, and often simplicity has a selective advantage. You know, it's just easier to simplify, to have less, less baggage, you know, so, so there, you know, there's nothing magical about selection as such. It can, it can act in an incre.
[00:13:22] Speaker A: Complexity increasing way.
[00:13:25] Speaker B: As with the Avita program, when complexity was rewarded, it can act in other ways. So selection itself can take many different forms and for it to act as.
[00:13:38] Speaker A: A probability amplifier for a given case.
[00:13:40] Speaker B: Means that it's taken a highly select.
[00:13:43] Speaker A: Highly selective form, you know, highly improbable form.
[00:13:47] Speaker B: And so that's, that's where I would say conservation of information.
[00:13:53] Speaker A: It acts as the, you know, two.
[00:13:56] Speaker B: Punch in the spend, you know, specified complexity is the number one punch.
[00:13:59] Speaker A: So it's a one, two punch specified complexity.
[00:14:02] Speaker B: If you've got it, you got design. Oh, you're claiming natural selection as a probability amplifier.
[00:14:08] Speaker A: Okay, now that probability amplifier is highly improbable. So you got the improbability problem again, you got the specified potential problem again.
[00:14:17] Speaker B: Then you want to get rid of that. Okay, now you got, now conservation of information kicks in now, specified complexity conservation. And it keeps going up and back, but you never escape that sort of.
[00:14:29] Speaker A: Dialectic between the two and you're trapped.
[00:14:33] Speaker B: You don't get around the design problem.
[00:14:36] Speaker A: And it's, you know, it's a point.
[00:14:39] Speaker B: I've made also for years. Maybe it's just that I haven't made.
[00:14:42] Speaker A: This stuff clearly enough. You know, maybe I just need to get on Joe Rogan, you know, and you know, just have, have some, some.
[00:14:51] Speaker B: Really nifty diagrams and ways of putting.
[00:14:54] Speaker A: It that will finally.
[00:14:55] Speaker B: Yeah, across.
[00:14:56] Speaker E: Well, I like the one two punch thing. I think Joe would as well.
[00:15:01] Speaker A: Yeah, we'll see about that.
[00:15:03] Speaker E: Yeah, no, no, that's good. The one two punch. And that does explain how these two concepts that you've put forward work together. I like that.
Now another term that you use is irreducible intelligence. And I hadn't heard that term until you used it here and I really like it. How do you define the term then? And how does the law of conservation of information lead us to a final resting place? So to speak, an irreducible intelligence rather.
[00:15:36] Speaker D: Than that infinite regress, you know, turtles.
[00:15:39] Speaker E: All the way down kind of thing.
[00:15:40] Speaker B: Yeah, well, I mean, just to backtrack a little bit. I mean this conservation of information, it does apply, imply a regress.
[00:15:49] Speaker A: So if you've got something highly improbable to explain, you have something that explains it, you know, raises the probability, then that thing is highly improbable.
[00:15:57] Speaker B: You want to explain that in naturalistic terms. Well then you've got another highly improbable.
[00:16:02] Speaker A: Event, you know, another amplifier, another improbability to explain.
[00:16:07] Speaker B: Now it seems you can only go back so far, you know, it's.
[00:16:11] Speaker A: If we're, if we agree that we're in a finite universe with a finite.
[00:16:16] Speaker B: History.
[00:16:19] Speaker A: You know, this sort of conservation of information regress has to stop at.
[00:16:23] Speaker B: Some point, point and there it seems that there are only a few possibilities.
[00:16:28] Speaker A: You know, I mean broadly, it seems you're either going to have to say that all the information was there from the start.
You can say that it is just pure chance, but you know, we're just.
[00:16:43] Speaker B: Going to live with it, you know, it's just chance got really lucky, you.
[00:16:46] Speaker A: Know, and we're here we are, you know, all this fine tuning, all this evolution took place maybe invoke a multi universe, you know, so there, I mean.
[00:16:59] Speaker B: If you're, you know, you can use.
[00:17:02] Speaker A: Metaphysics to, to, you know, destroy any scientific theory, you know, I mean it's, you know, you can make metaphysical assumptions, you know, and all your problems go away. So, you know, but then you raise other problems, right? I mean, so what's the, what's the status of this multi universe?
Well, what the metaphysical status. Do you really want to settle yourself with that?
[00:17:25] Speaker B: But broadly you've got, you know, all.
[00:17:29] Speaker A: The information was there from the start, you know.
[00:17:33] Speaker B: Well how did it start? You know, what's, is there anything behind that?
[00:17:40] Speaker A: You know, does it just become absurd? You know, it's just something we accept seems absurd, you know, but it's, do we go with a multi versus reverse, some sort of, you know, just brute chance explanation or what would be another possibility?
Another possibility is what, what I would call an irreducible intelligence, you know. What would be irreducible intelligence? Well, it's where you keep, you know, explaining the ample probability amplifier in terms of another probability amplifier, you know.
[00:18:10] Speaker B: But you know, at some point you.
[00:18:11] Speaker A: Come to find, you know, final resting.
[00:18:14] Speaker B: Place of explanation and say, okay.
[00:18:17] Speaker A: This.
[00:18:18] Speaker B: Is produced by an intelligence that does.
[00:18:19] Speaker A: Not have to be accounted for in terms of some Sort of improbability.
And that then becomes a third possibility.
It's not going to be naturalistically acceptable because, you know, from a materialistic bandage, intelligence is a byproduct of matter organizing itself. Intelligence results from an assemblage of particles, you know, you know, but from a certainly a Judeo Christian vantage, God is not an assemblage of particles. God's mind is not an assemblage of particles. It's not a pattern. And if we're made in the image of God, then, you know, we're also not in our mentality is not patterns. Which is actually very good news if you're a, you know, if you're not a computational reductionist.
So, you know, and I think what.
[00:19:23] Speaker B: I'm describing, I mean this is from.
[00:19:25] Speaker A: A Judeo Christian point of view, but I think just about any religious and philosophical tradition that takes teleology seriously as something real would be open to that.
[00:19:39] Speaker B: So in that conception, an irreducible intelligence is something that can basically solve problems.
[00:19:47] Speaker A: That, you know, if they had to be solved without intelligence would be highly improbable.
[00:19:54] Speaker B: So that's, you might say that intelligence.
[00:19:56] Speaker A: Is what they do is they create.
[00:19:59] Speaker B: Create information that, that, you know, if.
[00:20:06] Speaker A: It had to be accounted for in terms of chance would be highly improbable.
[00:20:09] Speaker B: So it's information rich.
And that's, you know, so that's, that's.
[00:20:15] Speaker A: What, what I'm referring to as an irreducible intelligence in this paper, you know, God, ourselves, angels, you know, other beings could fulfill that role.
[00:20:29] Speaker B: There could also be derived intelligences that.
[00:20:32] Speaker A: You know, is it where act as intermediaries between these irreducible intelligences.
[00:20:38] Speaker B: But since we're, you know, we're looking.
[00:20:40] Speaker A: At the origin of the universe or the origin of this information where you come from in the first place. Ultimately you do, it seems you are going to have to get to some sort of irreducible intelligence. You know, derived intelligence would be reducible, know to something more fundamental.
[00:20:55] Speaker E: Right. And even Richard Dawkins, you know, the eminent biologist and representer of all things Darwin, was prepared to admit, you know, as far back as excelled or expelled, sorry, the documentary, that hey, perhaps, you know, life was seated, sort of a directed panspermia approach. You know, life was seated by an alien force. And.
But then again, you know, what about the alien force? And you can go back ad nauseam, but I think it has to stop somewhere. There has to be an ultimate reality at the end of every worldview or philosophical approach to things.
[00:21:33] Speaker A: There can be some Other possibilities, you know, we're part of a simulation. And then that simulation, it's itself created, you know, by an irreducible intelligence. And, you know, you could have a cyclic universe, but even then, you know, but basically, I don't think you escape this, this idea of an irreducible intelligence. To your point about expelled, you know.
[00:21:56] Speaker B: Dawkins didn't just say that life may have been seated.
[00:21:59] Speaker A: He said that life may have been designed by such a. Basically an alien super intelligence.
[00:22:06] Speaker B: You know, so he allows design there, but then he immediately comes back and says, but even that alien intelligence must.
[00:22:12] Speaker A: Ultimately have been explained by some explicable process. And by explicable he means something purely naturalistic.
And the only candidate for him is natural selection.
[00:22:26] Speaker B: So everything has to funnel through natural.
[00:22:31] Speaker A: Selection at the end of the day for him.
But he basically, though, allows that life on Earth may have been designed, only it couldn't. Couldn't have been designed, by God. It would have been alien intelligence.
[00:22:44] Speaker B: But even that, I mean, that, that.
[00:22:46] Speaker A: Just becomes arbitrary, doesn't it? I mean, you know, we're an alien, but.
But not, you know, not an intelligence that's unembodied.
[00:22:58] Speaker E: Yeah.
Well, final question, Bill.
Hopefully we're piquing, you know, the curiosity of some folks out there with, with this new law that you've been working on.
What, you know, of all the work that you've put out and the books and the monographs. Where do you think is a good place to start for someone who wants to learn more about this and perhaps CSI as well as another pillar of id?
[00:23:24] Speaker A: Yeah, well, you know, it's interesting the.
[00:23:29] Speaker B: The second edition of Design Inference, which.
[00:23:31] Speaker A: Is where, you know, that's the place to go if you want to understand specified complexity.
We intended actually to include conservation of information in that. But.
[00:23:43] Speaker B: And the thing is, I mean, conservation of information, it's been an idea that's.
[00:23:46] Speaker A: Been out there for a long time. Even this law of conservation of information.
[00:23:50] Speaker B: The way I conceived it before was as a.
[00:23:53] Speaker A: As a sort of generalization of a number of theorems that were out there. It was only in the fall of.
[00:23:59] Speaker B: Of 2023 that it kind of the.
[00:24:01] Speaker A: The dime dropped for me where I saw. Wait a second. All these theorems are actually just special cases of a very simple probabilistic relationship. And so that's when I wrote this up.
[00:24:12] Speaker B: But I would say get second edition.
[00:24:16] Speaker A: Of the Design inference to understand specified complexity. Yes, there are some parts that are technical, but there's a lot that's just very accessible. I would say chapters one and two certainly, you know, I would say sections on specification and probabilistic resources dip into them. And then there's a chapter six on biology. I think that would be of general interest.
That's chapter seven. Chapter six is probably the most technical. It's about.
[00:24:48] Speaker B: But there's a lot that's just there.
[00:24:50] Speaker A: And I would say don't be intimidated by the time technical stuff just skip over it because there's always going to be stuff that's of interest with regard to the paper or monograph on law of conservation of information. That's paper that was published 20, 25 in Biocomplexity.
[00:25:12] Speaker B: It is a monograph. I mean it's not short, It's I.
[00:25:15] Speaker A: Think 50, 52,000 words. So it's, you know, it's book length.
[00:25:20] Speaker B: But you know, in many ways it functions as a review article.
[00:25:23] Speaker A: So I've tried to make it very accessible. There is a bit of math in there, but most of it is just motivated by a lot of examples and explanations. So, you know, I'm hoping. I'm actually working with two co authors on spinning that off into its own book, you know, and I think most of most of that text will be preserved and I think it'll be of general general interest.
[00:25:46] Speaker B: So I, I wouldn't be intimidated by.
[00:25:48] Speaker A: The fact that it's in the bio, this journal Bio Complexity. I would just go, I mean it's free, it's monograph and as I said.
[00:25:56] Speaker B: Also it serves as a review article. I have an appendix in there that's about 6,000 words which, which just covers.
[00:26:04] Speaker A: The entire, the, the history of conservation of information and law, conservation of information, information going up to the present. So it's, you know, I think there's just a lot of, you know, if you're, if you're really interested in this, I would, I would just go in there and you know, and if there are technical things that you're not comfortable with, just skip them. But I think there's a lot that you'll just pick up on and I.
[00:26:31] Speaker E: Do think that the everyday person can grasp this. I think you're right that it's, it's elementary and there's a simplicity there.
So again, the website's thedesigninference.com for getting a hold of the second completely revised new edition of Bill's work, the Design Inference and then we'll include a link to the monograph as well out of Bio Complexity. Well, Bill, thank you very much for your time. It's very valuable and appreciate the work you're doing and, and the chance to unpack this with you to.
[00:27:03] Speaker B: Yeah, it's my pleasure, Andrew.
[00:27:05] Speaker A: Thank you. I appreciate this thorough questioning. So it's nice to have to think on my feet and get these ideas out and hopefully in a more accessible way.
[00:27:16] Speaker E: Yeah, yeah. And we'll see if we can get Joe Rogan interested in the nature of information, you know, because it's, it's real and it's great. It's. It's really good stuff to wrap your head around because it does show you that there's a cost, you know, and when you apply it to Darwinian evolution, you start to, you know, understand the.
[00:27:39] Speaker D: Evaluation that we've made that.
[00:27:41] Speaker E: That says, look, you can't do this. This sort of approach is not going to get you mathematically to where we are.
And I think we need to be honest about that.
Well, don't forget, listeners and viewers, you can watch these interviews now as well as listen to them on our YouTube channel. Subscribe there and be the first to know about all the videos that are coming out. That's YouTube.com d the future.
[00:28:07] Speaker D: Well, I'm Andree Dermott.
[00:28:09] Speaker E: Thank you very much for joining us.
[00:28:11] Speaker C: Visit
[email protected] and intelligent design.org this program is copyright Discovery Institute and recorded by its center for Science and Culture.
