Episode Transcript
[00:00:07] Speaker A: Welcome to ID the Future, a podcast about intelligent design and evolution.
[00:00:14] Speaker B: Welcome. I'm Tom Gilson. On today's episode, we hear Andrew McDermid speaking with discovery Institute Senior Fellow J. Richards at the recent National Technology Summit conference in Bellevue, Washington, technology titled simply KOSM. Jay shares some history on an earlier such conference, Telecosm 98, and he talks about a book that came out of it on the topic of artificial intelligence, some debate on the topic, and also what's happened on issues since then.
[00:00:45] Speaker C: Jay, thanks for taking the time to speak with me today.
[00:00:47] Speaker D: Thanks, Andrew. It's great to be with you.
[00:00:49] Speaker C: Well, we're here at the COSM Technology Summit and one of the speakers is well known inventories entrepreneur and futurist Ray Kurzweil.
Back in 2002, you edited an anthology of essays called Are We Spiritual Machines? Ray Kurzweil versus the Critics of Strong AI can you tell us how this book came about and why it was written?
[00:01:09] Speaker D: It was actually based, interestingly enough, on another Gilder conference called Telecosm. And so George with Forbes Publishing did a series of conferences like this for years called telecosm. And in 1998 it was held in Lake Tahoe. And Ray Kurzweil had a book coming out called the Age of Spiritual Machines. And so George had him speak and we realized, so Kurzweil is, I'd say, an advocate of strong artificial intelligence. So the idea that machines are going to not just sort of exceed us in certain capacities, but they're going to actually essentially wake up that whatever we are, machines will not only equal that, but they will surpass us. He doesn't think that's a dystopian picture. He thinks it's utopian. He thinks we'll effectively merge ourselves with technology and off we'll go in this, what he calls a singularity. So this moment, say some 2043 or sometime in the fairly near future, in which technological acceleration, the acceleration of technology will be so rapid that it will essentially be discontinuous with prior history.
And so George had the idea to have a panel, a sort of debate on Kurzweil's book. And so Kurzweil gave a talk and then he had Bill Demski, who was a fellow at Discovery Institute, a guy named Tom Ray, who's a kind of computational biologist, Michael Denton, John Searle, a famous philosopher and critic of strong artificial intelligence, respond to Kurzweil. What was interesting was that this was at the conference where Bill Joy of Sun Microsystems attended and Joy when he heard Ray Kurzweil talk about the merger of nanotechnology and robotics and artificial intelligence, he saw this as a dystopian picture. And so Bill Joy went home after hearing this and talking with Kurzweil at the conference and wrote an article that became a cover story in Wired magazine called why the Future Doesn't Need Us, which was essentially Kurzweil's sunny optimism about AI Bill Joy's hardcore pessimism. And in fact, he called for the government to basically put, stop, regulate these industries sort of out of existence. And so that's the kind of reaction. And so we thought, you know, actually this debate deserves a wider audience than just people that could afford to come to the telecosm conference. And so that's why George and I in Discovery Institute and Discovery Institute Press decided to condense the debate. So he needed the authors, rather than just giving, you know, extemporaneous comments, to actually write things. And Ray was actually gracious enough both to participate and then also to write responses to all of them. So it's really Ray versus several critics and then his responses to it.
We just really wanted it to accurately represent the debate about artificial intelligence, not necessarily to settle the debate. George and I give our opinion. We have a particular view on this at the beginning, but it's had an amazing shelf life in many ways because, I mean, artificial intelligence is still with us. And now here we are in 2019. So this is 21 years after Telecosm 98 and artificial intelligence was one of the abiding themes of the COSM conference here in Bellevue. That's just ended.
[00:04:02] Speaker C: Wow. So these conversations have been going on a while.
What are some of the arguments that were presented in the book?
[00:04:08] Speaker D: Well, so the famous argument in artificial intelligence is from John Searle, his so called Chinese room argument, which essentially I think what he does is in this Chinese room argument, he shows that computers work at the level of abstract symbols, at what you could call the level of syntax and rules. They don't work at the level of semantics or of meaning. And so the image is of a man that's locked in a room, essentially, and he has a bunch of cards that have these symbols on them. It turns out they're Chinese characters. He doesn't know any Chinese though. And there's a slot in one side of the room and he'll get a card, will come in, and then what he has is a bunch of other cards with a system of rules. So if this image comes in, then you need to take this other image on this other card and put it through the slot on the other side of the room. And so he does this, and he's able inside the Chinese room to follow these rules consistently. What he doesn't know is that there are two Chinese speakers outside the room that are actually communicating with each other through him in a way. Now, his position is that of the computer, the computer. We use computers to store, to duplicate, to transmit, to analyze, to sort information.
Computers don't create meaning, right? In fact, they don't know anything. In fact, if anything, the man in the Chinese room, at least he's aware of the fact that he doesn't know what the symbols mean. The computer doesn't even know that. And so his point is that you can have something doing, computation, doing everything that computers actually do, and it's still going to be completely dumb. It doesn't do the most basic things that an agent does in terms of self awareness, in terms of what philosophers call qualia, in which that's basically the first person experience you have of your observations. And just this capacity for self transcendence to be able to choose. And so humans work at the level of meaning and purpose. These machines that we make, as amazing as they are because we make them, nevertheless they're not conscious and they're not ever going to sort of compete with us at that level because they're fundamentally different kinds of things. And so Searle has continued to update this argument. Ray Kurzweil thinks he has a way around it. I don't find it at all persuasive, but we wanted them to represent that. And there have been millions of probably pages written on this Chinese room critique. But if anything, I think it's still one of the most fundamental critiques of the whole strong AI program, which is this idea that man is a machine and that ultimately our machines are going to be able to do everything that man can do.
[00:06:31] Speaker C: And I think it was William Dembsky that also contributed and talked about Kurzweil's impoverished spirituality.
[00:06:38] Speaker D: That's right. And so, Bill, notice something that I think even Ray would admit, which is that his account in many ways is a theological account. In fact, it corresponds sort of to the kind of the Christian cosmic vision, right, in which you have this, you know, you have the original Garden of Eden, you have a fall, you have a sort of redemption, and then you have eternal life in the kingdom of God. And for Ray, who's, you know, effectively a materialist, still has these spiritual concerns. And it's quite clear that he sees this future and the singularity in which we merge with machines as the ultimate source of our immortality. So if we're going to have any kind of actual immortality, it's going to have to come come about by way of our technology. And of course, that's what Ray himself, even personally, I think, is hoping for, is that he himself will live long enough actually to see this happen. And, you know, I think that there's no mistaking that, and I think it's just especially clear, I think, in Kurzweil's own work. But those of us that have thought and written and talked about this for a long time realize that the artificial intelligence debate at one level looks like it's a debate about technology and engineering, and it is. But it's also a debate essentially about metaphysics, about the question is, what's the nature of reality? What is a machine? What is man? All those questions are brought to bear in this AI debate, which is, frankly, what makes it so fascinating.
[00:07:58] Speaker C: So if human intelligence is viewed as the result of an unintelligent process that didn't have us in mind, how does that inform our views of AI and the relationship between man and machine?
[00:08:08] Speaker D: Well, let me put it this way. If you think that we are this result of this blind and purposeless material processes, then ultimately you think that whatever human beings are, are reducible to those things. Right? And so the laws of physics must be ultimately determining what we take to be thoughts and deliberations and decisions. We think they're free, but they're ultimately caused and determined from the bottom up. And so if you think that, and you think, well, we became conscious, or at least seem to think we're conscious, and we do all these things and can build buildings and machines, why on earth wouldn't we be able to build machines that could do the same things that can even exceed us, Right. And so in some ways, if you're a MA that thinks we're purely the result of these blind material processes, you have something to worry about. And so I sort of. I sympathize. It makes sense. On the other hand, if you go the other way, either you have a prior theological commitment and you say, no, human beings are material spiritual hybrids. We're not just material beings. Or if you're just really thinking, I think commonsensically. So what is the thing you know most directly as a person? Well, you have direct access to your existence as an agent. You experience directly what it means sometimes to be coerced, sometimes to choose things freely. When you treat people as if they're morally culpable for their actions. You're assuming that they're a free agent, right, that could do otherwise. And so you already actually believe that we're more than mere matter, that we're more than mere machine. And so my argument would be, look, everyone, whether atheist or Christian or whatever, actually already knows that we're more than machines, we're intelligent agents. And if you start that way, then you want to say, okay, so what would need to be true about reality in order for that to be possible?
And if you start there, then you can look at machines and say, okay, well, well let's be open minded about what these things are going to achieve. But you don't need to worry that machines are going to replace us because they're not what we are anymore than you need to worry that oxes are going to disappear, right? If you make a really good tractor, or that if you make a tractor that's strong enough, it becomes an ox. I mean, it's just a weird kind of metaphysical mistake that we would never make. And in some ways, either because of kind of bad metaphysical assumptions or because of our language, we talk about machine learning, right? We talk about artificial intelligence. And that language, if we're not careful, which is really metaphorically, can get away from us so that we really do start imagining Skynet waking up and the robots deciding they're going to kill us and things like that. I just don't think there's any good reason to think that's true.
[00:10:32] Speaker C: Yeah, our imagination can run away with us sometimes with
[00:10:35] Speaker D: all this. And it makes sense from a sort of literary standpoint, right? I mean, nobody wants to read a book or watch a movie about a robot that doesn't wake up. I often joke, I say, look, nobody's going to watch a three hour movie about the ship computer on Star Trek. That's a very boring movie. You need a data, right, that's actually doing things and it seems like a person. And so that makes sense. There's kind of narrative needs to do that. But just don't mistake that for what's actually happening inside the guts and the so called brains of a computer.
[00:11:04] Speaker C: Yeah, for sure.
Well, Kurzweil, in the essay that begins this collection of essays, he said that the purpose of life is to evolve. And he also notes that technology goes beyond mere tool making.
It's really a continuation of the evolutionary process. Now I've actually looked up technology in the dictionary and it says it's the sum of the ways in which groups put Together the material objects of their civilization.
So it makes me wonder, is technology to be used as a tool or should it go beyond can it? That's really the topic here, isn't it?
[00:11:42] Speaker D: It is. And notice the dictionary definition is fin and much more commonsensical. I mean, we get technology from techne, right? And so these are these. I mean, he's challenging that traditional idea that technology is more than mere tools. Now. I think that's right, insofar as we don't just make tools, but tools also shape us, Right? And so once you learn to read, for instance, literate cultures, you know, we tend not to remember things as well as pre literate cultures. We tend to get nearsighted because we read text, right? We're being changed in some ways by our screen. So in that sense, it's more than tools. There's a feedback between us and the tools that we make. But his statement is ultimate, kind of a metaphysical claim, because what he's really saying is that effectively there's this long process of cosmic evolution that started in the physical world, so the inorganic physical world. And then there was an inflection point where it went to the biological stage, for instance, which has lasted, let's say, 4 billion years. And so we've had this stage of biological evolution. And what humans are is we're essentially the last stage of biological evolution and the beginning of the next stage, which is technological evolution. So you see how you sort of tied this together in this grand cosmic vision. We're right at the cusp. We're the last members of, you know, last installment of biological evolution, and we're the ones that are bringing about our successors, essentially. And so technology will take over from here. Well, that's quite an arresting vision, but I think it just allows you. Your mind is running away from you and you're thinking about technology in that way. I mean, the reality is that rather than looking at technology and saying, wow, it's going to totally replace us, every piece of technology that's ever existed has been created by human beings applying their ingenuity to the material world around them and transforming it, right? And so I actually think that the Christian theological context explains this much better.
If you think human beings are part of the material world made from the dust of the earth and the breath of God, right? So yes, we're material beings. We're also spiritual beings. And so we should expect that if we're made in the image of this creative God, that when we apply our minds, when we apply our spiritual reality to the material world, that we have this capacity to create things that were not there before. So we can take the sand that God has created and create fiber optic cables and computer chips, right, Made of silica sand. Effectively that actually, I think that picture actually makes sense of what's happening with technology and information technology. And in particular is where the role of creativity and information, the role of ideas comes to predominate over the physical kind of substrate of those things. And so I think that's what's actually happening in information technology. We're moving more and more from the world of atoms to the world of bits, right? From molecules to information.
And we are the kind of transition point between this, the kind of material and the spiritual, or the material and the informational. That actually makes a lot more sense. But in some ways what we've got is competing metaphysical visions that have implications and can make predictions. So I'll just make the opposite prediction that we're not going to have a singularity, anything like what Ray Kurzweil is predicting. And having a person is nowhere near as simple as being able to have a high resolution scan of the brain. That's just not going to work.
[00:14:58] Speaker C: Yeah.
Well, as we wrap up, we don't have too long here, but what has changed in the 20 years since this book came out and what remains the same?
[00:15:10] Speaker D: I would think that the thing, at least in the discussion that's changed the most is that we don't tend to think, for instance of software as kind of simple deterministic algorithms anymore. Of course we talk about machine learning. And so it's not like for instance, that we're going to get a system like Siri that's going to maybe eventually understand natural language. That's not going to happen by computer programmers literally programming every rule. What you have is a system of feedbacks, which is like the Google search mechanism, right? So the Page Rank system, it's not pre programmed at the beginning. It's an algorithm that's set up to take account of the decisions of people doing searches, right. And then to bias future answers based upon what people are doing. So it's really a hybrid of a complex algorithm that has a way of feeding in and taking account of the decisions of people. And so that's really, that's what machine learning is. And so you don't have to pre program everything in the beginning. You actually create an algorithm that's able to take account in a sense of experience. Right. And so that's how things actually work. So in some ways, if you thought, oh, Artificial intelligence will never be able to give us autonomous cars. It will never be able to do any kind of work doing language translation. Then you're going to be surprised because it's actually starting to happen. And so I think, if anything, that. So we need to be more realistic and in some ways more optimistic about the things technology can achieve. So that's changed, at least sort of at the level of the public conversation. What has not changed is the basic questions, right? Are we machines? Are machines going to become conscious? Are we intelligent agents? Those are these fundamental questions. And they're exactly the same in 2019 as they were in 1998.
[00:16:47] Speaker C: Wow. Well, this book remains important and relevant to the debate, and I hope we can continue reading it. And there might even be another edition that comes out to Mark 20 years later.
[00:16:59] Speaker D: Absolutely. It's good stuff, and I don't think it's going anyway anytime soon. And the idea of being able to intersect essentially theology, philosophy, engineering and science, I mean, who wouldn't want to talk about a subject like that? And that's what artificial intelligence is.
[00:17:13] Speaker C: Wow. Well,
[00:17:14] Speaker B: that was Jay Richards along with Andrew McDermott at the recent COSM conference in Bellevue, Washington.
If this podcast has been helpful and informative to you, please help us raise its visibility by going into your favorite podcast app and giving Intelligent Design the Future a five star rating.
I'm Tom Gilson for ID the Future.
Thank you for listening.
[00:17:38] Speaker A: This program was recorded by Discovery Institute's center for Science and Culture. ID the Future is copyright Discovery Institute.
For more information, visit Intelligent Design.org and IDTheFuture.com SA.
