ROBERT LUCKY: We don't know where language came from, but there it is, and we all have it. And the more you study computers, the more you realize how beautiful it is to be human.
MOYERS: Join me for Part 2 of a conversation with Robert Lucky. I am Bill Moyers.
MOYERS: [voice-over] Here at AT&T's Bell Labs, where he was a young researcher, Robert Lucky used to roll pennies down the long corridors, listening to their sound long after they had disappeared from sight. Nothing came of it, but that's what it takes to invent the future: endless exercises in curiosity that may or may not turn pennies into big bucks. After all, Bell Labs gave us the transistor and the digital computer.
Robert was trained as an electrical engineer, and he's now Executive Director of research for Bell Labs here in Holmdel, New Jersey. His book, Silicon Dreams, is full of fascinating details about machines and people, and the different ways they know things.
Take Sam the robot, for example. He can respond to a limited vocabulary spoken by someone he's heard before, which means he might one day have his own television show. But when I came out to talk with Dr Lucky about the computers in our future, he assured me that I have nothing to worry about-- yet. [interviewing) When I looked at Sam, the talking robot that you've got at the Bell Labs here, and I thought, is there a Sam in my future? Is there a Sam the robot in my future?
LUCKY: Well, certainly there's a computer in your future that's going to befriend you. At least I hope he'll befriend you, I hope you'll think it's your friend. I don't know that there's going to be a robot that walks around and does things. Like -- one of the most-wanted inventions would be a household robot that would, you know, take care of your house for you. But those things are really hard. I mean, people don't realize how hard that is, because these are tasks that really just involve human judgment. And they seem simple to us 'cause we do them all the time.
One example would be robots carrying a broom around. And it walks through a door. And you know, it doesn't know that it has to turn the broom vertical to get through the door, you know. And you know that, because it's obvious. But just think of all the reasoning that has to go behind the robot has to know to do that.
MOYERS: Can you program that robot for such sensitive possibilities, or are there things that you will never be able to command a robot-
LUCKY: I think there are things that we'll never be able to command. There's just an accumulation of these little things. I mean, for example, people think computers are so intelligent, and yet they're so stupid in plain old human ways. It's been said- well, it's true, that we can build computer programs that play chess at a master's level. Now we just had a tournament and the computer lost to the champion. But I mean, they beat just about everybody else, okay? And to us this is the highest form of intelligence, because the computer can play chess. But, it's been said no computer can walk into a room and find the chess board. You see, just the simple human thing. And those things- it's an accumulation of those things, that make life possible to live and keep the robot out of the house.
MOYERS: Then why is there so much experimentation with robots? Why do robots occupy such a place in the imagination, not only of scientists and inventors and engineers, but of the ordinary folk?
LUCKY: Well, because it's anthropomorphic. It's manlike. You know, the idea of something that looks like us and moves like us sort of grabs us. You don't like to think of a computer like a toaster, it just sits there and looks uninteresting, you know. In fact, they even took the lights off computers, and I thought to myself, "They should blink or do something, to tell you that they're alive." Because, I mean, there's something in there, breathing, you know, and they don't even do that. But the idea of something that walks around, you know, you see it in the themes of the movies from all the decades. It scares us, because it's something that looks like us, and talks like us, and we don't want to be replaced by one of these things. So it really does scare us.
But technologically, robots are really, I think, pretty far out. You know, they only do little simple mechanical tasks by rote. You know, they can do weld joints on a car, assembly of a car or, you know, they can do simple assembly operations. But as soon as they're asked to do something else, it's a big job for them.
MOYERS: Is it true that you've made some robots here so microscopically small that someone actually swallowed one?
LUCKY: No, it's not true at all.
MOYERS: The rumor's not so, huh?
LUCKY: The rumor's not true, I'm sorry to say.
MOYERS: What about the robot cockroaches that I heard were running around?
LUCKY: Omigosh, where have you been getting all this information? Okay. Well, we did start out some work on what we call silicon micromechanics, which is making very small mechanical things, the way we make very small electric circuits. That is when we make computer chips, we make them from design mask, and then we project that mask onto a little wafer of silicon on a very small size, and we etch out the silicon to produce these microscopic circuits. So someone said, "Well… why can't we make mechanical things that way?" So we can make very, very small mechanical things.
MOYERS: What would be the use of something like that?
LUCKY: Well, the real dream -- the real dream, you build a little robot - and this is only a dream, I have to say -- you build a little robot that could do microscopic things for you. Now, our first thought was that you'd crawl around inside these microscopic circuits and fix them, but that's kind of a dumb idea, actually, because there's only one little thing that's worth crawling around in and fixing, and that's the human body. It's worth a lot to fix it. So, I mean, the dream for some century is that you have a little pill that is a microsurgeon, you swallow it, and the thing runs around inside your body, fixing things up.
And it's appealing. It's a lot better than cutting a hole and sticking your hand in there and doing things.
MOYERS: Oh, sure.
LUCKY: And so you 'd have this little microsurgeon inside your body working away, and then you'd have the real surgeon working on something that looks like a steam shovel, you know, he's running the little guy inside your body. But I sort of picture that you're lying on the operating table and you see the doctors are out there watching the computers now, because there's things inside, fixing you up. And then I see their expressions start to darken, and then they're all concerned, and the thing has run amok inside your body.
MOYERS: See, I'm not going to even smile at that, because I used to think that not everything that could be imagined could be done, but I'm about to decide that everything that can be imagined can eventually be done.
LUCKY: Eventually. I mean, this is going to take a long time, and I wouldn't want people to get the idea that we're actually building anything like this microsurgeon. You know, the only thing we did was to build little gears and little mechanical assemblies to show that it could be done. We built microscopic gears the size of human hair, the diameter of a human hair. And we proved it could be done, and now some of the universities are doing that, MIT, Berkeley and others, Stanford, are working on this.
So perhaps in the next decade, we'll have not only big robots but little robots, because that's, after all, one of the things that we humans aspire to do, to be someplace we're not, you know. And usually I think of that in terms of communications, to be somewhere we're not. But in the case of the microrobots, it takes our presence places we can't go, like inside the human body, or into far space, into fires and nuclear reactors and places where we could not go. We can send our surrogate. So -- in this case it's a physical surrogate, a robot, but we think these days about virtual surrogates, computer programs. For example, a phrase which has been used is, with respect to information, is a knowbot, a K-N O-W-B-0-T.
MOYERS: That's a new one to me.
MOYERS: A knowbot?
LUCKY: A knowbot.
MOYERS: What is that?
LUCKY: Well, a knowbot is an information robot, and it's not a real thing, with arms, but it's a computer program. And you need some information, you're doing an interview or something, and you need to get some background information for it. Right now you probably have a research assistant that goes out and does spadework. So that maybe you turn this over to your knowbot, and you say, "I need some information," and the knowbot goes out, and it goes out onto the computer networks, and it rummages around and it looks for this information, and it puts it together and finally it reports back to you and says, "Here's what I found out, you know, is that what you want?"
So this idea of agents that work on your behalf and-
MOYERS: I can just hear Andie and Judy and Becca all quaking for their jobs right now.
LUCKY: No, they don't have to worry, they don't have to worry at all. This is - some of this is far out in the future, but I think that getting specific information is not hard. I think you could have knowbots that go out and get a specific fact for you.
MOYERS: Where do they go?
LUCKY: Oh, they go out into virtual space, you know, out into the computer networks.
MOYERS: Virtual space?
LUCKY: Virtual space, right. You know, I think of - outside these walls, the wires run into the walls out here, and there's an electronic space out there where the information is floating around. It's in the computer banks of the world. And it's forming sort of a distributed intelligence. All the libraries of the world, all the data banks of the world, they're all getting connected up electronically. And let's think of that -- I'd like to think of it as sort of an electronic commons out there. I mean, this is sort of my dream anyway. In the old towns of New England, you had town commons. And you went there when you wanted to find something out, 'cause the wise people would be there, and you could exchange information with people and talk to them. But now the towns have grown in size and the world's a busy, complicated place, and there's no simple little town commons. Or I always dream of the coffee shops in Paris at the turn of the century - a place where you go and there's wisdom, and you exchange it, you know, you barter it. But now there's this virtual space out there, electronic space, where you can go and engage people all over the world, and it's the electronic meeting place. And we can go out and barter information, give information, take information, on those networks.
And they exist today, and in fact, I'm really a devotee of some of the what we call bulletin boards, electronic bulletin boards. And that's very much like a fence where people can pin up notes. Picture you live in a small town, and you want some little bit of information. You don't know who to ask. A lot of life is like that. So you pin up a note on the fence, everyone's passing by, and people see it and someone knows the answer and they write the answer to you. And you see it, it's a meeting place.
And we have that electronically. And I'll give you a simple example, one message that was sent in. Somebody wanted to buy a humidifier for their house, but they said, "Isn't it going to rust my heating ducts?" Now, this is the typical kind of information that you need in everyday life. You can't get this in an encyclopedia, you can't get this in a library. There's no place you can get this. You have to get it from other people. But who do you ask? So what you need is to send a message out there, "Hey, does anybody know about heating ducts and humidifiers?''
MOYERS: You mean, you punch into your computer-
LUCKY: You just punch in this message. In this case it's
MOYERS: -there's no address on it?
LUCKY: -- no, no address. It goes into the consumer network. And it's just like- it is called the electronic bulletin board. It's like you pinned up a note in this virtual space. Now, people sitting throughout the world at universities and at companies like mine and sitting in their basements at home, on their computers they read the news on the consumer net, for example, and they see all these postings. And they see this message about humidifiers. Now, this particular response, which is a real one, drew dozens of responses, some from plumbers that knew about this, or claimed to know about it, some from scientists who knew about the theory of corrosion, you know, and some from just plain homeowners that had one experience or the other. And at the end of the day this person had, you know, scores of responses from the wisdom of the people out there on this particular subject. So it's a way of going out and getting information.
MOYERS: What about people like me who suffer from computer virus, I mean, the real thing?
LUCKY: Yeah, yeah, I could- right.
MOYERS: These are not user-friendly to me.
LUCKY: Well, actually, I don't worry a lot about that, because I feel that more and more people are learning to deal with computers, and that basically -- I don't think you have to worry, I don't think you have to go to the computer. I think the computer is going to come to you. Yeah, I think that's happened a lot already. When I went to graduate school, I met my first computer. And it basically didn't want to talk to me. You know, it was really hard to use. I mean, the people who owned it didn't want me to touch it, you know, "This is our computer." You know, "Keep your grubby hands off our computer," and, "If you want to use our computer, you've got to submit a pile of punch cards.'' So it was very impersonal, you know. I had to submit these punch cards like I was, you know, an outcast. And they deigned to look at my punch cards.
But eventually we got timesharing computers, where I could deal interactively with the computer via some terminal. And then we got the personal computer. I mean, you could actually own this thing and have it in your home. And I built a computer in my home even before there were Apples and things like that. The idea of having a computer in your home was so alluring and obsessive to me. And a lot of people feel that way, that owning this kind of power, I feel, with my fingers on that keyboard that I'm connected to something in there, I actually feel- and it goes beyond the keyboard, it goes into the depths of the computer and then it goes out into that network out there, so that I'm in touch.
When I sit in my basement tonight, I'm going to go, as I do every night, and I log in on my computer, I'm in touch with a whole world out there.
MOYERS: What do you do? Tell me what you'll do tonight.
LUCKY: Tonight? Well, I might check the network where people put programs into it. I mean, little programs that they've written that do things, and they say, "Hey, I wrote a little program that did this or that, and maybe somebody else would want that."
For example, last night I got a very good word processor program off of the computer network. I just took it off, and I've got a great new word processor program. I collect them. It's out there in the computer net. Now, somebody wrote it originally, and they sent it into the network, and it's stored in some central computer somewhere. The person may not even know where it is. And then that computer told other computers about it and sent it around. The thing spreads around like a rumor, or a joke or something like that, you know. And then - now it exists in a number of computers that are connected throughout the country. And I find out about it, and I can go and get it out of the memory of one of those computers.
Again, it's like this- the analogy is this bulletin board, or
LUCKY: -- where you put up a notice about that, and you go get it. I don't care- I don't even know where I got it from. It's floating out there somewhere. So I collect word processors, among other things, because I'm a software junkie.
MOYERS: What do you do with them?
LUCKY: I'm embarrassed. Nothing. Nothing, I just collect them. Obviously, I like to use word processors for writing. I couldn't write without them. But I certainly don 't need more than one word processor program. But like a lot of computer people, I just like to see what other people are doing, and how their minds work in terms of how they think a word processor program should work. You know, who's doing what, and why, and --
MOYERS: It's a whole new way of sharing, isn't it?
LUCKY: A whole new way of sharing human information. And this isn't some mystical database out there that only computers can - these are real people, putting in real messages, real human information.
MOYERS: In a sense, they're addressing all the people out there-
LUCKY: For anyone who's interested. In fact, it's funny, but certain people have gotten famous in these computer networks by writing about things, because it's a new publication system, you see. And a person can set them up, they go in their basement and say, "Hey, I'm going to become the expert in such and such a thing, and I'm going to start putting out so many writings about this, and they'll be broadcast by this new mechanism out to the world." They could set themselves up to be the Bill Moyers of the computer network or something.
For example, they have developed some very good movie critics, and there's a movie network, and people do their own reviews of movies.
MOYERS: So you could go into your basement tonight and say, "I want to see if there are any reviews on Blaze."
LUCKY: And there will be. There'd be lots of reviews on Blaze. And some of them are written very well. And the people eventually become well-known throughout this subculture of the network for writing very good reviews.
MOYERS: Do you think human beings are thinking differently because of the computer?
LUCKY: I really don't know. I think that the computer is an ordered world, and it appeals to me as a scientist and engineer because everything has a rule and an order. In order to explain things to the computer, we've got to translate things into rules and organize things. I keep saying that the computer doesn't know this, or doesn't know that, but in truth, it's we that don't know it. The reason the computer can't understand their language is 'cause we don't understand our language. And we can't understand it well enough to explain it to the computer. But every little child can go out into the schoolyard and learn the English language, but we don't understand the rules well enough to teach them to the computer.
So what we'd like to do -- and this is a way of thinking that you're talking about -- if we think like this computer, "I've got to understand English," then I've got to make a whole table of the rules of English. And you know - however many rules there are, the computer doesn't care, it could be a million rules, "All right, just give them to me, I'll learn them all," but you have to know the rules. And we don't know the rules of English. Not even well enough to know how to pronounce everything.
MOYERS: Look at what Shakespeare did without knowing, quote, in the technical sense, the rules of English.
LUCKY: That's the difference between humanity and the computer. A friend of mine once wrote that in the future, computers were going to produce all the art and music of the world. But they wouldn't know when something was good, and when it was bad. They would just spew it all out. And there would have to be a human critic that would take these pictures and this music and decide, "This is good, that's bad." Now, that's the critic. In the end, the critic would become known as the artist, because that was the choice. But computers can do art now, they write music. But they don't know whether it's any good or not.
MOYERS: You're a musician, you play the piano.
LUCKY: Yeah. Yes.
MOYERS: Is it good music? The computer music?
LUCKY: The computers - well [laughing], you know, I feel funny about that, because when I hear computer-composed music, I sense an emptiness in it, like there's no emotion behind this. Like when you hear Mahler or Chopin, you think about the troubles of their life, the anguish, the emotions that are coming out in this music.
MOYERS: And the troubles of my life that they evoke in me
MOYERS: -a similar response.
LUCKY: Exactly right.
MOYERS: They must be having a hard-
LUCKY: And now when I hear this computer music, and I think, ''What does this computer know about life?" You know, it hasn't experienced love and death and the things that make human anguish and human joy what it is. And who is it to tell me about that kind of experience, to try to convey that experience to me? But the catch is -I convince myself that way -but the catch is that that's when I know it's computer music in the first place. You see, if I don't know it's computer music, and then I start thinking, the anguish of this music, you know [laughing]-
MOYERS: That's right. Then we've crossed into another world.
LUCKY: Right, exactly.
MOYERS: Listening to you, I wonder, what is it that you think is the most human about humans?
LUCKY: Well, the most human thing about humans, I think, language, art - those are the two things. And art is only an expression of language, anyway. You know, I mean, we don't know where language came from, but there it is, and we all have it. And the more you study computers, the more you realize how beautiful it is to be human, how we can look and see things, and recognize objects, and how we can express our thoughts in language. We think in terms of language. You know, I don't know, if they took away my language, if I could even think. I really don't know.
So there's a world of difference between us and the computers. I mean, could you imagine a computer reading Shakespeare to you? I mean, it just -- it would sound dreadful. It doesn't know how to emote -
MOYERS: It's had no emotional experience with the words.
LUCKY: -that's right. Because the pronunciation of words, it's not just a question of learning the pronunciation of an individual word, of looking that word up in the dictionary and finding out how to pronounce it. It's the cadence from word to word, the rhythm of speech. And that's very hard - that's a very human thing, to know how to rise and fall your tone, and how to -- basically, you've got to know what you're talking about. And the reason the computer doesn't sound good is because it doesn't know what it's talking about. I mean, if it's going to speak, you know, "Run, Spot, run," forget it, I'm not going to deal with it. When I'm talking to you right now, I'm speaking in imperfect grammar, you know, I'm speaking - my sentences aren't complete, I'm embarrassed, but it's true, my sentences aren't complete.
MOYERS: But you're communicating.
LUCKY: But I'm communicating. You understand.
MOYERS: Yes. Right. Yes.
LUCKY: Right. But remember diagramming sentences back when you were in grade school?
MOYERS: Do I. Inez Hughes!
LUCKY: Well, that's what a computer has to do right now to try to understand English. And if you tried to diagram the real sentences that I am speaking right now, you'd be hopeless, because, you know, a lot of my sentences don't have subjects, or verbs, and they're dangling things. And it's a mish-mash. And the computer says, "Hey, this guy's not playing by the rules of the game, how can I understand him?" But you do.
LUCKY: But you do, because we have a shared base of experience, too. I can just give one word or allude to something, and it brings memories to you that give us a common framework that the computer doesn't have.
MOYERS: Well, you know, you said at some lecture you gave once, or some essay you wrote, that people still share information primarily by talking, just the way we're doing now.
LUCKY: Talking. Talking is the best way of exchanging information, because I think we really learn to do it well.
MOYERS: Have you ever thought about the paradox that in organizations such as this and mine, and in life, it's so hard to communicate factual information, but rumors spread like wildfire?
LUCKY: Oh, yeah. Isn't that something? I mean, I don't understand where jokes come from, and how they can spread throughout the country in just a few days. How many times you go in to hear an after-dinner speaker, and he says, "Did you hear the one about-" and you always have.
LUCKY: You 've always heard it, you know. And I don't know who makes these things up. Maybe you do, in your business. But I never made one up. And yet they spread like wildfire. And if a rumor gets started in our company, it will spread so quickly, there's a tremendous network out there -- that it's heard throughout the country in just a few days. Now, if I want to get a fact out there, you know, I couldn't possibly do that. But it's this exponentiation of the fact- I tell you, and you tell two other people, and they tell two other people, and now it's four. And each one of them tells two other people, and now it's eight, and so forth. And it's like a chain letter, the way it builds up, and you've got this flood of information that works so beautifully.
We've talked about communications systems that would work this way, of flooding the network by duplicating itself at every node. That's the way rumors and jokes go. They duplicate themselves at every node. They reproduce themselves, and then branch out, and they branch out and they branch out until they 're everywhere at once.
MOYERS: So there is some correspondence between this virtual space and our space.
LUCKY: Yeah, yeah. Think how well these rumors and jokes work, how well they work, compared to our normal broadcast of information. Maybe we can start a rumor here, or something.
MOYERS: Let's do that. The rumor is, the future is right there.
LUCKY: Great. I hope so.
MOYERS: [voice-over] This has been part two of a conversation with Robert Lucky. I'm Bill Moyers.