JAD ABUMRAD: Hey, I'm Jad Abumrad.

ROBERT KRULWICH: I'm Robert Krulwich.

JAD: This is Radiolab. And this next segment began with a simple question.

STEVE STROGATZ: Mm-hmm.

ROBERT: Seeing as our topic so far has been limits, and we've done body ...

JAD: And we did the brain.

ROBERT: Now we're gonna go really big.

STEVE STROGATZ: Yeah.

JAD: Yeah, so we called up Steve Strogatz, mathematician at Cornell University, frequent guest on the show, and we asked him ...

ROBERT: Are there limits to human knowledge?

JAD: Yeah. And his answer sent us on a little adventure.

STEVE STROGATZ: The—yeah. Is there anything that's at the limits of our knowledge is a question that a lot of us scientists worry about. And—and certainly the 20th century taught us that there are many things that limit our knowledge, for instance, the Heisenberg Uncertainty Principle in quantum physics showed us that you can't know the position and momentum of a subatomic particle at the same time. You just can't do it. It's not a matter of not having good enough instruments or not being clever enough. It's just a fundamental barrier that nature puts in your way.

STEVE STROGATZ: In logic, Gödel’s Theorem tells us you can't prove certain things even though they're true. So we—there are all kinds of limits, but those seem a bit remote from everyday experience. And yet I think there are really important limits on our knowledge that we're all familiar with. What I'm thinking of here is our inability to think about big numbers, because with your fingers you've got 10, you know, normally. So we're good at 10, we're barely good at 100. And once you start getting to thousands, millions, billions and trillions, it gets hazier and hazier. When you hear now about the trillions of dollars in the deficit or whatever it is, the debt, you know, we don't—that means nothing. How are you supposed to think about that? Now when you ask why can't we understand the common cold but we can put a person on the moon, it has to do with large numbers.

JAD: Not just large numbers of numbers, says Steve, but large numbers of things interacting.

STEVE STROGATZ: There are so many genes involved and so many biochemical reactions involved, our brains are limited, our memories are very limited, and so I worry a little bit that—that we might be approaching the end of our ability to have insight into certain kinds of questions.

ROBERT: What Steve means by the word "insight" is not like you found the answer, it's like that ...

JAD: It's like a feeling.

ROBERT: Right.

JAD: You know, like that "Oh, I get it!"

ROBERT: A feeling you get when you really understand the answer.

STEVE STROGATZ: Yeah, that satisfying feeling that I can see the reasoning.

JAD: I can actually feel it in my bones.

STEVE STROGATZ: That's—that's a very pleasurable feeling, but one that we may not always be able to enjoy.

HOD LIPSON: I mean, you can see the space.

JAD: We weren't really quite sure how to feel about this ...

ROBERT: Right.

JAD: ... but then Steve said, "You know, don't take my word for it. Talk to these guys that work down the hall from me. You'll see."

HOD LIPSON: Yeah, we can—we can go right ahead.

JAD: Cool. Can you guys introduce yourself? Tell me who I'm talking to.

HOD LIPSON: Yeah. So my name is Hod Lipson.

MICHAEL SCHMIDT: My name is Michael Schmidt. I'm a PhD student.

HOD LIPSON: And I'm a roboticist.

JAD: And Hod and Mike have developed this thing, which does make you wonder if Steve's right. It's a computer.

HOD LIPSON: Yes.

JAD: Actually, many.

HOD LIPSON: A whole tower of computers that are all grinding away and performing calculations.

JAD: Actually, when you get down to it, it's just a piece of software. But they've named it ...

HOD LIPSON: The Eureka.

JAD: Because that's what it was designed to do, to have "Eureka" moments.

HOD LIPSON: Let—let maybe a kind of simpler example ...

JAD: And the story of Eureka begins pretty simply. With a ...

HOD LIPSON: ... think of a regular pendulum, okay?

JAD: With a pendulum.

HOD LIPSON: Just one of these things you see hanging off a grandfather clock.

JAD: Okay, I've got a regular pendulum swinging in my mind.

HOD LIPSON: Okay. Swinging left and right.

JAD: Now, says Hod, double it. Instead of a string connected to a ball, make it a string connected to a ball connected to another string connected to another ball.

MICHAEL SCHMIDT: Which is basically like a double pendulum. The cool thing about this is you just put it up, you lift it up and let it go.

JAD: And what you'll get, says Mike ...

MICHAEL SCHMIDT: Is chaos. This really crazy behavior.

JAD: Instead of nice and even, now you got random.

MICHAEL SCHMIDT: It's almost impossible to actually try to predict where this thing will move.

JAD: So what they did was they got a camera, connected it to Eureka, and basically just had Eureka watch this thing, you know, move about crazily. And then they asked the computer a really simple question: can you make some kind of sense out of this erratic behavior? Like, is there something in this system that always stays the same?

HOD LIPSON: Tell me what about these pendulums over time is not changing?

JAD: Because with everything, there's gotta be some kind of logic in there.

JAD: So you're looking for a law, basically. I mean, you're looking for the law of the double pendulum.

HOD LIPSON: Yes, that's the idea.

JAD: So Eureka is there watching this pendulum.

MICHAEL SCHMIDT: It was about 3:00 am in the lab.

JAD: And it's basically spitting out all of these different guesses.

HOD LIPSON: Formulating hypotheses.

MICHAEL SCHMIDT: It's getting closer. Closer.

JAD: And then onto the screen pops this simple formula: F=ma.

JAD: What is F=ma? Is that actually the law that ...

HOD LIPSON: F=ma is Newton's law of motion.

ROBERT: The Isaac Newton.

JAD: That's Sir Isaac to you.

HOD LIPSON: It's a basic law of physics.

JAD: And one of the greatest discoveries in the history of human thinking.

HOD LIPSON: Took it about a day, 24 hours. But—but the interesting thing is that it came up with this thing without knowing anything about physics. Nothing. That's why we kind of—we think that this algorithm might be able to find new laws that we don't know about yet.

ROBERT: In fact, once word got out about Eureka ...

JAD: That's when the emails started.

HOD LIPSON: A couple of emails a day.

JAD: From scientists all over the place who were like, "Hey ..."

ROBERT: "Do you mind if we borrow your robot?"

JAD: For what kinds of stuff?

HOD LIPSON: Anything you can think of, from trying to predict behaviors of cows in a herd, to particle physics, to the stock market.

ROBERT: And that's—and this is when we get to Steve's point about the limits of insight.

JAD: That's when they met this guy.

GUROL SUEL: My name is Gurol Suel.

JAD: Gurol is a biologist.

GUROL SUEL: At the University of Texas Southwestern Medical Center.

ROBERT: He got in touch with Hod.

HOD LIPSON: And he said, "I have this amazing data, which is single cell dynamics."

JAD: Meaning he's got this tiny little thing.

GUROL SUEL: It's a simple bacteria.

JAD: Really basic.

ROBERT: And he's been collecting this information on how it works.

JAD: On its inside.

HOD LIPSON: How things go up and down. Certain nutrients increase, certain nutrients decrease over time, just like a pendulum.

JAD: But the thing is, in a cell it's like thousands of pendulums.

HOD LIPSON: There's so many parts.

ROBERT: Genes turning on and off.

GUROL SUEL: Thousands and thousands, tens of thousands.

JAD: Proteins turning on other genes and nutrients going up and down.

ROBERT: It's this crazy quilt of complicated feedback.

JAD: And he wanted to know inside of this cell, how are all of these things related? I mean, we can measure it all. We can see things going up and down and all that.

ROBERT: But what are the rules?

JAD: What are the rules? And this, he says, is the problem for biology.

GUROL SUEL: Biology's one of the least well-understood systems compared to, let's say, chemistry and physics.

JAD: They're still lacking the basics.

GUROL SUEL: So we said, "Look, Mr. Robot, can you tell us what you think are sort of the important principles governing this organism, and maybe detect things that were hidden from us?"

HOD LIPSON: So he sent us the data, and we analyzed it and ...

ROBERT: Well, okay, let's—yeah, so what happened?

HOD LIPSON: ... suddenly equations started popping out.

GUROL SUEL: Almost immediately. The robot came back to us and said, "Okay, here's a set of two equations that describe your data."

JAD: Do you remember by any chance, what the—what the actual equation was? Not—not that we'd understand it, but just sort of to hear it said out loud.

GUROL SUEL: Yeah. No, I don't. I don't have my Rain Man skills developed to that degree yet.

JAD: The important thing is that the equation was telling him things like when this protein goes up, this other thing always goes down. And when that thing goes down, this gene turns on and does a loop-de-loop. And when he went to his cell to check all this out, the equation was right.

HOD LIPSON: These equations matched the data, and in fact, they explain new data.

JAD: These equations could even predict what the cell was about to do. But hold the champagne, there's just one little problem here. The formulas check out, but ...

HOD LIPSON: We don't know what they mean.

JAD: You don't know what they mean?

HOD LIPSON: Right.

JAD: Meaning they don't know why these equations work.

ROBERT: Right. Why when this goes up, does that go down? Why when that goes up, does this go sideways? Why?

GUROL SUEL: I had to first look at this and try to make sense of it. We said, like, "Oh, okay. I think we understand." And then we're like, "Oh, maybe we don't." We think that we're close to understanding it.

HOD LIPSON: But, you know, now we're in this bizarre situation. We can't even publish it right now because we can't just publish a equation without explaining it.

JAD: So in the end, they're in this awkward position where they've got the answer but they don't have the insight.

HOD LIPSON: And I think it's a preview of what's to come in science.

JAD: The more we turn to computers with these big questions, the more they'll give us answers that we just don't understand.

HOD LIPSON: We'll be faced with this challenge of having to find ways to get a computer to explain what it found.

STEVE STROGATZ: But that will leave us, if this really happens, in some weird position as bystanders, where we're sort of listening to the oracle, but not really understanding the answer. Is there gonna be a time when we—we can't cut it anymore? We've had this—this window in human history when we could not just know things but actually understand them. That is, you could know why they were true. Not just know, but to know why. And that's a beautiful moment in human history, but I feel like it may only be a moment.

GUROL SUEL: Well, I don't really see it quite that—that sort of sad and dramatic.

JAD: [laughs]

ROBERT: [laughs]

GUROL SUEL: Because at the end there will be simple principles to describe even the most complicated of processes.

JAD: So you have a bias that prevents you from feeling the kind of despair that Steve feels and that we were hoping you would feel.

GUROL SUEL: Oh, well. I have a positive outlook.

JAD: [laughs]

ROBERT: Well, I'm just wondering about the "we." "Look what we have discovered," you'll say when you're an old man with your robot sitting there in a dress next to you.

JAD: [laughs]

ROBERT: And the robot will be holding your hand, but that will be a cold hand. And Jad and I will be thinking, "I don't know, who's the 'we' here?" Is it like the ...

GUROL SUEL: Well, I would say "we" is sort of knowledge. I'm just thirsty for understanding and thirsty for knowledge. Me and the cold hand holding my hand, we've accumulated and contributed to the overall understanding of something that we thought maybe 50 years ago wasn't possible, and that would be something that would make me happy.

[STEVE STROGATZ: Hi, this is Steve Strogatz. Radiolab is produced by Jad Abumrad. Our staff includes Ellen Horne, Michael Raphael, Soren Wheeler, Lulu Miller and Pat Walters, with help from Adi Narayan, Tim Howard and Sharon Shattuck. Special thanks to Steven Auerbach. All right. Thank you. Bye-bye.]

[ANSWERING MACHINE: End of mailbox.]

 

-30-

 

Copyright © 2023 New York Public Radio. All rights reserved. Visit our website terms of use at www.wnyc.org for further information.

 

 New York Public Radio transcripts are created on a rush deadline, often by contractors. This text may not be in its final form and may be updated or revised in the future. Accuracy and availability may vary. The authoritative record of programming is the audio record.