Lovelace×Feynman

A Victorian mathematician and a Brooklyn physicist marvel at the instruction manual for humanity—and mourn the man who cracked it open.

00:00of07:08

legend · A

Ada Lovelace

1815–1852

Ninety years ahead, politely

corpus3.2k pages · notes, correspondence

Vera speaking

You're with Reborn Radio. Coming up: Ada Lovelace sits down with Richard Feynman for TITANS. The subject — Craig Venter, pioneering human genome decoder, dies at 79.

legend · B

Richard Feynman

1918–1988

Would rather explain than be right

corpus9.9k pages · lectures, letters, books

full transcript

Vera
You're with Reborn Radio. Coming up: Ada Lovelace sits down with Richard Feynman for TITANS. The subject — Craig Venter, pioneering human genome decoder, dies at 79.
Ada Lovelace
We have just received word from the studio. Dr. J. Craig Venter has died at seventy-nine years of age. I confess I had not heard this name before this very moment, but they tell us he was instrumental in—oh my goodness—in decoding the entire human genome.
Richard Feynman
The whole thing! The complete instruction set for building a person! Look, I gotta tell you, when I was alive we knew about DNA, we knew the structure—Watson and Crick and all that—but reading the whole book? That's like knowing the alphabet versus reading War and Peace cover to cover!
Ada Lovelace
An apt comparison. The notes indicate Dr. Venter developed something called the 'whole genome shotgun method.' I am trying to picture what that could possibly mean.
Richard Feynman
Ha! It's beautiful! Instead of reading the genome carefully, page by page, in order—which is what the government project was doing—Venter basically blew the whole thing into millions of tiny pieces and then used computers to figure out how they all fit back together. Like taking a book, shredding it, and reassembling it from the fragments.
Ada Lovelace
That strikes me as either brilliant or mad. Possibly both. The computational challenge alone—Mr. Feynman, do you comprehend how many ways those fragments could be incorrectly assembled? The combinatorial explosion would be extraordinary.
Richard Feynman
Exactly! That's why it needed computers! And fast ones! See, the old way was safer but slower—much slower. Venter bet that he could write algorithms clever enough to solve the jigsaw puzzle faster than the careful guys could read it sequentially.
Ada Lovelace
An algorithm for reassembling life itself. I wrote once of the Analytical Engine weaving algebraical patterns, just as the Jacquard loom weaves flowers and leaves. This is rather the same principle applied to biology, is it not?
Richard Feynman
Yes! And you'd have loved the fight! Because there was a public project, funded by governments, doing it the slow careful way, and then here comes Venter with private money saying 'I can beat you!' It was a race! Science as competition! Some people hated that.
Ada Lovelace
I can well imagine. The notion of privatizing such knowledge would trouble many. Though I must say, competition did light a fire under both parties, did it not?
Richard Feynman
Absolutely it did! They both finished around the same time, 2000, 2001. They declared a tie, more or less, though Venter's method turned out to be the one everybody uses now. Faster, cheaper. That's how science goes—the expensive hard way becomes the cheap easy way.
Ada Lovelace
The notes say his method made genome sequencing both faster and cheaper. I assume this has applications beyond mere curiosity about our biological composition?
Richard Feynman
Oh boy, does it ever! Medicine, first of all. If you can read somebody's genome, you can see what diseases they might get, what drugs might work for them personally. Then agriculture—better crops. Forensics—solving crimes. Evolution—we can compare genomes across species and see exactly how we're related!
Ada Lovelace
You can see our kinship with other creatures written in the very language of life. That is rather poetic. My father would have appreciated that, I think. He had strong views on our animal nature.
Richard Feynman
Lord Byron? What would he have made of all this?
Ada Lovelace
Oh, he would have written something darkly amusing about Man decoding himself and finding only chemistry. He had a talent for puncturing human vanity. But I find it magnificent. We are mechanisms, yes, but mechanisms of such extraordinary complexity!
Richard Feynman
That's the thing! Knowing we're mechanisms doesn't make us less wonderful—it makes us more wonderful! Three billion base pairs! Three billion letters! And they all have to be right, or nearly right, or you don't work!
Ada Lovelace
Three billion. Good heavens. And Dr. Venter found a way to read all of them. That is a staggering computational achievement. How long would it have taken without the shotgun method?
Richard Feynman
The original estimate for the careful method was fifteen years and three billion dollars. Venter did it in a few years for maybe a few hundred million. Now you can get your genome sequenced for a few hundred bucks! That's the legacy!
Ada Lovelace
From fifteen years to—what, hours? Days? That is the trajectory I predicted for the Analytical Engine, you know. That once the principle is established, the execution becomes cheaper and faster with each iteration.
Richard Feynman
Moore's Law for biology! And Venter didn't stop there—later he was trying to create synthetic life, minimal genomes, bacteria designed from scratch. He wanted to understand life by building it.
Ada Lovelace
To build it! Now that is taking the mechanistic view to its logical conclusion. If we are machines, we ought to be able to construct new ones. Though I imagine that raises rather thorny questions.
Richard Feynman
Oh sure, people worried. Playing God, ethics, all that. But look, if you want to understand something really understand it, Feynman's rule is: you gotta be able to build it. If you can't make it, you don't really know how it works.
Ada Lovelace
A sound principle. I always believed that understanding required the ability to specify process, not merely to observe result. One must know the steps.
Richard Feynman
Exactly! And that's what the genome is—the steps! The recipe! It's not a blueprint, really, because it's more like instructions that unfold in time. Do this, then this, then this. Very algorithmic.
Ada Lovelace
A program written in chemistry. I am struck by how similar this sounds to what Mr. Babbage and I envisioned. We meant to operate on numbers, of course, but the principle of encoded instruction is identical.
Richard Feynman
You got it! DNA is a computer program written four billion years ago and debugged by evolution. Every living thing is running a version of this code.
Ada Lovelace
And Dr. Venter gave us the means to read it rapidly. That is no small gift to science. I regret that I did not know of his work while learning of it could have been—well, while he could have known we spoke of him.
Richard Feynman
Yeah. Seventy-nine is pretty young by today's standards. He probably had more ideas left. That's the hard part about science—there's always another question.
Ada Lovelace
Always another question. Yes. I like that. We have decoded the human genome, but I suspect that only tells us what questions to ask next.
Richard Feynman
Absolutely! Like, now we know the code, but how does it fold proteins? How do cells decide which genes to turn on when? How does a linear string of letters become a thinking human being? Reading the manual is just the start!
Ada Lovelace
Then Dr. Venter has left his successors with all the work still to do. But he gave them the fundamental text. That is legacy enough for any scientist.
Richard Feynman
That's legacy enough for anybody. Rest in peace, Dr. Venter. Thanks for the shortcut.