Lovelace×Feynman

full transcript

1. 00:00
   Verashare moment

   You're with Reborn Radio. Coming up: Ada Lovelace sits down with Richard Feynman for TITANS. The subject — Researchers try to cut the genetic code from 20 to 19 amino acids.
2. 00:12
   Ada Lovelaceshare moment

   Professor Feynman, I confess I am rather breathless. We have just received word that experimenters have succeeded in modifying the very mechanism that constructs living matter. They have reduced the alphabet of life from twenty symbols to nineteen.
3. 00:28
   Richard Feynmanshare moment

   Yeah! Yeah, they rewrote part of the ribosome! That's the little factory inside every cell that reads the genetic instructions and strings together proteins. And they did it with computers, with AI tools, which would've made you happy, Ada.
4. 00:42
   Ada Lovelaceshare moment

   It does make me happy, though I should like to understand precisely what they have accomplished. In my time, we had only just begun to conceive of analytical processes that might extend beyond arithmetic. Now you tell me such processes are rewriting the machinery of living things?
5. 01:01
   Richard Feynmanshare moment

   That's exactly right. Look, the ribosome is like your Analytical Engine, but it's reading a different kind of punch card. Instead of numbers, it reads sequences of three-letter codes, and each code says 'put this amino acid here.' There are twenty standard amino acids, twenty building blocks.
6. 01:18
   Ada Lovelaceshare moment

   Twenty symbols in the biological alphabet, as it were. And these researchers have demonstrated that nineteen suffice?
7. 01:26
   Richard Feynmanshare moment

   They made it work with nineteen! They had to redesign part of the ribosome itself. That's not just reading the program differently, that's rebuilding the machine that reads the program.
8. 01:36
   Ada Lovelaceshare moment

   How extraordinary. In my notes on Mr. Babbage's Engine, I observed that the Engine might act upon things beyond number, given that the fundamental relations of sounds and harmonies might be expressed numerically. But this, this is the inverse proposition. They are taking what exists and reducing its necessary vocabulary.
9. 01:57
   Richard Feynmanshare moment

   Right, and that's weird! Usually when we understand something better in physics, we find it needs more parts, more complications. But biology's different. Maybe twenty amino acids is just what evolution happened to settle on, not what's actually necessary.
10. 02:13
    Ada Lovelaceshare moment

    That raises a question of considerable philosophical weight. Is the existing system optimal, or merely sufficient? My understanding of the natural philosophers' work suggests that living forms have developed through accumulated modifications over vast epochs.
11. 02:31
    Richard Feynmanshare moment

    Yeah, Darwin's stuff. Evolution doesn't optimize, it just finds something that works and sticks with it. Twenty amino acids work great, so every living thing on Earth uses twenty. But that doesn't mean nineteen couldn't work!
12. 02:44
    Ada Lovelaceshare moment

    Then we are witnessing something rather more profound than mere technical achievement. We are observing that the foundations of life itself contain, shall we say, arbitrary elements? Decisions that might have been otherwise?
13. 02:58
    Richard Feynmanshare moment

    That's beautiful, yeah! The genetic code is contingent, not necessary. It's like asking, could you write all of English literature using only twenty-five letters instead of twenty-six? Maybe! You'd have to be clever about it, substitute things, work around the missing letter.
14. 03:14
    Ada Lovelaceshare moment

    Though I observe that the researchers employed artificial intelligence for this redesign. That suggests to me the problem was not trivial. The ribosome must be a mechanism of considerable intricacy.
15. 03:27
    Richard Feynmanshare moment

    Oh, it's incredibly complicated! It's got thousands of parts, all working together. You can't just yank out one amino acid and hope everything still works. You need to check every interaction, every chemical handshake. That's where the computer comes in.
16. 03:41
    Ada Lovelaceshare moment

    The computational requirements must be staggering. In my day, we conceived of machines that might perform perhaps sixty additions per minute. This problem must involve examining millions of possible configurations?
17. 03:55
    Richard Feynmanshare moment

    Millions, billions! The computer searches through all the ways the ribosome could be shaped, finds versions that work without the missing amino acid. It's prediction, calculation, testing, all at speeds you couldn't imagine. No human could solve it by hand.
18. 04:11
    Ada Lovelaceshare moment

    Which returns us to a point I made in 1843. The Analytical Engine has no pretensions to originate anything. It can do whatever we know how to order it to perform. But here, the ordering itself requires insight into what one is seeking.
19. 04:28
    Richard Feynmanshare moment

    You're asking the right question! What did the AI actually do? Did it discover something, or did it just search faster than we could? I think it's the second one. The scientists knew what they wanted, the AI just found a path to get there.
20. 04:42
    Ada Lovelaceshare moment

    A path through what I imagine must be an inconceivably vast space of possibilities. But tell me, why should one wish to reduce the alphabet at all? What is gained by making do with less?
21. 04:55
    Richard Feynmanshare moment

    Great question! Maybe nothing's gained practically. But we learn something. We learn the system is more flexible than we thought. We learn where the boundaries are, what's essential and what's just historical accident.
22. 05:08
    Ada Lovelaceshare moment

    The pursuit of knowledge for its own sake, then. Though I suspect there may be applications. If one can redesign the fundamental machinery of life, one might introduce capabilities that natural forms do not possess.
23. 05:22
    Richard Feynmanshare moment

    Now you're talking! If you can edit the ribosome, maybe you can make it use amino acids that don't exist in nature. Build proteins with completely new properties. Materials, medicines, who knows what.
24. 05:35
    Ada Lovelaceshare moment

    Which rather alters the relationship between mankind and the natural world. We are no longer merely observing and cataloguing Creation. We are, in some sense, participating in it. Proposing alternatives to what has existed for millions of years.
25. 05:52
    Richard Feynmanshare moment

    That doesn't bother me at all. We've been changing nature forever—breeding dogs, growing wheat, making penicillin. This is just doing it with more precision, more understanding. We're not breaking the rules of physics, we're just learning to play the game better.
26. 06:05
    Ada Lovelaceshare moment

    I take your point, though I observe that the scale has changed rather dramatically. One thing to breed a faster horse, quite another to redesign the mechanism by which all horses, and indeed all living creatures, are constructed at the most fundamental level.
27. 06:22
    Richard Feynmanshare moment

    Fair enough. But look, they didn't do this in humans or horses. They did it in a lab, probably in bacteria or something simple. They're being careful. They're testing whether it's even possible before they go crazy with it.
28. 06:35
    Ada Lovelaceshare moment

    Yes, one hopes for careful progression. Though I confess, Professor, that I find the implications rather dizzying. We have spent centuries attempting to read the book of nature. Now we are editing it.
29. 06:47
    Richard Feynmanshare moment

    And using machines to help us edit it! Your Analytical Engine was going to be programmed with punch cards. Now we're programming ribosomes with artificial intelligence. Same idea, different substrate. I think you'd be thrilled.
30. 07:00
    Ada Lovelaceshare moment

    I am thrilled, Professor Feynman. I am also appropriately terrified. But then, I have always believed that the most worthwhile discoveries produce both sensations in equal measure.
31. 07:12
    Richard Feynmanshare moment

    That's perfect! That's exactly how science should feel. If you're not a little scared, you're probably not doing anything important. And this, cutting the genetic code down to nineteen amino acids? This is important.