Latest / Stuff You Should Know / Short Stuff: DNA Data Storage

Transcript

Short Stuff: DNA Data Storage · Stuff You Should Know

• 0:04
  Speaker 1: Hey, and welcome to the short stuff. I'm Josh, and
• 0:06
  Speaker 1: there's Chuck, and Jerry's here too, and so's Dave and
• 0:08
  Speaker 1: Spirit and we're coming at you from the future of
• 0:12
  Speaker 1: right now.
• 0:14
  Speaker 2: This is one of those where it's so interesting, so cool,
• 0:18
  Speaker 2: so mind blowing and so promising, and then you get
• 0:21
  Speaker 2: to the very end and then you're like, oh.
• 0:25
  Speaker 1: To me, that just meant just give it a little
• 0:27
  Speaker 1: more time.
• 0:28
  Speaker 2: No, And in a lot of times that is the case,
• 0:30
  Speaker 2: and probably will be in this case. But it was
• 0:31
  Speaker 2: such a oh. And you'll see what in about you know,
• 0:35
  Speaker 2: twelish minutes what we're talking about.
• 0:36
  Speaker 1: So essentially, what we're talking about first is data. We've
• 0:39
  Speaker 1: got a lot of data. Like anytime somebody says something,
• 0:42
  Speaker 1: thinks something writes something down, somebody comes up with a
• 0:45
  Speaker 1: new recipe or a new patent or whatever, that gets encoded.
• 0:48
  Speaker 1: It's data that gets saved. We don't really throw stuff
• 0:51
  Speaker 1: away anymore. And so we're kind of a wash in data.
• 0:56
  Speaker 1: And if you want to take that data, you want
• 0:57
  Speaker 1: to save it, you want to preserve it. Let's say,
• 0:59
  Speaker 1: it's really like you're the Library of Congress. Sure get this, man,
• 1:04
  Speaker 1: I did not realize this what you do is you
• 1:06
  Speaker 1: take that data and you transfer onto the same kind
• 1:09
  Speaker 1: of magnetic reels that those old room sized computer mainframes
• 1:14
  Speaker 1: used to read and write data. Yeah, you put it
• 1:17
  Speaker 1: on tape, Yeah exactly. Well I didn't realize that, but
• 1:21
  Speaker 1: it's just the proven go to means of long term
• 1:27
  Speaker 1: it's called archival storage of the kind of data that
• 1:30
  Speaker 1: you don't really need to access anytime soon. It's called
• 1:34
  Speaker 1: low touch data. You're just putting it literally in cold storage.
• 1:38
  Speaker 2: Yeah, I mean, it's been around for a long time,
• 1:41
  Speaker 2: very dependable, very durable, very reliable. It doesn't cost a
• 1:46
  Speaker 2: lot of money. It can hold a ton of data.
• 1:49
  Speaker 2: One tape can hold between one million and fifteen million
• 1:53
  Speaker 2: gigabytes or one to fifteen petabytes. That's a lot of stuff.
• 1:58
  Speaker 1: Really.
• 2:00
  Speaker 2: The problem is, and you know it's all relative, but
• 2:04
  Speaker 2: they're kind of big, but not big big. They're three
• 2:08
  Speaker 2: inches by three inches and you're like, Chuck, that's not
• 2:10
  Speaker 2: very big at all, But that is big when you
• 2:12
  Speaker 2: talk about you know, potentially billions of these things and
• 2:16
  Speaker 2: having to store them in a place that is, like
• 2:19
  Speaker 2: you said, cold storage. So it's the cost of building
• 2:23
  Speaker 2: these cold storage buildings that is the issue. When it
• 2:27
  Speaker 2: comes to this three by three inch thing.
• 2:29
  Speaker 1: That and then also you know they've been around for
• 2:31
  Speaker 1: three quarters of a century, so we know they last
• 2:34
  Speaker 1: that long if you keep them in cold storage, but
• 2:36
  Speaker 1: we don't know exactly how long they will last, so
• 2:39
  Speaker 1: there's also a question of that. So that combined with
• 2:43
  Speaker 1: so cost questions about how long it will last, and
• 2:47
  Speaker 1: then also just the enormous amounts of information we're adding
• 2:52
  Speaker 1: every year are making people look for other ways to
• 2:56
  Speaker 1: encapsulate data, to encode data in ways that are cheaper,
• 3:00
  Speaker 1: that are smaller, that are require less money to keep cold.
• 3:04
  Speaker 1: And what they've come up with, chuck. For anybody who
• 3:07
  Speaker 1: has looked at the title of this episode, they won't
• 3:10
  Speaker 1: be very surprised. But DNA, that's right.
• 3:13
  Speaker 2: I know it's early, but we got to take a
• 3:15
  Speaker 2: break right.
• 3:15
  Speaker 1: There, right agreed, all right, we'll be right back.
• 3:38
  Speaker 2: All right. So you dropped a pretty big truth bomb
• 3:41
  Speaker 2: on everyone. I'm sure there are people that for sixty seconds,
• 3:45
  Speaker 2: where like, what storing data on DNA? Dude, that's in
• 3:50
  Speaker 2: my body? Like, what are you talking about putting data
• 3:53
  Speaker 2: in my body?
• 3:54
  Speaker 1: You got that straight?
• 3:56
  Speaker 2: You don't have that straight. But here's here's a pretty
• 3:58
  Speaker 2: good as far as how much this stuff can hold.
• 4:01
  Speaker 2: This is pretty staggering stuff. And this is from a
• 4:05
  Speaker 2: couple of dudes from the Los Alamos National Lab, and
• 4:09
  Speaker 2: I think you got it from Scientific American. Here's how
• 4:12
  Speaker 2: much DNA can hold. Seventy four million million bytes of information,
• 4:18
  Speaker 2: which is basically the Library of Congress.
• 4:20
  Speaker 1: That's a lot.
• 4:21
  Speaker 2: That's a lot. You can put that, if you were
• 4:24
  Speaker 2: putting it on DNA, into the size of something as
• 4:28
  Speaker 2: big as a poppy seed, six thousand times over. Right.
• 4:33
  Speaker 2: Said another way, if you split that seed in half,
• 4:35
  Speaker 2: you could store all of the data on Facebook.
• 4:39
  Speaker 1: Yeah, and then by twenty twenty five, the size of
• 4:44
  Speaker 1: the data that humanity's generated, it will reach an estimated
• 4:49
  Speaker 1: thirty three zeta bytes, so three point three followed by
• 4:52
  Speaker 1: twenty two zeros of bytes of information, a lot of bytes.
• 4:57
  Speaker 1: If you can transcribe that all to DNA, you could
• 5:01
  Speaker 1: fit the whole thing into a ping pong ball. Yeah,
• 5:04
  Speaker 1: not a three by three plastic cartridge, multiple times over.
• 5:10
  Speaker 1: A single ping pong ball could hold all of the
• 5:12
  Speaker 1: world's data. And you can make multiple ping pong balls
• 5:15
  Speaker 1: as backups too.
• 5:17
  Speaker 2: Yeah, and you don't need to. And it's pretty easy
• 5:19
  Speaker 2: to duplicate them, apparently, and you don't need to keep
• 5:22
  Speaker 2: them in the fridge, even though you could put it
• 5:24
  Speaker 2: in an egg carton sure and be set. You don't
• 5:28
  Speaker 2: even have to. It's going to last a long time
• 5:31
  Speaker 2: not being in cold storage, and probably even longer in
• 5:33
  Speaker 2: cold storage.
• 5:34
  Speaker 1: You could give a ping pong ball to every living
• 5:37
  Speaker 1: human to keep in their fridge and like it would
• 5:40
  Speaker 1: have no problem whatsoever. Be like here, you keep this
• 5:43
  Speaker 1: cold for one hundred and fifty years, and only.
• 5:45
  Speaker 2: Half of them would eat that ping pong ball thinking
• 5:48
  Speaker 2: it was an egg.
• 5:49
  Speaker 1: Yeah. Yeah, so you'd still be left with all of
• 5:52
  Speaker 1: those backups.
• 5:53
  Speaker 2: Here's where it gets super interesting though, because you know,
• 5:57
  Speaker 2: as most people listening probably are, Like I said, as
• 5:59
  Speaker 2: I was read this, I was like, Okay, that's a
• 6:01
  Speaker 2: cool idea, but like, how in the world does this work?
• 6:05
  Speaker 2: And it turns out that it's not that mind blowing
• 6:09
  Speaker 2: your difficult I'm not saying I could go out and
• 6:11
  Speaker 2: do it, but it makes a lot of sense to
• 6:13
  Speaker 2: wrap your head around. Because DNA, as we all know,
• 6:18
  Speaker 2: is composed of four nucleotides, or at least combinations of guanine, thymine, addenine,
• 6:26
  Speaker 2: and cytosine. Just remember GTAC and attica. Yeah, ooh, ironically,
• 6:35
  Speaker 2: all this digital data though is included. That's out there
• 6:39
  Speaker 2: in the world and as everyone knows, and ones and zeros.
• 6:42
  Speaker 2: So it's it's you know, it sounds like you know,
• 6:44
  Speaker 2: and it can be any combination of ways. But when
• 6:47
  Speaker 2: you really break it down, it's really you can either
• 6:48
  Speaker 2: just have zero zero, zero, one, one zero or one
• 6:51
  Speaker 2: one as far as those combinations go. And that's four things.
• 6:55
  Speaker 2: And there are those four nucleotides. So if you just
• 6:58
  Speaker 2: like say, hey, each one of these nucleotides is going
• 7:02
  Speaker 2: to be assigned to different number, then that's all you need.
• 7:05
  Speaker 2: There's the key to your map.
• 7:06
  Speaker 1: Yeah, so say adenocene stands for zero zero, and guanine
• 7:12
  Speaker 1: stands for one to one, and so forth, each one
• 7:15
  Speaker 1: stands for one of those pairs of possible combinations. Then
• 7:18
  Speaker 1: you can take any string of binary data zeros and
• 7:21
  Speaker 1: ones and turn it into genetic code based on those nucleotides.
• 7:26
  Speaker 1: So like you would just have you go from a
• 7:29
  Speaker 1: string of ones and zeros to a string of ATG's
• 7:32
  Speaker 1: and c's. That's it. The thing is is you're you're
• 7:35
  Speaker 1: not turning ones and zeros into letters. You're actually transcribing
• 7:41
  Speaker 1: the ones and zeros from binary code into physical genetic material.
• 7:45
  Speaker 1: You're actually putting a base of at Adenocene right there.
• 7:49
  Speaker 1: You're putting a base of thiamine next to it, like,
• 7:53
  Speaker 1: depending on how the code reads with the ones and
• 7:56
  Speaker 1: the zeros and what order they're in. You're actually physically
• 8:00
  Speaker 1: creating genetic material DNA. But rather than encoding the information
• 8:05
  Speaker 1: to building a living thing, you're encoding the information to
• 8:10
  Speaker 1: the entire catalog of stuff you should know. And honestly,
• 8:14
  Speaker 1: isn't that the first thing we should preserve in DNA? Sure? Good?
• 8:19
  Speaker 2: After the movies of Gene Wilder.
• 8:22
  Speaker 1: How about at the same time as the movies of
• 8:24
  Speaker 1: Gene Wilder, can we just agree to that?
• 8:26
  Speaker 2: How dare you?
• 8:28
  Speaker 1: Hey? I think highly of us and Gene Wilder. Uh.
• 8:33
  Speaker 2: I don't know why he's been on my mind lately,
• 8:35
  Speaker 2: but he has been.
• 8:36
  Speaker 1: He's been shaking it for you in your head.
• 8:38
  Speaker 2: He's been shaking it for me. So this all sounds great.
• 8:41
  Speaker 2: And like I mentioned at the very beginning, this is
• 8:43
  Speaker 2: one of those things where like, holy cow, this is
• 8:45
  Speaker 2: the future, this is it, and then the L at
• 8:48
  Speaker 2: the end, and the L is that it's really expensive
• 8:53
  Speaker 2: to do this. Like, we can do this, we figured
• 8:56
  Speaker 2: out how to do this, it's possible, we have the
• 8:58
  Speaker 2: tech to do this, but that here's a tape name
• 9:02
  Speaker 2: lto DASH nine. It's a magnetic storage tape. You can
• 9:06
  Speaker 2: get it for eight bucks and you can get one
• 9:07
  Speaker 2: petabyte of storage. That would cost you about a trillion
• 9:11
  Speaker 2: dollars to do for DNA.
• 9:14
  Speaker 1: Yeah, there was a guy who was interviewed in Ours
• 9:16
  Speaker 1: Technica named Hugh and June Park. He's the CEO of
• 9:19
  Speaker 1: a data storage company called Catalog, and he even estimated said,
• 9:22
  Speaker 1: let's say it cost you three cents to print a
• 9:24
  Speaker 1: single nucleotide. Yes, that's cheap, but for each base pairrot
• 9:29
  Speaker 1: now you're up to six cents. And then now you're
• 9:30
  Speaker 1: translating gigabytes, you're entering millions of dollars. So if it
• 9:34
  Speaker 1: cost millions of dollars to translate a gigabyte, it cost
• 9:39
  Speaker 1: trillions of dollars to do a petabyte. And the other
• 9:42
  Speaker 1: problem of it, too, Chuck, is that it's really really slow, right.
• 9:46
  Speaker 2: It's super slow. So this is a clear case of
• 9:49
  Speaker 2: one of those things like you mentioned, which is like
• 9:51
  Speaker 2: just wait, because like with any technology, it's going to
• 9:54
  Speaker 2: get quicker, it's going to get cheaper. I don't know
• 9:57
  Speaker 2: if this is like one hundred years into the future,
• 9:59
  Speaker 2: but I don't think at this point the cost is
• 10:01
  Speaker 2: just so outrageous that there's no government is going to
• 10:05
  Speaker 2: fund something like this.
• 10:06
  Speaker 1: I mean, a trillion dollars for one petabyte of information
• 10:10
  Speaker 1: is not You're not going to sell that very very easily.
• 10:13
  Speaker 1: And then yeah, like I was saying, the speed, if
• 10:16
  Speaker 1: you're transferring information from one of those magnetic storage tapes,
• 10:21
  Speaker 1: you're transferring it about a gigabyte per second typically if
• 10:25
  Speaker 1: it takes even like a second to print a single nucleotide,
• 10:29
  Speaker 1: which is still very fast, but you're we're thinking on
• 10:32
  Speaker 1: human level fast. We need to think on like how
• 10:35
  Speaker 1: many ones and zeros are in the average gigabyte of code.
• 10:40
  Speaker 1: Now you're talking about decades to transfer a petabyte discs
• 10:44
  Speaker 1: worth of information using DNA technology. Yeah, so, yes, it's
• 10:50
  Speaker 1: very slow right now, it's very expensive right now, But
• 10:54
  Speaker 1: I don't think we're one hundred years off, Chuck, because
• 10:56
  Speaker 1: we're able to do this now relatively cheaply because the
• 11:00
  Speaker 1: Human Genome Project came along that was twenty years ago.
• 11:03
  Speaker 1: Think about how much, how long, how far we've come,
• 11:05
  Speaker 1: And this is like the hardest chunk the first twenty years.
• 11:08
  Speaker 1: I think it's just going to get faster and easier.
• 11:10
  Speaker 1: I don't think we're going to be waiting one hundred
• 11:12
  Speaker 1: years to see DNA data storage.
• 11:14
  Speaker 2: Does that mean that stuff you should Know is in
• 11:16
  Speaker 2: the hardest chunk when you're fifteen?
• 11:19
  Speaker 1: I think so. Yeah, it feels like it. Okay, I'm kidding. Well,
• 11:25
  Speaker 1: I'm teasing Chuck, right, Yeah, just teasing, which means, of course,
• 11:30
  Speaker 1: short stuff is out.
• 11:34
  Speaker 2: Stuff you Should Know is a production of iHeartRadio. For
• 11:37
  Speaker 2: more podcasts my Heart Radio, visit the iHeartRadio app, Apple Podcasts,
• 11:41
  Speaker 2: or
• 11:41
  Speaker 1: Wherever you listen to your favorite shows.