1
00:00:15,316 --> 00:00:24,516
Speaker 1: Pushkin. Every cell in your body is an incredible data

2
00:00:24,636 --> 00:00:29,276
Speaker 1: storage device. Inside the nucleus of almost every cell is

3
00:00:29,316 --> 00:00:34,636
Speaker 1: your entire genetic code, all of your DNA arranged just so,

4
00:00:34,956 --> 00:00:41,476
Speaker 1: providing the complete blueprint of you in every cell. Unbelievable.

5
00:00:41,916 --> 00:00:45,476
Speaker 1: A few decades ago, some scientists considered this fact and thought,

6
00:00:46,196 --> 00:00:49,636
Speaker 1: DNA has stored the data of pretty much every living

7
00:00:49,676 --> 00:00:54,356
Speaker 1: thing on Earth for billions of years. It's incredibly efficient, reliable.

8
00:00:54,556 --> 00:00:57,436
Speaker 1: We know it works. What if we could use it

9
00:00:57,516 --> 00:01:01,196
Speaker 1: to store other kinds of data for certain things? It

10
00:01:01,236 --> 00:01:04,876
Speaker 1: could be profoundly better than the hard drives and tapes

11
00:01:04,916 --> 00:01:08,996
Speaker 1: were currently using for data storage. And now they've figured

12
00:01:08,996 --> 00:01:15,996
Speaker 1: out how to do it. I'm Jacob Goldstein and this

13
00:01:16,116 --> 00:01:19,596
Speaker 1: is What's Your Problem, the show where engineers and entrepreneurs

14
00:01:19,756 --> 00:01:21,996
Speaker 1: talk about how they're going to change the world once

15
00:01:22,036 --> 00:01:25,796
Speaker 1: they solve a few problems. My guest today is Emily Laprust.

16
00:01:26,236 --> 00:01:30,196
Speaker 1: She's the co founder and CEO of Twist Bioscience, one

17
00:01:30,236 --> 00:01:33,916
Speaker 1: of the leading companies working on storing data in DNA.

18
00:01:34,276 --> 00:01:38,876
Speaker 1: Sounds like science fiction, it's just science. Emily's problem, how

19
00:01:38,916 --> 00:01:42,196
Speaker 1: do you make storing data in DNA as cheap as

20
00:01:42,236 --> 00:01:47,756
Speaker 1: storing it on a hard drug. Emily's company, Twist, is

21
00:01:47,796 --> 00:01:51,796
Speaker 1: in the DNA synthesis business. They make and sell custom

22
00:01:51,916 --> 00:01:56,596
Speaker 1: DNA to researchers and healthcare companies, and Twist is also

23
00:01:56,716 --> 00:01:59,876
Speaker 1: part of a group of companies, including Microsoft and the

24
00:02:00,036 --> 00:02:03,356
Speaker 1: data storage company Western Digital, that are trying to bring

25
00:02:03,716 --> 00:02:06,636
Speaker 1: DNA storage to market. It's clear that it can work

26
00:02:06,676 --> 00:02:08,796
Speaker 1: in the lab. The trick now is to make it

27
00:02:08,836 --> 00:02:11,676
Speaker 1: work in the real world and have the economics makes sense.

28
00:02:12,476 --> 00:02:15,516
Speaker 1: We started our conversation by talking about a key problem

29
00:02:15,796 --> 00:02:20,036
Speaker 1: with how data storage works today. Hard drives and flash

30
00:02:20,116 --> 00:02:22,796
Speaker 1: drives and even old fashioned tape which is still used,

31
00:02:23,156 --> 00:02:28,036
Speaker 1: are all based on magnetization, and magnetization just does not

32
00:02:28,196 --> 00:02:32,596
Speaker 1: work for long term storage. It's unreliable, it degrades, and

33
00:02:32,676 --> 00:02:35,156
Speaker 1: so if you want to achive data for a very

34
00:02:35,156 --> 00:02:37,516
Speaker 1: long time, what you have to do is every five

35
00:02:37,556 --> 00:02:39,956
Speaker 1: to seven years, you have to take the data from

36
00:02:39,996 --> 00:02:42,596
Speaker 1: one hob drive to the next, or from one day

37
00:02:42,636 --> 00:02:44,436
Speaker 1: to the next, or from one flash to the next

38
00:02:44,556 --> 00:02:47,796
Speaker 1: every five years, because you just can't trust that the

39
00:02:47,876 --> 00:02:51,196
Speaker 1: data will stay there more than five years. And so

40
00:02:51,636 --> 00:02:55,476
Speaker 1: the system we have now, obviously it's great, right, like

41
00:02:55,636 --> 00:02:58,516
Speaker 1: data storage is this amazing miracle of the modern world.

42
00:02:58,556 --> 00:03:02,956
Speaker 1: But you're saying a flaw with it is it doesn't

43
00:03:03,076 --> 00:03:08,036
Speaker 1: work forever. In fact, it becomes unreliable after five years on.

44
00:03:09,116 --> 00:03:11,716
Speaker 1: One of the big effort that's going on there is

45
00:03:12,076 --> 00:03:17,076
Speaker 1: exactly that it's constantly replacing hard drives that are broken

46
00:03:17,156 --> 00:03:20,596
Speaker 1: with new ones, and so the migration, the maintenance, and

47
00:03:20,676 --> 00:03:25,156
Speaker 1: the energy that you need to do that is a

48
00:03:25,196 --> 00:03:29,556
Speaker 1: big issue for archiving, and archiving is more than sixty

49
00:03:29,556 --> 00:03:31,956
Speaker 1: percent of the market of data. And so that's where

50
00:03:32,076 --> 00:03:36,436
Speaker 1: DNA and totally changings is you could archive data for

51
00:03:36,436 --> 00:03:40,116
Speaker 1: a long time with no migration, no maintenance, no energy,

52
00:03:40,236 --> 00:03:43,436
Speaker 1: and that could be a game changer. So the key

53
00:03:43,636 --> 00:03:46,276
Speaker 1: use case is like, it's not my phone, it's not

54
00:03:46,356 --> 00:03:48,916
Speaker 1: my laptop, it's not anything I'm sort of personally doing.

55
00:03:48,996 --> 00:03:52,556
Speaker 1: It's maybe if I'm saving something to the cloud, you know,

56
00:03:52,676 --> 00:03:57,636
Speaker 1: my whatever, kids, baby pictures, and presumably more significantly big

57
00:03:57,716 --> 00:04:01,596
Speaker 1: institutional sort of corporate or governmental kind of files. I mean,

58
00:04:01,636 --> 00:04:04,876
Speaker 1: it's that the real opportunity that you're talking about here. Yeah, exactly,

59
00:04:04,996 --> 00:04:07,236
Speaker 1: DNA will never be on on your computer. It's not

60
00:04:07,476 --> 00:04:10,956
Speaker 1: for what's called data. So the data will get in

61
00:04:10,956 --> 00:04:15,556
Speaker 1: and out all the time. It's it's more fall cool data.

62
00:04:15,716 --> 00:04:18,276
Speaker 1: That's the data that you read that infrequently, but that

63
00:04:18,476 --> 00:04:22,436
Speaker 1: is the majority of the data out there that DNA

64
00:04:22,516 --> 00:04:25,516
Speaker 1: can solve. If you're a bank, if you are an

65
00:04:25,716 --> 00:04:27,916
Speaker 1: insurance company, you have to store the data for a

66
00:04:27,956 --> 00:04:31,156
Speaker 1: very long time. If you're a government, you want you

67
00:04:31,196 --> 00:04:33,076
Speaker 1: need to store that that for a long time. So

68
00:04:33,156 --> 00:04:35,236
Speaker 1: the longer you're going to need to keep the data,

69
00:04:35,396 --> 00:04:41,316
Speaker 1: the more it makes sense to store it in DNA exactly.

70
00:04:41,516 --> 00:04:44,436
Speaker 1: And that's why the first project we're going to launch

71
00:04:44,516 --> 00:04:47,716
Speaker 1: is we call it a century Archive, so it's basically

72
00:04:47,876 --> 00:04:51,236
Speaker 1: you purchase one hundred years of storage. I like that.

73
00:04:51,236 --> 00:04:54,396
Speaker 1: That's good branding. Did a marketing person come up with that?

74
00:04:54,436 --> 00:04:56,396
Speaker 1: Did you come up with that? I did not come

75
00:04:56,476 --> 00:05:00,356
Speaker 1: up with that, but yes, definitely a very clever marketing

76
00:05:00,356 --> 00:05:03,476
Speaker 1: person came up with that. And then after that we'll

77
00:05:03,876 --> 00:05:08,036
Speaker 1: sell the Millennium Archive, the Millennium, the Millennium Archive. But well,

78
00:05:08,116 --> 00:05:10,116
Speaker 1: let's get to one hundred first. We can I can

79
00:05:10,196 --> 00:05:12,316
Speaker 1: have you on a year. We can talk about the

80
00:05:12,396 --> 00:05:15,916
Speaker 1: Millennium Archive, but so are you right now selling a

81
00:05:15,956 --> 00:05:19,836
Speaker 1: product where if a company wants to save something for

82
00:05:19,876 --> 00:05:21,596
Speaker 1: a hundred years, they can give you money and you

83
00:05:21,636 --> 00:05:24,476
Speaker 1: will store their data in DNA. Does that exist? Yeah,

84
00:05:24,556 --> 00:05:26,316
Speaker 1: we've done it before. Right now, we are in the

85
00:05:26,476 --> 00:05:31,436
Speaker 1: early access, so it's not fully broadly available. But someone

86
00:05:31,516 --> 00:05:34,516
Speaker 1: comes today and they have money, will definitely do it

87
00:05:34,556 --> 00:05:38,196
Speaker 1: for them. The one thing though, today it's still is

88
00:05:38,236 --> 00:05:43,316
Speaker 1: still quite expensive, and we're working to make it affordable.

89
00:05:43,476 --> 00:05:45,196
Speaker 1: It will couse a little bit more up front, not

90
00:05:45,236 --> 00:05:47,876
Speaker 1: too much because people don't really want to be a

91
00:05:47,876 --> 00:05:51,756
Speaker 1: premium and this market is the elastic so it's kind

92
00:05:51,756 --> 00:05:56,476
Speaker 1: of ironic. But the lower the price of what we sell,

93
00:05:56,676 --> 00:06:00,036
Speaker 1: the more we're gonna sell. We're gonna sell it. Sure. No,

94
00:06:00,196 --> 00:06:06,476
Speaker 1: that is a very intuitive relationship and a classic issue

95
00:06:06,476 --> 00:06:09,036
Speaker 1: in technology. Right you have this new thing and it's

96
00:06:09,236 --> 00:06:12,436
Speaker 1: really expensive, so nobody's buying it, and the way it

97
00:06:12,476 --> 00:06:14,836
Speaker 1: gets cheaper is lots of people buy it and then

98
00:06:14,876 --> 00:06:17,676
Speaker 1: you figure out the economy of scale and it gets cheaper.

99
00:06:17,676 --> 00:06:19,756
Speaker 1: I mean, it's that the dynamic you're talking about here,

100
00:06:20,036 --> 00:06:22,916
Speaker 1: that's exactly the dynamic. And DNA is amazing for that

101
00:06:23,316 --> 00:06:28,076
Speaker 1: um because DNA is extremely tiny. Yeah, it's extremely dense.

102
00:06:28,436 --> 00:06:31,916
Speaker 1: You could take all the data in the Internet and

103
00:06:32,596 --> 00:06:36,196
Speaker 1: which is youtue, I mean everything, it's the whole world,

104
00:06:37,036 --> 00:06:39,276
Speaker 1: and it will fit in a shoe box if you

105
00:06:39,316 --> 00:06:44,836
Speaker 1: put it in DNA. So I want to understand a

106
00:06:44,876 --> 00:06:47,916
Speaker 1: little bit what that means, because it's so not intuitive.

107
00:06:48,076 --> 00:06:51,316
Speaker 1: So all the data on the Internet means like every

108
00:06:52,156 --> 00:06:55,636
Speaker 1: video that's on YouTube, everything everybody has ever posted to

109
00:06:55,636 --> 00:06:59,716
Speaker 1: the internet, every tweet, every Facebook post, everything, everything, So

110
00:06:59,876 --> 00:07:03,756
Speaker 1: that all exists now on hard drives, right, and giant

111
00:07:03,836 --> 00:07:07,716
Speaker 1: rooms full of computers all over the world, cooled by fans. Right,

112
00:07:07,796 --> 00:07:10,476
Speaker 1: Like all the data on the Internet exists now sort

113
00:07:10,476 --> 00:07:15,276
Speaker 1: of distributed across many, many warehouses. Like just do you

114
00:07:15,396 --> 00:07:17,196
Speaker 1: do you know how much space it takes up? Now?

115
00:07:17,356 --> 00:07:19,356
Speaker 1: I don't know. I don't have any idea. I don't know,

116
00:07:19,436 --> 00:07:23,156
Speaker 1: but it's it's it's millions and minus of square foot.

117
00:07:23,156 --> 00:07:27,676
Speaker 1: What I do know is that one Facebook data center

118
00:07:29,156 --> 00:07:33,836
Speaker 1: in Texas uses two percent of the Texas electricity. When

119
00:07:33,836 --> 00:07:36,996
Speaker 1: we talk about storing data in DNA, can you walk

120
00:07:37,036 --> 00:07:40,356
Speaker 1: me through, actually what happens. What happens somebody wants to

121
00:07:40,396 --> 00:07:43,916
Speaker 1: save whatever, a movie, a Netflix movie, and instead of

122
00:07:43,956 --> 00:07:46,756
Speaker 1: saving it on a little chip, they save it in DNA.

123
00:07:46,996 --> 00:07:51,476
Speaker 1: So storage of data is a bunch of zeros and ones. Right,

124
00:07:51,556 --> 00:07:56,236
Speaker 1: It's the binary code DNA four letters ACGT, And just

125
00:07:56,276 --> 00:07:58,356
Speaker 1: to be super clear, when you say four letters, I

126
00:07:58,356 --> 00:08:01,676
Speaker 1: mean all. All DNA is made up of four basically

127
00:08:01,836 --> 00:08:05,236
Speaker 1: types of molecules, and only four. That's right. If you

128
00:08:05,396 --> 00:08:09,876
Speaker 1: look at any DNA of any leaving organism from the

129
00:08:09,876 --> 00:08:15,556
Speaker 1: biggest to the smoot one, that DNA is made of foditos, acg, NT,

130
00:08:16,276 --> 00:08:21,836
Speaker 1: and the function of that organism is based on in

131
00:08:21,836 --> 00:08:27,396
Speaker 1: which orders the acgs and ts linked together. Emily says,

132
00:08:27,476 --> 00:08:30,196
Speaker 1: to use DNA to store a movie, or to store

133
00:08:30,236 --> 00:08:33,036
Speaker 1: any kind of data, you let each of the four

134
00:08:33,156 --> 00:08:37,596
Speaker 1: letters represent a binary pair. So for example, you could

135
00:08:37,596 --> 00:08:41,276
Speaker 1: say zero zero is A and zero one is C,

136
00:08:41,676 --> 00:08:45,356
Speaker 1: one one is T, and one zero is G. Once

137
00:08:45,396 --> 00:08:48,356
Speaker 1: you've done that, you can take any digital file. You

138
00:08:48,356 --> 00:08:51,796
Speaker 1: can take that movie and on a computer convert the

139
00:08:51,916 --> 00:08:54,996
Speaker 1: zeros and ones that represent the movie into a string

140
00:08:55,036 --> 00:08:59,196
Speaker 1: of the letters ACTG. That string of letters is your blueprint.

141
00:08:59,716 --> 00:09:03,356
Speaker 1: And then the next step is kind of amazing because

142
00:09:03,436 --> 00:09:06,276
Speaker 1: we are so used to everything happening, you know, digitally

143
00:09:06,316 --> 00:09:09,956
Speaker 1: on the computer. The next step is a machine about

144
00:09:09,996 --> 00:09:13,956
Speaker 1: the size of an SUV actually starts printing strands of

145
00:09:14,036 --> 00:09:17,716
Speaker 1: DNA based on that digital blueprint. What we're going to

146
00:09:17,916 --> 00:09:21,916
Speaker 1: actually physically synthesize, We're going to make it from scratch,

147
00:09:22,596 --> 00:09:25,716
Speaker 1: is a bunch of little pieces of DNA that has

148
00:09:25,756 --> 00:09:28,476
Speaker 1: two to three hundred litters. Okay, so is that the

149
00:09:28,516 --> 00:09:30,356
Speaker 1: next step. It's still on the computer, but now the

150
00:09:30,396 --> 00:09:37,036
Speaker 1: computer tells some DNA synthesis machine what molecules to synthesize.

151
00:09:36,916 --> 00:09:39,116
Speaker 1: Where the a's go, where the teas go, where the

152
00:09:39,116 --> 00:09:41,796
Speaker 1: ceas go, where the gee's go. Yeah, so exactly, So

153
00:09:41,876 --> 00:09:46,796
Speaker 1: a twist, we build a DNA three D printer basically,

154
00:09:47,196 --> 00:09:49,716
Speaker 1: ok and we have a we have a piece of silicon.

155
00:09:50,316 --> 00:09:52,836
Speaker 1: It's a citicin chip, and so on. Our citic and chip.

156
00:09:53,276 --> 00:09:58,636
Speaker 1: The first generation that we made at one million pieces

157
00:09:58,676 --> 00:10:02,076
Speaker 1: of DNA. The current generation we're working on as two

158
00:10:02,236 --> 00:10:06,036
Speaker 1: hundred fifty six million. Okay, And so we have a

159
00:10:06,036 --> 00:10:08,116
Speaker 1: we have a cilican chip, and then we have a printer.

160
00:10:08,356 --> 00:10:10,436
Speaker 1: So I want to keep going with our sort of

161
00:10:10,476 --> 00:10:13,996
Speaker 1: how does it work narrative here. So we got the

162
00:10:14,076 --> 00:10:16,116
Speaker 1: movie that was already ones and zeros. The ones and

163
00:10:16,196 --> 00:10:23,956
Speaker 1: zeros became ATCNG, those got split up into little packets. Yeah,

164
00:10:24,036 --> 00:10:28,916
Speaker 1: the computer told the three D printer basically what actual

165
00:10:28,956 --> 00:10:33,516
Speaker 1: ATCNG to print physically the actual chemicals on a silicon chip,

166
00:10:33,636 --> 00:10:37,716
Speaker 1: hundreds of millions of them, and then we got delivered

167
00:10:37,796 --> 00:10:40,556
Speaker 1: that ACGT at the right location at the right time.

168
00:10:40,636 --> 00:10:42,796
Speaker 1: And so we start from the citic ins ship. There

169
00:10:42,836 --> 00:10:44,836
Speaker 1: is no DNA on it. And then we come in

170
00:10:44,876 --> 00:10:46,836
Speaker 1: and we print the first layer. So we put the

171
00:10:46,876 --> 00:10:49,396
Speaker 1: first ACGT on the surface, and then we come in

172
00:10:49,516 --> 00:10:53,756
Speaker 1: again to put the second layer of SEGT again the third.

173
00:10:53,836 --> 00:10:56,476
Speaker 1: We do that two to three hundred times, and so

174
00:10:56,516 --> 00:11:01,876
Speaker 1: we are building DNA up to three hundred letters and

175
00:11:01,956 --> 00:11:05,956
Speaker 1: we do that millions of time on the Citicin chip. Okay,

176
00:11:06,036 --> 00:11:10,036
Speaker 1: so now you've got a silicon chip with DNA on it.

177
00:11:10,676 --> 00:11:14,236
Speaker 1: What happens next? Yeah, So at that point the density

178
00:11:14,356 --> 00:11:16,476
Speaker 1: is not that great. It's kind of like a hard drive.

179
00:11:16,876 --> 00:11:19,876
Speaker 1: But then the beauties. You can remove the DNA from

180
00:11:19,916 --> 00:11:22,276
Speaker 1: the citikinship. I want to keep track of that DNA.

181
00:11:22,356 --> 00:11:25,316
Speaker 1: You just synthesize. Where does it go? You just dry

182
00:11:25,356 --> 00:11:28,596
Speaker 1: it down and it's so tiny what you've made. It's

183
00:11:28,636 --> 00:11:33,396
Speaker 1: a speck of dust. You can't see it. So now

184
00:11:33,436 --> 00:11:37,236
Speaker 1: it goes into a tiny vessel. But it's called the

185
00:11:37,356 --> 00:11:42,036
Speaker 1: DNA shell. Yeah, it's a good name. It's not ours.

186
00:11:42,076 --> 00:11:46,276
Speaker 1: We buy it, and it's a piece of stainless steel

187
00:11:46,796 --> 00:11:50,516
Speaker 1: and there's a glass cutting inside. You put your the

188
00:11:50,676 --> 00:11:53,796
Speaker 1: DNA in it, you dry down the water, you sell it,

189
00:11:54,316 --> 00:11:57,356
Speaker 1: and you put ilium, so there's no oxygen, no water.

190
00:11:58,116 --> 00:12:02,196
Speaker 1: And in that DNAs shell you can put hundreds of

191
00:12:02,316 --> 00:12:06,956
Speaker 1: Google Data Center in that DNA shell, like every movie

192
00:12:07,076 --> 00:12:10,076
Speaker 1: on Netflix. You could pack in a out of DNA

193
00:12:10,196 --> 00:12:12,516
Speaker 1: in it. And and then you see it and that

194
00:12:12,636 --> 00:12:16,516
Speaker 1: is stable for thousands of years. And how big is it?

195
00:12:16,516 --> 00:12:19,956
Speaker 1: It's uh so, you know, I live in Montana. Everybody

196
00:12:19,956 --> 00:12:21,996
Speaker 1: has guns here, so it's the size of a small

197
00:12:22,076 --> 00:12:29,036
Speaker 1: caliber bullet. That's nice. Okay, So about like I live

198
00:12:29,076 --> 00:12:32,116
Speaker 1: in Brooklyn where fewer people have guns at least as

199
00:12:32,116 --> 00:12:33,716
Speaker 1: far as I know. So would it be about like

200
00:12:33,756 --> 00:12:37,236
Speaker 1: the size of like a like a bean. It would

201
00:12:37,276 --> 00:12:40,636
Speaker 1: be the size of two black beans. Okay. You can

202
00:12:40,716 --> 00:12:44,236
Speaker 1: put every movie on Netflix in there, that's right, and

203
00:12:44,316 --> 00:12:47,956
Speaker 1: then you can store it your room. Tompret show um

204
00:12:47,996 --> 00:12:52,516
Speaker 1: in the sun. You know, it's it's extremely stable for forever.

205
00:12:52,796 --> 00:12:55,116
Speaker 1: And then when you want the data bag. So now

206
00:12:55,116 --> 00:12:57,516
Speaker 1: it's fifty years later and I want to watch that

207
00:12:57,596 --> 00:13:00,596
Speaker 1: movie again. Because I forgot that it wasn't very good.

208
00:13:00,676 --> 00:13:03,236
Speaker 1: Well oh you oh, let me say, jeez, you know

209
00:13:03,316 --> 00:13:07,716
Speaker 1: there's something happened to to the Netflix that are center,

210
00:13:07,876 --> 00:13:11,196
Speaker 1: and you know that movie is gone. We need to

211
00:13:11,196 --> 00:13:13,956
Speaker 1: get it back. That's the real practical use for this.

212
00:13:13,996 --> 00:13:16,356
Speaker 1: I would imagine it's that like, at least for now,

213
00:13:16,356 --> 00:13:18,916
Speaker 1: it's like a backup. It's like a super safe backup,

214
00:13:18,996 --> 00:13:22,156
Speaker 1: even if all of the data centers get hacked or

215
00:13:22,156 --> 00:13:25,276
Speaker 1: blown up or whatever. You have this volt Yeah, yeah,

216
00:13:25,316 --> 00:13:27,676
Speaker 1: it's a very it's it's the vault exactly so exactly

217
00:13:27,716 --> 00:13:30,796
Speaker 1: that that's the ultimate vault. And the beauty is DNA.

218
00:13:30,876 --> 00:13:34,076
Speaker 1: You can make a copy very easily. Let's go back.

219
00:13:34,076 --> 00:13:38,276
Speaker 1: So we have our all of Netflix in a little

220
00:13:38,356 --> 00:13:44,116
Speaker 1: cylinder size of two beans. Well, now we want to

221
00:13:44,156 --> 00:13:46,356
Speaker 1: get the data. How do we get the data off there?

222
00:13:46,556 --> 00:13:50,116
Speaker 1: So you you crack open the shell and you put

223
00:13:50,156 --> 00:13:53,156
Speaker 1: it in what's called a sequencer. So it's a machine

224
00:13:53,196 --> 00:13:56,436
Speaker 1: that reads DNA. And the sequencer is not fast. It's

225
00:13:56,516 --> 00:14:00,556
Speaker 1: twelve hours. So twelve hours later you'll get your data.

226
00:14:00,756 --> 00:14:03,836
Speaker 1: And now it's still ACGT in the computer and you

227
00:14:03,876 --> 00:14:07,916
Speaker 1: have to do the reverse decoding. You have to turn

228
00:14:07,996 --> 00:14:10,596
Speaker 1: the ACGT in two zeros on one and that's the

229
00:14:10,716 --> 00:14:12,316
Speaker 1: end of the story, right, and then you can watch

230
00:14:12,396 --> 00:14:19,436
Speaker 1: the movie that you encoded today exactly exactly. Sounds amazing,

231
00:14:19,956 --> 00:14:22,636
Speaker 1: DNA all of Netflix and a cylinder the size of

232
00:14:22,636 --> 00:14:26,956
Speaker 1: two black beans. But DNA storage is not yet a

233
00:14:27,036 --> 00:14:30,036
Speaker 1: thing that is really happening in the world at scale.

234
00:14:30,756 --> 00:14:33,316
Speaker 1: In a minute, the problem Emily has to solve for

235
00:14:33,356 --> 00:14:36,236
Speaker 1: that to happen, for DNA storage to become a real

236
00:14:36,316 --> 00:14:45,956
Speaker 1: thing in the world. Now, let's get back to the show.

237
00:14:46,996 --> 00:14:50,796
Speaker 1: So it sounds amazing. I'm sold the Internet in a shoebox.

238
00:14:51,916 --> 00:14:56,676
Speaker 1: It sounds incredible. Why Why isn't it really a thing yet?

239
00:14:56,916 --> 00:14:59,636
Speaker 1: Why does Why why aren't people doing it? Yeah, it

240
00:14:59,676 --> 00:15:04,516
Speaker 1: goes back to economics. You can do it today, but

241
00:15:04,596 --> 00:15:08,556
Speaker 1: today it's too expensive. And it's too expensive because today

242
00:15:08,596 --> 00:15:13,316
Speaker 1: we only pack a million pieces of adigos on the

243
00:15:13,356 --> 00:15:16,836
Speaker 1: civic and chip. It works. We've done dozens of administrations

244
00:15:16,836 --> 00:15:21,276
Speaker 1: with again Netflix, Microsoft, University of Washington, with the Montro

245
00:15:21,356 --> 00:15:26,956
Speaker 1: Jazz Festival, with the United Nations, with the Olympic Games. Right,

246
00:15:26,996 --> 00:15:30,996
Speaker 1: so it works, but it is expensive, and that's why

247
00:15:31,516 --> 00:15:34,636
Speaker 1: at Weiss we are pushing the technology further we're packing

248
00:15:34,676 --> 00:15:38,076
Speaker 1: more and more pieces of DNA on the same citkinship. Again,

249
00:15:38,676 --> 00:15:40,876
Speaker 1: the state of the art used to be one hundred

250
00:15:40,956 --> 00:15:45,196
Speaker 1: pieces of DNA. The first Twist product was a million

251
00:15:45,236 --> 00:15:47,796
Speaker 1: pieces of DNA. Right now we're working working on a

252
00:15:47,876 --> 00:15:50,236
Speaker 1: chip that we have two hundred and fifty six millions.

253
00:15:50,316 --> 00:15:53,516
Speaker 1: So okay, so you're packing more and more a DNA

254
00:15:53,916 --> 00:15:56,676
Speaker 1: on a chip to make DNA storage cheaper. Right, that's

255
00:15:56,716 --> 00:16:00,396
Speaker 1: the fundamental problem. You have to solve. How much cheaper

256
00:16:00,796 --> 00:16:02,996
Speaker 1: have you made it so far? And how much cheaper

257
00:16:03,036 --> 00:16:05,076
Speaker 1: does it have to get. We've already done a thousand

258
00:16:05,116 --> 00:16:11,076
Speaker 1: times less and were we're in field of and those

259
00:16:11,156 --> 00:16:15,516
Speaker 1: a thousand times cost reduction and have to add by

260
00:16:15,636 --> 00:16:19,596
Speaker 1: packing more sequences. You do that by packing more pieces

261
00:16:19,636 --> 00:16:24,436
Speaker 1: of DNA on the city Kin chip. And so it's weird.

262
00:16:24,556 --> 00:16:27,036
Speaker 1: I mean, we're used to thinking about this kind of thing.

263
00:16:27,196 --> 00:16:29,876
Speaker 1: I think because of Moore's law, right, because this thing

264
00:16:29,916 --> 00:16:34,516
Speaker 1: has happened with computers for whatever it is, fifty years now.

265
00:16:34,876 --> 00:16:37,796
Speaker 1: That exactly what you're talking about in DNA, right, every

266
00:16:37,836 --> 00:16:40,116
Speaker 1: two years you get twice as many transistors on a

267
00:16:40,196 --> 00:16:43,916
Speaker 1: chip or whatever. But it's not really a law. It

268
00:16:43,916 --> 00:16:47,476
Speaker 1: didn't happen naturally. Right, It happened because like many, many, many,

269
00:16:47,556 --> 00:16:50,636
Speaker 1: really really really smart people kept coming up with clever

270
00:16:51,236 --> 00:16:54,196
Speaker 1: ways of getting more transistors on a chip. Right, it

271
00:16:54,316 --> 00:16:58,316
Speaker 1: doesn't have to keep getting better, getting more efficient. So

272
00:16:58,716 --> 00:17:00,756
Speaker 1: how do you do that? Like, is there some way

273
00:17:00,756 --> 00:17:03,636
Speaker 1: without getting too technical, that we can talk about some

274
00:17:03,716 --> 00:17:07,156
Speaker 1: of the problems you have to solve to keep doubling

275
00:17:07,196 --> 00:17:10,036
Speaker 1: to get a thousand times cheaper. Yeah, it's it's a

276
00:17:10,156 --> 00:17:14,076
Speaker 1: very it's a very difficult engineering problem. And I love

277
00:17:14,116 --> 00:17:17,156
Speaker 1: that it's difficult because if it was easy, every idiots

278
00:17:17,196 --> 00:17:19,796
Speaker 1: would be doing it. And so it is really really hard,

279
00:17:19,796 --> 00:17:22,156
Speaker 1: and you have to be very good at sumer conductor,

280
00:17:22,236 --> 00:17:25,516
Speaker 1: at silicon engineering, you have to be extremely good at

281
00:17:25,716 --> 00:17:30,876
Speaker 1: electrical engineering, at mechanical engineering, and chemical engineering at computation.

282
00:17:31,996 --> 00:17:36,716
Speaker 1: It just a What you have to do is build

283
00:17:36,796 --> 00:17:41,556
Speaker 1: a team where everybody is the best at what they

284
00:17:41,596 --> 00:17:44,716
Speaker 1: do in their own field and have those people be

285
00:17:45,116 --> 00:17:49,436
Speaker 1: extremely great team players, and then you have to demand performance.

286
00:17:49,436 --> 00:17:52,676
Speaker 1: So in some ways you have to run this lack

287
00:17:52,716 --> 00:17:55,476
Speaker 1: of sports team. I always say at Twist, we're not

288
00:17:55,516 --> 00:17:58,236
Speaker 1: a family. Right, in the family, you tolerate bad behavior.

289
00:17:58,676 --> 00:18:02,156
Speaker 1: We out sports team. We hire the best athlete for

290
00:18:02,196 --> 00:18:06,996
Speaker 1: each position to demand performance, and you demand that they

291
00:18:07,076 --> 00:18:10,236
Speaker 1: work as a team and if you give them an

292
00:18:10,316 --> 00:18:15,036
Speaker 1: amazing tough problem, they'll find a way. So you have

293
00:18:15,076 --> 00:18:18,516
Speaker 1: this team, how do they actually solve the problem of

294
00:18:19,036 --> 00:18:23,396
Speaker 1: you know, building denser chips to make DNA storage cheaper. Yeah,

295
00:18:23,436 --> 00:18:25,556
Speaker 1: so first you have to design the Citicans ship where

296
00:18:25,596 --> 00:18:28,516
Speaker 1: it's it's a design a computer that cost millions of

297
00:18:28,556 --> 00:18:31,756
Speaker 1: dollars just to do design right, And then you press

298
00:18:31,796 --> 00:18:34,756
Speaker 1: the button and you say, get me that chipmate. That's

299
00:18:34,796 --> 00:18:37,996
Speaker 1: another set of millions of dollars. And then you get

300
00:18:37,996 --> 00:18:41,676
Speaker 1: the chip in and you literally hook it up to

301
00:18:42,036 --> 00:18:45,436
Speaker 1: a giant amount of set of flectionics and you literally

302
00:18:45,516 --> 00:18:49,156
Speaker 1: boot the chip like you boot your computer, and you

303
00:18:49,196 --> 00:18:52,796
Speaker 1: know the chip doesn't work. It's like, oh, now I

304
00:18:52,796 --> 00:18:55,476
Speaker 1: have to go back either redesign it, which you know

305
00:18:55,516 --> 00:18:58,316
Speaker 1: it's money and time that you want to do. Oh,

306
00:18:58,396 --> 00:19:02,116
Speaker 1: you have to debug it. See what, See what happens.

307
00:19:02,156 --> 00:19:05,796
Speaker 1: And when you finally get the chip to work, that's

308
00:19:05,836 --> 00:19:07,956
Speaker 1: just a Citicans chip. Now you have to bring the

309
00:19:08,036 --> 00:19:11,076
Speaker 1: ACGT on it. Okay, yeah, you have to. You have

310
00:19:11,116 --> 00:19:14,236
Speaker 1: to build the DNA and that's very hard as well.

311
00:19:15,196 --> 00:19:18,356
Speaker 1: And that's why we be one experiment and we have

312
00:19:18,436 --> 00:19:22,996
Speaker 1: to do the the engineering principle of design, build tests.

313
00:19:23,036 --> 00:19:26,436
Speaker 1: You design an expense, you build the DNA, you test it,

314
00:19:26,556 --> 00:19:28,476
Speaker 1: you learn a little bit, you go back and you

315
00:19:28,516 --> 00:19:31,396
Speaker 1: do it again and again, and every cycle of learning

316
00:19:32,076 --> 00:19:35,276
Speaker 1: is time and money and again, if you have the

317
00:19:35,356 --> 00:19:39,436
Speaker 1: right team, that cycle of learning will be faster than

318
00:19:39,476 --> 00:19:42,356
Speaker 1: the competition, will be cheaper on a competition, and that's

319
00:19:42,356 --> 00:19:46,516
Speaker 1: how we win. It is extremely, extremely difficult, and again

320
00:19:46,916 --> 00:19:50,516
Speaker 1: thankfully it's difficult. That that means we have a shot

321
00:19:50,316 --> 00:19:55,316
Speaker 1: at greatness. So yes, the good news is it's super

322
00:19:55,316 --> 00:19:57,636
Speaker 1: hard and nobody knows how to do it. So the

323
00:19:57,796 --> 00:20:01,756
Speaker 1: goal ultimately is to be able to store things in

324
00:20:01,836 --> 00:20:03,756
Speaker 1: DNA for a price that's comparable to what you could

325
00:20:03,756 --> 00:20:05,836
Speaker 1: put it on a hard drive for now, and in

326
00:20:05,956 --> 00:20:07,676
Speaker 1: that universe it would be cheaper in the long run

327
00:20:07,676 --> 00:20:12,596
Speaker 1: in the DNA. When do you think that'll happen? So

328
00:20:12,636 --> 00:20:14,476
Speaker 1: I know the answer, and I get that question from

329
00:20:14,516 --> 00:20:17,916
Speaker 1: all our investors all the time. Unfortunately I can't. I

330
00:20:17,956 --> 00:20:22,356
Speaker 1: can't tell you, but it's it's it's very soon. And

331
00:20:22,436 --> 00:20:25,436
Speaker 1: the chip we're working on now, that is the chip

332
00:20:25,956 --> 00:20:30,196
Speaker 1: that we're going to go commercial with and so the

333
00:20:30,276 --> 00:20:33,116
Speaker 1: chip is designed, the ship is build, and now we're

334
00:20:33,116 --> 00:20:36,556
Speaker 1: in the debugging phase and when when that is finished,

335
00:20:37,156 --> 00:20:40,676
Speaker 1: then we'll be able to launch commercially. And so does

336
00:20:40,676 --> 00:20:43,636
Speaker 1: that mean in a year? Does that mean in five years?

337
00:20:45,236 --> 00:20:47,796
Speaker 1: So if it's not done in five years, I'll definitely

338
00:20:47,876 --> 00:20:51,116
Speaker 1: get a lot of love letters from our investors. So

339
00:20:51,156 --> 00:20:56,076
Speaker 1: it's definitely a lot less than five years. In a minute,

340
00:20:56,076 --> 00:20:59,636
Speaker 1: it's the lightning round, including Emily's advice for solving hard

341
00:20:59,636 --> 00:21:03,476
Speaker 1: problems and what she actually does to solve hard problems.

342
00:21:10,316 --> 00:21:12,916
Speaker 1: Now let's get back to the show. The last thing

343
00:21:13,316 --> 00:21:15,956
Speaker 1: is a lightning round, a bunch of questions, but but

344
00:21:16,076 --> 00:21:19,396
Speaker 1: we'll do them fast. What is one piece of advice

345
00:21:19,436 --> 00:21:24,996
Speaker 1: you'd give to someone trying to solve a hard problem?

346
00:21:25,076 --> 00:21:31,316
Speaker 1: If I first, you don't succeed, try again, Right, you

347
00:21:31,836 --> 00:21:34,796
Speaker 1: have a choice. Right when you don't succeed, you can

348
00:21:34,916 --> 00:21:37,716
Speaker 1: call mom, you can order a pizza, you can drink

349
00:21:37,756 --> 00:21:40,556
Speaker 1: a beer. You try again. You have a favorite gene.

350
00:21:41,596 --> 00:21:44,556
Speaker 1: I do not have a favorite gene. I love all

351
00:21:44,636 --> 00:21:49,916
Speaker 1: the genes. They're like children, right, I guess you're their children.

352
00:21:49,956 --> 00:21:55,156
Speaker 1: We're their children. As a scientist, As a geneticist, do

353
00:21:55,196 --> 00:21:58,236
Speaker 1: you feel like you understand something about the world or

354
00:21:58,276 --> 00:22:02,316
Speaker 1: about people that most people don't really understand. I don't

355
00:22:02,356 --> 00:22:07,236
Speaker 1: think so. Yeah, I'm not very to shephilia as a person.

356
00:22:07,276 --> 00:22:13,116
Speaker 1: I'll probably somewhere on on some spectrum. So I know,

357
00:22:13,236 --> 00:22:15,156
Speaker 1: to make DNA and or to write, you know, to sell,

358
00:22:16,276 --> 00:22:19,836
Speaker 1: to do fundraising. But those are probably the four things.

359
00:22:20,036 --> 00:22:22,996
Speaker 1: Do you think you'll ever leave twist? You think they're

360
00:22:22,996 --> 00:22:24,676
Speaker 1: all kind of time when you want to go do

361
00:22:24,756 --> 00:22:26,916
Speaker 1: something else, I would be in the seat until I

362
00:22:26,996 --> 00:22:30,836
Speaker 1: get fired by by the board. So this is probably

363
00:22:30,876 --> 00:22:34,556
Speaker 1: my last job. I've read that you play the piano

364
00:22:34,676 --> 00:22:39,356
Speaker 1: when you're thinking about work. Is there some something in

365
00:22:39,396 --> 00:22:42,956
Speaker 1: particular you've been enjoying playing lately? A favorite piece these days?

366
00:22:44,556 --> 00:22:47,476
Speaker 1: So I practice the piano every day. I don't know

367
00:22:47,516 --> 00:22:50,756
Speaker 1: if I'm really playing it. Some people make coll it

368
00:22:51,076 --> 00:22:54,556
Speaker 1: butchering or murdering the piano, but I totally enjoy it.

369
00:22:54,796 --> 00:22:56,436
Speaker 1: What I love about the piano is that you know,

370
00:22:56,476 --> 00:22:59,276
Speaker 1: the left hand is easy, the right hand is easy.

371
00:22:59,316 --> 00:23:02,916
Speaker 1: But putting the two hands together, that's a total mind

372
00:23:03,676 --> 00:23:06,996
Speaker 1: bending experience. And it's you have to use one hundred

373
00:23:07,076 --> 00:23:09,836
Speaker 1: percent of your conscious mind. And when I find is

374
00:23:10,236 --> 00:23:13,276
Speaker 1: if I have a hard problem at work, I can't

375
00:23:13,276 --> 00:23:14,876
Speaker 1: figure that out, I go up to the piano. I'm

376
00:23:14,876 --> 00:23:17,596
Speaker 1: one hundred percent. I'm using one hundredercent of my brain

377
00:23:17,996 --> 00:23:21,756
Speaker 1: and somehow, in the background somewhere I think of a

378
00:23:21,836 --> 00:23:25,916
Speaker 1: solution and that that is quite exciting. And yeah, all

379
00:23:25,996 --> 00:23:28,636
Speaker 1: my tough problem. I never get them resolved sitting on

380
00:23:28,756 --> 00:23:31,196
Speaker 1: by desk, routing on my computer. It's I get them

381
00:23:31,316 --> 00:23:37,036
Speaker 1: resolved either playing piano or taking the dogs out for

382
00:23:37,236 --> 00:23:39,996
Speaker 1: a walk. So maybe that's your real advice for how

383
00:23:39,996 --> 00:23:42,396
Speaker 1: to solve a hard problem is don't think about the

384
00:23:42,436 --> 00:23:45,956
Speaker 1: problem or do something else. Yeah, take your prim out

385
00:23:45,996 --> 00:23:53,436
Speaker 1: of the building, take a hike. Emily Laprus is the

386
00:23:53,516 --> 00:23:59,676
Speaker 1: co founder and CEO of Twist Bioscience. Today's show was

387
00:23:59,716 --> 00:24:03,476
Speaker 1: produced by Edith Russlo, edited by Robert Smith, and engineered

388
00:24:03,476 --> 00:24:07,036
Speaker 1: by Amanda K. Wong. You can reach us at problem

389
00:24:07,196 --> 00:24:10,676
Speaker 1: at pushkin dot fm, or you can find me on

390
00:24:10,676 --> 00:24:14,636
Speaker 1: Twitter at Jacob Goldstein. I'm Jacob Goldstein and I'll be

391
00:24:14,676 --> 00:24:23,356
Speaker 1: back next week with another episode of What's Your Problem.