WEBVTT

1
00:00:00.051 --> 00:00:08.257
Hello Internet, this is your captain speaking Lester Nare, joined as always by my co-host and our resident PhD, Krishna Chowdhury.

2
00:00:08.637 --> 00:00:10.499
This is From First Principles.

3
00:00:10.879 --> 00:00:25.610
We have some great stories for you lined up this week, starting with an update on the interstellar visitor 3i Atlas as we've seen four new instruments capture images as it makes its way towards Earth, followed up by a story about longevity.

4
00:00:25.990 --> 00:00:30.373
If you were born after 1939, you may not live to see 100 years of age.

5
00:00:30.734 --> 00:00:35.317
We'll talk about that and Brian Johnson, aka the man who wants to live forever.

6
00:00:35.837 --> 00:00:41.201
Our third story is about a new no-sort plastic that will allow us to recycle more efficiently.

7
00:00:41.702 --> 00:00:46.866
And we will end today with an interesting analog AI story coming out of Microsoft.

8
00:00:47.206 --> 00:00:47.886
This is FFP.

9
00:00:48.166 --> 00:00:48.847
Let's get after it.

10
00:01:04.371 --> 00:01:04.791
My friend.

11
00:01:05.031 --> 00:01:05.511
How's it going?

12
00:01:06.392 --> 00:01:07.012
We're back again.

13
00:01:07.312 --> 00:01:07.732
Yeah.

14
00:01:08.072 --> 00:01:08.552
Another one.

15
00:01:09.152 --> 00:01:10.893
I think we have a little something here.

16
00:01:10.933 --> 00:01:13.634
People seem to be excited about talking about science.

17
00:01:13.654 --> 00:01:14.334
Yes, dude.

18
00:01:14.434 --> 00:01:14.954
In the weeds.

19
00:01:15.154 --> 00:01:17.675
I'm really enjoying seeing all the feedback that we're getting.

20
00:01:17.695 --> 00:01:18.395
100%.

21
00:01:18.475 --> 00:01:18.775
Yeah.

22
00:01:19.335 --> 00:01:22.976
We're going to start with a very popular science story that has been all over the headlines.

23
00:01:22.996 --> 00:01:23.436
That's right.

24
00:01:24.397 --> 00:01:25.357
About 3A Atlas.

25
00:01:26.177 --> 00:01:31.200
The headline this week, four powerful telescopes agree, interstellar comet 3I Atlas really is bizarre.

26
00:01:31.521 --> 00:01:35.083
This is from a variety of outlets, but from space.com is where we sort it.

27
00:01:35.103 --> 00:01:49.773
And the new information here is both NASA and ESA, the sort of North American and European space agencies, have used their instruments, Hubble, Sphear X, JWST, and TESS to capture images of the object as it makes its way towards our sun.

28
00:01:49.973 --> 00:01:50.233
Yeah.

29
00:01:50.693 --> 00:01:51.494
So what's new here?

30
00:01:52.014 --> 00:02:03.457
So what's new here is we've actually gotten really nice data from the James Webb Space Telescope, from Hubble, and from the VLT and from a lot of these things.

31
00:02:03.497 --> 00:02:09.098
And they're converging on trying to identify really what this object is actually like.

32
00:02:09.578 --> 00:02:10.338
Okay.

33
00:02:10.358 --> 00:02:13.119
And some of the, you know, nitty gritty characteristics.

34
00:02:13.499 --> 00:02:17.300
So just to recap our viewers about what 3i Atlas is.

35
00:02:17.680 --> 00:02:17.880
Right.

36
00:02:18.980 --> 00:02:28.688
It was discovered on the 1st of July of this year by the ATLAS program, which is the Asteroid Terrestrial Impact Last Alert System.

37
00:02:29.128 --> 00:02:36.074
And this is a system that is part of University of Hawaii's Institute of Astronomy.

38
00:02:36.394 --> 00:02:39.257
There's basically a bunch of telescopes all around the globe.

39
00:02:39.297 --> 00:02:40.317
There's two in Hawaii.

40
00:02:40.338 --> 00:02:42.659
There's one on Maui, one on the Big Island.

41
00:02:42.960 --> 00:02:45.502
Then there's one in South Africa, Chile, Australia.

42
00:02:46.283 --> 00:02:49.046
and Spain in Tenerife in the Canary Islands.

43
00:02:49.387 --> 00:02:51.650
This particular one was discovered in the one in Chile.

44
00:02:51.930 --> 00:02:52.231
Okay.

45
00:02:53.392 --> 00:02:59.821
The whole point of this guy is just to peruse the sky looking for transient objects and objects that move around.

46
00:02:59.881 --> 00:03:00.122
Okay.

47
00:03:00.482 --> 00:03:02.405
So it was discovered there and then it turns out that

48
00:03:02.765 --> 00:03:07.449
You know, other telescopes had already seen it but hadn't identified it as this thing.

49
00:03:07.489 --> 00:03:11.813
We'd talked about the Zwicky transient facility in Palomar Mountain near San Diego, right?

50
00:03:11.874 --> 00:03:12.054
Yes.

51
00:03:12.114 --> 00:03:18.440
That thing actually had already found it in May but hadn't, like, pinpointed it as, like, a thingy that we should be worried about.

52
00:03:18.580 --> 00:03:18.740
Yep.

53
00:03:19.000 --> 00:03:22.223
So once we identified it with the Atlas program…

54
00:03:23.584 --> 00:03:27.187
everybody all over the globe was like, oh, let's just peruse the data because we know where it is, right?

55
00:03:27.207 --> 00:03:27.887
And where it would have been.

56
00:03:28.167 --> 00:03:31.990
And so from that, we could then get the trajectory of this thing.

57
00:03:32.110 --> 00:03:32.330
Yes.

58
00:03:32.571 --> 00:03:34.352
And once we calculated that trajectory...

59
00:03:35.753 --> 00:03:38.235
It was like, oh, whoa, whoa, whoa, whoa, whoa, whoa, whoa.

60
00:03:38.255 --> 00:03:38.475
Okay?

61
00:03:38.776 --> 00:03:44.501
It's moving really fast, 61 kilometers per second, all the way out in Jupiter.

62
00:03:44.821 --> 00:03:44.981
Yep.

63
00:03:45.061 --> 00:03:45.342
Okay?

64
00:03:45.482 --> 00:03:49.946
And if it's moving that fast all the way out there, that means it's not going to stay inside the solar system.

65
00:03:49.986 --> 00:03:50.166
Right.

66
00:03:50.306 --> 00:03:52.428
You can do the Newton's laws and do the calculation.

67
00:03:52.748 --> 00:03:55.491
It'll go right past the sun and then on its way.

68
00:03:55.511 --> 00:03:55.691
Mm-hmm.

69
00:03:56.901 --> 00:04:00.083
Then we got into these things about people who had these really blurry images.

70
00:04:00.103 --> 00:04:01.224
They're like, is it aliens?

71
00:04:01.244 --> 00:04:03.226
It's never aliens.

72
00:04:03.446 --> 00:04:04.247
Much to my chagrin.

73
00:04:04.267 --> 00:04:05.167
Yeah, yeah, yeah.

74
00:04:05.407 --> 00:04:06.228
We always hope, right?

75
00:04:06.308 --> 00:04:07.069
One can always hope.

76
00:04:07.429 --> 00:04:09.711
But time always tells that it's not aliens.

77
00:04:09.811 --> 00:04:14.474
And it turns out to be the third interstellar object, right?

78
00:04:14.855 --> 00:04:21.740
Which is an object that is not part of our solar system, that is literally going from star system to star system.

79
00:04:22.340 --> 00:04:24.541
around the galaxy of the Milky Way, right?

80
00:04:24.561 --> 00:04:29.464
Because the idea is there are these large gaps of void space in between, you know, star systems.

81
00:04:29.524 --> 00:04:35.248
And so it's from a probabilities perspective to because of the vastness of space.

82
00:04:35.328 --> 00:04:35.568
Yeah.

83
00:04:36.528 --> 00:04:39.290
And our ability to detect both of those reasons.

84
00:04:39.470 --> 00:04:41.672
We haven't seen a lot of these types of objects before.

85
00:04:41.692 --> 00:04:43.013
Yeah, we haven't seen a lot.

86
00:04:43.053 --> 00:04:47.318
Now we will be able to because of things like the Vera Rubin coming online and things like that, right?

87
00:04:47.358 --> 00:04:47.898
Yes.

88
00:04:47.918 --> 00:04:48.799
This is our third one.

89
00:04:48.819 --> 00:04:50.000
That's why it's called 3i.

90
00:04:50.040 --> 00:04:51.622
The i is actually for interstellar.

91
00:04:51.642 --> 00:04:52.022
Oh, there you go.

92
00:04:52.182 --> 00:04:56.006
And 3i Atlas is because Atlas was the program that discovered it.

93
00:04:56.206 --> 00:04:56.406
Right.

94
00:04:56.467 --> 00:04:57.067
Very.

95
00:04:57.107 --> 00:04:59.931
The first one is one eye Oumuamua.

96
00:05:00.812 --> 00:05:03.656
That was the really weird elongated thingy.

97
00:05:04.077 --> 00:05:07.621
And that one we caught on the way out, so we couldn't really get good data on it.

98
00:05:07.662 --> 00:05:11.727
The second one was two eye Borisov, which turned out to be an interstellar comet.

99
00:05:12.087 --> 00:05:13.810
And then now this was three eye Atlas.

100
00:05:14.110 --> 00:05:28.264
And so a key difference between Oumuamua, let's say, and Atlas is that we were able to catch Atlas on its way in, not when it was on its way out, which gives more time to capture more data.

101
00:05:28.324 --> 00:05:28.965
Yeah.

102
00:05:29.065 --> 00:05:32.148
As well as being able to see how it behaves over longer time periods.

103
00:05:32.288 --> 00:05:33.389
Yeah, exactly.

104
00:05:33.409 --> 00:05:36.232
It's like that scene from Office when I was like, don't panic.

105
00:05:36.633 --> 00:05:37.474
I said, stay calm.

106
00:05:37.494 --> 00:05:38.915
Like, OK, now it's like it's happening.

107
00:05:38.955 --> 00:05:39.156
Right.

108
00:05:39.516 --> 00:05:53.011
So now everyone's like scrambling to figure out how to image this thing, because once we've got it, like every astronomer who's a planetary astronomer, who is, you know, or planetary origins or an exoplanets astronomer trying to understand how solar systems form.

109
00:05:53.351 --> 00:05:54.352
This is a very big deal.

110
00:05:54.692 --> 00:05:54.892
Okay.

111
00:05:54.912 --> 00:06:03.737
Because the other thing is if you turn back the clock, this thing came from above the Milky Way, which is where there's an old population of stars that are much older than our sun.

112
00:06:03.977 --> 00:06:04.138
Right.

113
00:06:04.178 --> 00:06:04.338
Right.

114
00:06:04.598 --> 00:06:07.319
So this thing could be something like 7 billion years old.

115
00:06:07.780 --> 00:06:10.521
Our solar system is tops 4.6, 4.7.

116
00:06:11.022 --> 00:06:14.984
So this thing is older than, you know, anything that we've really had a close encounter with.

117
00:06:15.024 --> 00:06:15.384
Right.

118
00:06:15.424 --> 00:06:16.305
Now, the...

119
00:06:17.787 --> 00:06:22.492
The problem is when this thing gets close to us, we're going to be on the other side of the sun.

120
00:06:22.733 --> 00:06:22.933
Right.

121
00:06:23.093 --> 00:06:23.874
So it's coming in.

122
00:06:24.114 --> 00:06:27.578
We're on the front side of the sun right now and it's coming in towards us.

123
00:06:27.698 --> 00:06:29.040
But we're going to be ordering.

124
00:06:29.320 --> 00:06:30.061
It won't be here.

125
00:06:30.201 --> 00:06:33.484
It'll still be coming this direction and we'll be blocked by the sun.

126
00:06:33.525 --> 00:06:34.325
By the sun, right?

127
00:06:34.365 --> 00:06:35.006
It's like during –

128
00:06:35.907 --> 00:06:38.608
If you wanted to see it, it would be during daytime behind the sun.

129
00:06:38.648 --> 00:06:38.829
Yep.

130
00:06:39.049 --> 00:06:39.269
Right?

131
00:06:39.449 --> 00:06:43.191
Which, because of the atmosphere, we're not going to see anything.

132
00:06:43.211 --> 00:06:49.734
The sky is going to be blue, and the air glows more than any stars or this 3i atlas.

133
00:06:49.774 --> 00:06:53.496
This is why Vera Rubin is doing a survey, sky survey at night.

134
00:06:53.656 --> 00:06:53.896
Yeah.

135
00:06:54.216 --> 00:06:55.016
And not during the day.

136
00:06:55.116 --> 00:06:56.057
Yes, exactly.

137
00:06:56.257 --> 00:07:00.458
And furthermore, we can't really point the James Webb Space Telescope towards it either.

138
00:07:00.918 --> 00:07:10.582
Because, like, then the sun, the whole point of the James Webb Telescope, the reason why it's so nice and, like, sensitive is because it's got this giant shade.

139
00:07:11.222 --> 00:07:15.404
That shades it from the sun so that the instruments can be kept at like four Kelvin.

140
00:07:15.604 --> 00:07:15.784
Right.

141
00:07:15.844 --> 00:07:24.848
Like colder than like actually no even less than like there's like a literal like dilution refrigerator on that thing that is like keeping it colder than outer space.

142
00:07:25.028 --> 00:07:25.208
Right.

143
00:07:25.268 --> 00:07:28.089
Because then you get these instruments that are super cold.

144
00:07:28.489 --> 00:07:30.370
The instruments the atoms aren't jiggling.

145
00:07:30.470 --> 00:07:30.630
Right.

146
00:07:30.650 --> 00:07:33.831
So then the only signal you're getting is not from the instrument itself.

147
00:07:33.931 --> 00:07:34.111
Right.

148
00:07:34.151 --> 00:07:35.672
But from like the starlight.

149
00:07:35.692 --> 00:07:35.892
Right.

150
00:07:36.372 --> 00:07:37.733
There's less or no noise.

151
00:07:37.833 --> 00:07:38.334
Yeah, yes.

152
00:07:38.434 --> 00:07:42.756
But if you wanted to point it towards that thing, that defeats the whole purpose.

153
00:07:42.776 --> 00:07:43.337
That makes sense.

154
00:07:43.357 --> 00:07:44.357
So you can't really do that.

155
00:07:44.457 --> 00:07:45.158
That makes sense.

156
00:07:45.758 --> 00:07:53.343
So what we'd like to do is to observe it right now with our terrestrial-based telescopes like the James Webb.

157
00:07:53.443 --> 00:07:54.123
Yes.

158
00:07:54.624 --> 00:07:56.605
Now when it's sort of away from the sun.

159
00:07:56.725 --> 00:07:56.925
Yes.

160
00:07:56.985 --> 00:07:58.186
Get as much data as possible.

161
00:07:58.286 --> 00:07:58.466
Yes.

162
00:07:58.746 --> 00:08:02.149
And then as it goes through the solar system and it reaches Mars.

163
00:08:02.169 --> 00:08:02.309
Yes.

164
00:08:02.469 --> 00:08:02.649
Yes.

165
00:08:03.069 --> 00:08:04.350
It's going to be on the other side of the sun.

166
00:08:04.390 --> 00:08:05.811
It's going to be near Mars' orbit.

167
00:08:05.871 --> 00:08:06.752
Yes.

168
00:08:06.892 --> 00:08:10.074
We can have all of our orbiters that are orbiting Mars.

169
00:08:10.234 --> 00:08:10.474
Right.

170
00:08:11.275 --> 00:08:16.058
Sort of... And then point towards this thing and try to get as much data as possible.

171
00:08:16.338 --> 00:08:20.521
So the idea is we have orbiters around Mars that have a certain mission right now, right?

172
00:08:20.541 --> 00:08:21.442
A Mars-based mission.

173
00:08:23.103 --> 00:08:23.783
They can take a break.

174
00:08:24.484 --> 00:08:25.405
This is one of those, like...

175
00:08:26.365 --> 00:08:28.926
Not that often in the lifetime opportunity.

176
00:08:28.946 --> 00:08:48.931
It's like we can take a couple hours a day, reorient, use some of our fuel to do so, to capture, to provide that data when we on Earth, as well as JWST, don't have the capability or the angle of viewing to be able to actually properly capture it as it gets closer to our orbit.

177
00:08:49.071 --> 00:08:49.891
Yeah, and exactly.

178
00:08:49.911 --> 00:08:53.932
And with comets, you really want to look at the thing when it's close to the sun.

179
00:08:54.552 --> 00:08:57.695
Because that's when it's doing the most cool, crazy stuff.

180
00:08:57.995 --> 00:08:58.136
Right.

181
00:08:58.176 --> 00:08:58.436
Okay?

182
00:08:58.756 --> 00:09:00.918
When the comet is far away, it's just a chunk of ice.

183
00:09:01.519 --> 00:09:04.802
But as it gets closer to the sun, it's going to start interacting with the solar wind.

184
00:09:05.162 --> 00:09:08.565
It's going to start interacting with the photons that the sun is giving out.

185
00:09:08.866 --> 00:09:13.210
So then there's going to be chemical reactions and physical reactions happening on top of the comet.

186
00:09:13.470 --> 00:09:14.611
And that's what we want to see.

187
00:09:14.771 --> 00:09:14.971
Right?

188
00:09:15.292 --> 00:09:16.032
In the very, like...

189
00:09:17.414 --> 00:09:24.139
Way out there, when we tried to look at 3i Atlas, it was really bright for how far away it was.

190
00:09:24.419 --> 00:09:29.743
And that's what gave rise to a lot of these alternative theories about it's artificial and all this other kind of stuff.

191
00:09:29.763 --> 00:09:31.564
The light's not reflecting from the sun.

192
00:09:31.604 --> 00:09:32.725
It's generating its own light.

193
00:09:32.745 --> 00:09:34.626
It's generating its own light because it's so bright.

194
00:09:34.806 --> 00:09:35.667
How is that possible?

195
00:09:36.047 --> 00:09:41.011
Now, as it got closer, actually, so August 6th is when James Webb looked at it.

196
00:09:41.411 --> 00:09:43.493
And the preprint came out on August 25th.

197
00:09:43.853 --> 00:09:46.515
They were fast about this because they know that people want it.

198
00:09:46.615 --> 00:09:47.075
Right.

199
00:09:47.256 --> 00:09:54.542
So the preprint came out like literally they had like what, like 20 days to do all of the people were upset about how long it took.

200
00:09:54.622 --> 00:09:54.842
Yeah.

201
00:09:54.882 --> 00:09:55.062
Yeah.

202
00:09:55.082 --> 00:10:00.767
But dude, like 20 days to like massage data is kind of like create a story.

203
00:10:00.827 --> 00:10:00.967
Right.

204
00:10:01.007 --> 00:10:02.869
It's not just like you're analyzing the data.

205
00:10:02.889 --> 00:10:04.630
You have to like interpret the data.

206
00:10:04.650 --> 00:10:06.992
Try to create a story about like what you're trying to say.

207
00:10:07.112 --> 00:10:07.332
Right.

208
00:10:07.693 --> 00:10:09.494
Science is all about stories at the end of the day.

209
00:10:09.514 --> 00:10:10.335
It's not just numbers.

210
00:10:10.495 --> 00:10:10.655
Right.

211
00:10:10.675 --> 00:10:12.196
You have to have some kind of narrative.

212
00:10:12.336 --> 00:10:12.496
Right.

213
00:10:12.557 --> 00:10:12.777
So.

214
00:10:14.478 --> 00:10:17.081
They analyzed the data from James Webb Space Telescope.

215
00:10:17.581 --> 00:10:20.584
It used one of the instruments that is a spectrograph.

216
00:10:20.924 --> 00:10:23.447
So it can tell us things like the composition of this thing.

217
00:10:23.887 --> 00:10:28.611
And what they found was a really high mixing ratio for carbon dioxide and water.

218
00:10:29.192 --> 00:10:29.472
Okay.

219
00:10:29.832 --> 00:10:32.655
So usually the mixing ratio is quite low.

220
00:10:33.075 --> 00:10:34.297
Here it's eight to one.

221
00:10:34.877 --> 00:10:35.117
Okay.

222
00:10:35.478 --> 00:10:35.738
Okay.

223
00:10:35.778 --> 00:10:37.659
There's eight times as much CO2 as water.

224
00:10:37.759 --> 00:10:37.940
Right.

225
00:10:38.040 --> 00:10:39.161
That's like a lot.

226
00:10:39.321 --> 00:10:39.481
Yes.

227
00:10:39.621 --> 00:10:39.861
Okay.

228
00:10:39.881 --> 00:10:43.664
And that's very unusual for comets in our solar system.

229
00:10:43.844 --> 00:10:44.064
Correct.

230
00:10:44.105 --> 00:10:44.365
Okay.

231
00:10:44.665 --> 00:10:47.407
So that means that this thing, there's two possibilities.

232
00:10:47.527 --> 00:10:47.767
Okay.

233
00:10:48.348 --> 00:10:50.530
One is that when this thing was born, Mm-hmm.

234
00:10:51.591 --> 00:10:56.575
it had a very different birth environment than the comets in our solar system.

235
00:10:56.876 --> 00:11:00.119
It could be that this thing was born in something called the CO2 ice line.

236
00:11:00.499 --> 00:11:00.779
Okay.

237
00:11:00.819 --> 00:11:06.044
You can imagine when a star is forming and the solar system is forming around it, the star starts...

238
00:11:07.105 --> 00:11:10.589
making light and thermal energy, that starts going out, right?

239
00:11:10.609 --> 00:11:11.630
There's going to be...

240
00:11:11.710 --> 00:11:24.263
The disk is going to have a mixture of silicon, hydrogen, CO2, water, and all the dust from that primordial nebula that sort of became the star, right?

241
00:11:24.643 --> 00:11:27.606
So all of this stuff is going to be in the disk.

242
00:11:28.006 --> 00:11:29.127
There's going to be a line...

243
00:11:29.728 --> 00:11:38.353
Where inside that line, close to the star, there's going to be enough radiation to put CO2 in a gaseous form.

244
00:11:38.613 --> 00:11:38.913
Okay.

245
00:11:38.933 --> 00:11:39.093
Okay?

246
00:11:39.333 --> 00:11:44.976
But afterwards, there's not going to be enough radiation, not going to be enough heat, and that CO2 is going to become dry ice.

247
00:11:45.557 --> 00:11:45.797
Okay?

248
00:11:46.137 --> 00:11:53.441
So there's going to be this ice line where after a certain distance away from the star, you just have dry ice and before you have gaseous ice.

249
00:11:53.661 --> 00:11:54.862
Because also... Gaseous CO2.

250
00:11:55.082 --> 00:11:56.723
Carbon dioxide does not have...

251
00:11:56.743 --> 00:11:57.503
There's no liquid form.

252
00:11:57.683 --> 00:11:57.983
Yeah, I mean...

253
00:11:58.003 --> 00:11:59.343
So it goes straight from gas to solid.

254
00:11:59.383 --> 00:11:59.804
Yeah, yeah.

255
00:11:59.884 --> 00:12:04.285
I mean, there is a liquid form, but at, like, very peculiar pressures, right?

256
00:12:04.305 --> 00:12:04.965
Okay, that's fair.

257
00:12:05.265 --> 00:12:15.307
Like, in the low pressure of space, and even at, like, one atmosphere, there's... Like, at one atmosphere, it'll go directly from dry ice to CO2, right?

258
00:12:15.407 --> 00:12:17.648
If you put in high enough pressure, you can liquefy CO2.

259
00:12:17.668 --> 00:12:18.528
Okay, that makes sense.

260
00:12:18.788 --> 00:12:18.988
Right?

261
00:12:19.028 --> 00:12:22.449
But, like, out here, it's just going to go ice to gas.

262
00:12:22.489 --> 00:12:23.229
And back and forth.

263
00:12:23.249 --> 00:12:23.409
Right?

264
00:12:23.490 --> 00:12:23.990
Back and forth.

265
00:12:24.030 --> 00:12:24.290
So...

266
00:12:25.830 --> 00:12:33.274
You have this like dry ice line now you could be that like the comet was forming in this right past that ice line, right?

267
00:12:33.334 --> 00:12:41.658
So it's like gathering up a lot of co2 not a lot is a lot of water right because water the water line is elsewhere at right right right so

268
00:12:43.873 --> 00:12:48.996
That could be one reason why 3i Atlas has a lot of CO2 versus water.

269
00:12:49.316 --> 00:12:57.621
The other thing could be that it formed or it was subjected to a really harsh radiative environment.

270
00:12:57.821 --> 00:12:58.042
Okay.

271
00:12:58.102 --> 00:12:58.362
Okay.

272
00:12:58.662 --> 00:13:04.445
And then what can happen is you can have these reactions between carbon monoxide, which is also fairly common.

273
00:13:04.585 --> 00:13:05.126
Yes.

274
00:13:05.206 --> 00:13:10.349
Carbon monoxide and water with this high energy radiation to create CO2.

275
00:13:10.509 --> 00:13:10.709
Got it.

276
00:13:10.729 --> 00:13:11.750
And that's what happened, right?

277
00:13:11.930 --> 00:13:12.110
Got it.

278
00:13:12.130 --> 00:13:18.076
So there's these two possible scenarios, but in any case, it's still very, very different from any comet that we've seen.

279
00:13:18.557 --> 00:13:30.029
Which is, I think, what's interesting is there's this leap, you know, that folks like me who are not as educated always make where it's like, oh, we have no reference points, so that must mean it's aliens.

280
00:13:30.409 --> 00:13:32.751
And I think kind of what the through line you're drawing here is...

281
00:13:33.472 --> 00:13:41.717
We do know that it was birthed in an environment that is different than the environment where we have these data sets.

282
00:13:41.837 --> 00:13:42.037
Yeah.

283
00:13:42.397 --> 00:13:42.577
Yeah.

284
00:13:42.717 --> 00:13:44.738
And so those initial conditions matter.

285
00:13:44.998 --> 00:13:46.139
Yeah.

286
00:13:46.299 --> 00:13:49.061
And the data sets, to your point, the data sets that we have,

287
00:13:49.921 --> 00:13:53.523
on what is normal is like from our solar system.

288
00:13:53.663 --> 00:14:01.428
As you say, right, there's millions and millions of solar systems out there with different types of stars, different galactic environments.

289
00:14:01.848 --> 00:14:05.730
Like we are a tiny little sample size, right?

290
00:14:06.070 --> 00:14:13.935
And one of the great things and why everyone is so excited about interstellar objects is because it gives us a sampling of what's out there.

291
00:14:14.135 --> 00:14:14.315
Yes.

292
00:14:14.395 --> 00:14:14.615
Right.

293
00:14:14.715 --> 00:14:19.737
And so that's why everyone is like scrambling to try to get as much data as possible on this guy.

294
00:14:19.777 --> 00:14:19.957
Right.

295
00:14:20.118 --> 00:14:32.043
Because from that data, like we kind of talked about at the beginning of the story, there are multiple insights that can be gleaned, not only at time of capture, but also afterwards when we learn other things.

296
00:14:32.083 --> 00:14:34.924
And it's like, oh, let's go look back at the Atlas data.

297
00:14:35.064 --> 00:14:35.284
Yeah.

298
00:14:35.364 --> 00:14:42.507
Now that we have some different context about how these highly radiative environments work.

299
00:14:42.827 --> 00:15:11.825
work yeah or any number of other yeah options yeah exactly we can like start i mean you know we work we're kind of scrambling because like this thing's going to be out by september right right so right now we're just in the observe phase right okay after this thing gets out and we get all the data then i'm pretty sure there's going to be a big cohort of people that are going to go into the modeling phase right of trying to understand okay theoretically how is something like this possible right like if i were to model it this way would it match the data no well then that's not

300
00:15:12.205 --> 00:15:14.086
That's probably not what happened.

301
00:15:14.286 --> 00:15:20.608
And so, yeah, at this point, it's just like everyone's just like just pointed at the thing and get as much light as you can.

302
00:15:20.688 --> 00:15:32.272
What's so funny about this is, as I mentioned in a previous episode, I get the pleasure of being able to work with Dr. Avi Loeb on a regular basis, who is a professor of astrophysics at Harvard.

303
00:15:33.872 --> 00:15:40.777
you know, one of the things we'd reached out to him to do was to do a white paper on this new update from 3i Atlas.

304
00:15:40.817 --> 00:15:50.864
And to go with what you're just saying, his response was almost a more expanded version of what you just said, which is like, we need to wait for the dust to settle.

305
00:15:50.924 --> 00:15:53.325
Right now, it's just an observational stage.

306
00:15:53.786 --> 00:15:57.228
We need to focus on capturing as much data and information as possible.

307
00:15:57.908 --> 00:16:08.110
Once it's gone and that opportunity has left, then there's going to be time to actually settle in and do analysis on what it is that we just captured.

308
00:16:08.331 --> 00:16:17.153
So I think there's an important distinction to identify here, which is like even if it was aliens, right, we don't have any real baseline to reference for it, to identify it as such.

309
00:16:17.413 --> 00:16:26.975
And if it doesn't land on the White House lawn, there's going to be this two-stage data capture and observation, which is where we are now, and then sort of analysis and modeling, which happens after.

310
00:16:27.635 --> 00:16:38.018
And we will get more clarity, generally speaking, once the observation period is over and we move into this analysis and modeling timeframe, which is, again, how science works.

311
00:16:38.058 --> 00:16:38.998
Yeah, exactly.

312
00:16:39.218 --> 00:16:46.740
But there's a big public desire for answers, which is understandable.

313
00:16:48.201 --> 00:16:51.062
And like many things, it's just a process.

314
00:16:51.082 --> 00:16:53.123
So it's interesting you say that because it makes me think of that.

315
00:16:53.343 --> 00:16:53.824
Yeah.

316
00:16:53.844 --> 00:16:54.324
That letter we got.

317
00:16:54.344 --> 00:17:03.088
And like, and you know, it's definitely not aliens at this point, but it's still like one of the coolest objects that's ever visited us.

318
00:17:03.729 --> 00:17:03.949
Right.

319
00:17:04.369 --> 00:17:07.711
And so, I mean, I don't think people should be disappointed.

320
00:17:08.451 --> 00:17:08.671
Right.

321
00:17:09.011 --> 00:17:09.832
It's kind of like.

322
00:17:10.923 --> 00:17:11.563
A cop-out.

323
00:17:12.384 --> 00:17:15.845
I mean, you know, as the Buddhists say, like, don't have these expectations.

324
00:17:15.925 --> 00:17:19.006
Because the world itself is a wonderful place.

325
00:17:19.146 --> 00:17:19.346
Right.

326
00:17:19.526 --> 00:17:19.986
Right?

327
00:17:20.086 --> 00:17:24.968
Imagine this thing, like, going through the Milky Way for seven billion years.

328
00:17:25.148 --> 00:17:25.448
Right.

329
00:17:25.488 --> 00:17:27.228
Visiting, like, who knows what, right?

330
00:17:27.448 --> 00:17:29.949
This isn't... We're not the first star system that it's seen.

331
00:17:30.089 --> 00:17:30.269
Mm-hmm.

332
00:17:30.349 --> 00:17:31.590
Imagine from its point of view.

333
00:17:31.670 --> 00:17:34.031
It's gone from one star system to the next.

334
00:17:34.311 --> 00:17:38.232
The fact that it's going so fast means that it's had a pretty close encounter.

335
00:17:38.252 --> 00:17:38.492
Yeah, right.

336
00:17:38.632 --> 00:17:40.053
With something pretty massive.

337
00:17:40.073 --> 00:17:40.234
Right.

338
00:17:40.434 --> 00:17:42.295
To like be zooming through like this.

339
00:17:42.356 --> 00:17:42.556
Right.

340
00:17:42.816 --> 00:17:46.840
The sun is not really going to change its velocity that much.

341
00:17:46.920 --> 00:17:47.180
Got it.

342
00:17:47.260 --> 00:17:47.460
Right.

343
00:17:47.500 --> 00:17:50.623
It's just, it's sort of like redirecting it in one direction.

344
00:17:50.763 --> 00:17:50.903
Yep.

345
00:17:50.923 --> 00:17:53.226
But like this thing had a very close encounter.

346
00:17:53.526 --> 00:17:57.850
Like this, you know, like imagine from its perspective, it's been around...

347
00:17:58.150 --> 00:18:01.512
for like half the age of the universe.

348
00:18:02.272 --> 00:18:02.833
That's a long time.

349
00:18:02.873 --> 00:18:03.593
That's a long time.

350
00:18:03.613 --> 00:18:04.994
We've only been around for a third.

351
00:18:05.294 --> 00:18:08.696
So I think it's an incredible story, right?

352
00:18:08.776 --> 00:18:13.779
And to get back to the idea of we're trying to observe this thing as much as possible, right?

353
00:18:13.999 --> 00:18:17.441
So that's where the European Space Agency and this new stuff is coming in, right?

354
00:18:17.721 --> 00:18:20.483
Where the European Space Agency has a bunch of orbiters.

355
00:18:21.143 --> 00:18:42.099
around mars that they're like okay we're gonna we're gonna repurpose for this one of them is um the mars express and the other one is the exo-mars trace gas orbiter both of these both of these things have cameras that are pretty good this thing is still going to be pretty far away from mars but you know getting any sort of anything getting anything when it's on the other side of the sun and we can't put anything there

356
00:18:42.699 --> 00:18:45.300
And it's going to be the most volatile when it's on the other side of the sun.

357
00:18:45.540 --> 00:18:54.144
So getting any data at that point of its orbit, of its trajectory, I should say, not orbit because it's not orbiting, is going to be incredibly valuable.

358
00:18:54.224 --> 00:18:57.966
And the guys at the ESA are definitely tempering everyone's expectations.

359
00:18:57.986 --> 00:18:59.686
It's like, guys, it's going to be far away.

360
00:18:59.766 --> 00:19:01.707
Like, our stuff wasn't designed for this.

361
00:19:01.727 --> 00:19:06.049
It's like, okay, dude, like, no one's going to blame you if you get shit data.

362
00:19:06.269 --> 00:19:08.010
Like, just get whatever you can.

363
00:19:08.930 --> 00:19:13.312
And then the other thing is, so we also talked about the Europa Clipper in one of the previous episodes.

364
00:19:13.652 --> 00:19:21.516
So on its way back to Earth, it's actually going to cross through where the comet was.

365
00:19:21.616 --> 00:19:25.698
So as a comet goes across the solar system, it leaves behind...

366
00:19:26.518 --> 00:19:29.960
a trail of particles, of debris in its wake.

367
00:19:30.240 --> 00:19:32.341
That's actually how we get meteor showers.

368
00:19:32.801 --> 00:19:40.404
Meteor showers are literally just the Earth moving through cometary paths that have these particles.

369
00:19:40.824 --> 00:19:45.986
And just like planets, the particles are going to be going through the sun.

370
00:19:46.186 --> 00:19:47.787
They're orbiting the sun as well.

371
00:19:48.127 --> 00:19:52.949
So as the Earth goes through these cometary paths...

372
00:19:53.835 --> 00:20:01.225
any sort of particles that get trapped in the Earth's gravity fall down to Earth, and then that's how you get the Perseids and all of these meteor showers, right?

373
00:20:01.525 --> 00:20:06.992
But you can now imagine the Europa Clipper probe is going to be going through this thing, right?

374
00:20:07.272 --> 00:20:09.155
So we could be getting observations about

375
00:20:10.763 --> 00:20:13.386
the particulate matter that, you know, was a part of that.

376
00:20:13.406 --> 00:20:13.627
Yeah.

377
00:20:13.687 --> 00:20:13.827
Yeah.

378
00:20:13.847 --> 00:20:15.249
And it's not just new Europa Clipper.

379
00:20:15.269 --> 00:20:19.294
I think there's other places like Hira and Lucy, they may fly through the commentary tail.

380
00:20:19.694 --> 00:20:20.836
They're all NASA.

381
00:20:21.016 --> 00:20:21.697
Yep.

382
00:20:21.717 --> 00:20:25.882
So that's going to be super exciting to get like actual, like, you know, data.

383
00:20:26.563 --> 00:20:28.265
It's going to be both optical, like imagery.

384
00:20:28.685 --> 00:20:31.266
capture as well as like physical matter analysis.

385
00:20:32.027 --> 00:20:47.254
And I think this sort of, you know, this brings up the, when we build these really expensive, highly precise instruments for observation and data capture, they're not necessarily single purpose in this exact context.

386
00:20:47.274 --> 00:20:47.775
Yeah, yeah.

387
00:20:47.815 --> 00:20:54.558
It's like we send a bunch of stuff to look at Mars, but we can still use those same instruments for other use cases, such as,

388
00:20:54.898 --> 00:20:58.120
Hey, go just look at this thing that's coming through and let's see what we can get.

389
00:20:58.220 --> 00:20:58.640
Yeah.

390
00:20:58.901 --> 00:20:59.101
Yeah.

391
00:20:59.141 --> 00:21:08.306
It's pretty crazy because like I'm pretty sure when the ESA and when NASA like made these things, we didn't – you know, Oumu'a wasn't even a thing.

392
00:21:08.346 --> 00:21:08.527
Right.

393
00:21:08.787 --> 00:21:09.007
Right?

394
00:21:09.247 --> 00:21:11.388
When they were planning like – These were made well before we even knew.

395
00:21:11.408 --> 00:21:18.413
Well before we even – and then now it's like – now it's kind of nice that humanity has a solar system-wide presence.

396
00:21:18.433 --> 00:21:18.553
Right.

397
00:21:18.573 --> 00:21:18.713
Right.

398
00:21:18.993 --> 00:21:31.045
You know, where even though Earth is in a crappy part of its orbit and we're kind of handcuffed in that way, we can still get that data from other parts because we have this presence that's so big.

399
00:21:31.426 --> 00:21:32.307
That's kind of cool.

400
00:21:32.387 --> 00:21:33.388
It's really cool.

401
00:21:33.448 --> 00:21:36.331
It's why both near and deep space matter quite a bit.

402
00:21:37.012 --> 00:21:41.016
And while it may be the case...

403
00:21:42.602 --> 00:21:46.045
based on our second article, that we may not live past 100.

404
00:21:46.486 --> 00:21:47.046
Yeah.

405
00:21:47.407 --> 00:21:52.752
Let's hope that these tools and instruments maybe one day do capture alien life in our local neighborhood.

406
00:21:53.012 --> 00:21:53.973
Oh, that would be dope.

407
00:21:53.993 --> 00:21:54.894
That would be fantastic.

408
00:21:54.914 --> 00:21:56.035
That would be fantastic.

409
00:21:56.335 --> 00:22:02.781
But our second story is a similarly interesting one about longevity no longer increasing.

410
00:22:03.041 --> 00:22:03.482
Headline...

411
00:22:03.802 --> 00:22:09.244
Studies find generations born after 1939 unlikely to reach age 100.

412
00:22:09.504 --> 00:22:11.925
This is covered in the Independent.

413
00:22:12.126 --> 00:22:22.430
But the research was out of it was an international collaboration between the Max Planck Institute of Demographic Research, the Institut National d'Etudes Démographiques,

414
00:22:23.010 --> 00:22:27.992
And the University of Wisconsin-Madison, who we've mentioned on this podcast a couple of times now.

415
00:22:28.493 --> 00:22:34.015
And they published their paper in PNAS, which is the Proceedings of the National Academy of Sciences.

416
00:22:34.735 --> 00:22:45.920
And so people are probably most familiar with longevity as a concept or studies, either from the Netflix documentary, The Man Who Thinks He Can Live Forever, which is about this hundred millionaire, Brian Johnson.

417
00:22:46.260 --> 00:22:46.901
What is he doing?

418
00:22:47.021 --> 00:22:49.422
So he's doing all, he's getting the best doctors in the world.

419
00:22:49.502 --> 00:22:49.842
Okay.

420
00:22:50.082 --> 00:22:51.683
To like give him a regiment.

421
00:22:52.643 --> 00:23:01.854
So he's now allegedly decreased his actual age from his mid-40s, where he's in his 40s, to the bodily age of a 20-something-year-old.

422
00:23:02.314 --> 00:23:02.514
Right?

423
00:23:02.995 --> 00:23:04.416
Wow.

424
00:23:04.637 --> 00:23:09.342
So his whole thing is, unless I die in a freak accident, my body is literally...

425
00:23:09.702 --> 00:23:12.824
Is he one of these guys who gets blood transfusions?

426
00:23:13.304 --> 00:23:15.125
So I don't want to say specifically.

427
00:23:15.145 --> 00:23:15.726
Yeah, okay.

428
00:23:15.766 --> 00:23:20.269
But the only two angles are either the Brian Johnson story or the idea that all these billionaires are vampires.

429
00:23:20.289 --> 00:23:20.409
Yeah.

430
00:23:20.429 --> 00:23:22.150
Getting their young employees to give them blood.

431
00:23:22.190 --> 00:23:24.091
Yeah, because that's like a real thing.

432
00:23:24.111 --> 00:23:26.172
I saw news stories about that.

433
00:23:26.232 --> 00:23:27.713
I'm not sure with Brian specifically.

434
00:23:27.733 --> 00:23:29.234
Yeah, maybe Brian Johnson isn't doing that.

435
00:23:29.294 --> 00:23:29.534
Okay.

436
00:23:29.554 --> 00:23:31.275
But this is how in the pop culture.

437
00:23:31.335 --> 00:23:32.236
Everyone's obsessed with it.

438
00:23:32.336 --> 00:23:32.976
Right.

439
00:23:33.037 --> 00:23:34.718
People want to live longer, all this stuff.

440
00:23:35.938 --> 00:23:41.820
It's always been the case over the course of the year that like as generations go by we get older We are able to live longer.

441
00:23:41.960 --> 00:23:44.321
Yeah, this study is sort of suggesting not anymore.

442
00:23:44.601 --> 00:23:45.101
Not anymore.

443
00:23:45.141 --> 00:23:48.722
We've yeah We might have come to the end of the road with the low-hanging fruits.

444
00:23:48.802 --> 00:23:50.002
That's what this thing is suggest.

445
00:23:50.062 --> 00:23:50.302
Okay.

446
00:23:50.522 --> 00:23:55.444
Okay So longevity there's been this thing called the longevity revolution over the over the past century, right?

447
00:23:55.764 --> 00:23:59.985
life expectancy has been increasing like crazy over the last century and

448
00:24:00.205 --> 00:24:03.166
And that has everything to do with advancements in science.

449
00:24:03.386 --> 00:24:03.727
100%.

450
00:24:03.767 --> 00:24:03.967
Right?

451
00:24:04.787 --> 00:24:09.309
We've, like, figured out biology at a nitty-gritty level.

452
00:24:09.689 --> 00:24:14.251
We figured out, hey, like, sickness actually comes from little tiny pathogens.

453
00:24:14.691 --> 00:24:22.655
And we figured out ways to advance public health with vaccines, with antibiotics, with all kinds of, like, revolutionary technologies.

454
00:24:22.655 --> 00:24:24.857
technology, medical innovations.

455
00:24:25.117 --> 00:24:27.438
We've also had a lot of socioeconomic development.

456
00:24:27.859 --> 00:24:38.766
So in sanitation, in the eradication of infectious diseases, smallpox is no longer a thing, which is an incredible thing to think about, dude.

457
00:24:38.827 --> 00:24:41.468
Smallpox doesn't exist in the world.

458
00:24:41.729 --> 00:24:46.272
Which is controversial to bring up the fact that that's an incredible feat for humanity.

459
00:24:46.312 --> 00:24:47.953
For humanity and science and everything.

460
00:24:48.193 --> 00:24:48.954
Dude, I mean, and

461
00:24:50.835 --> 00:25:04.440
You know, I can't imagine it happening today, which is so sad, which is so sad, because like back when in the I think it was in the 1980s when, you know, we were at the last stages of eradicating smallpox.

462
00:25:04.540 --> 00:25:07.441
And there was just a tiny little bit of smallpox in Africa.

463
00:25:07.801 --> 00:25:09.862
OK. And in the middle of a civil war.

464
00:25:10.382 --> 00:25:22.908
Both sides decided to have a ceasefire so that health workers could go in and because there had been a small, smallpox outbreak there and health workers could go in and vaccinate everyone there.

465
00:25:23.828 --> 00:25:25.289
And then they started fighting each other.

466
00:25:25.309 --> 00:25:28.490
It's like, it's like, just stop fighting for like, for like a few days.

467
00:25:28.590 --> 00:25:29.771
Give us the time here.

468
00:25:30.131 --> 00:25:31.472
And then everyone's like, probably like.

469
00:25:32.232 --> 00:25:34.754
planning how to restart the fighting.

470
00:25:35.034 --> 00:25:39.698
But even in the Civil War, they respected the fact that, okay, nobody wants smallpox.

471
00:25:39.758 --> 00:25:39.978
Right.

472
00:25:40.638 --> 00:25:47.183
Like, I want to win the Civil War, but not at the cost of a nation with smallpox.

473
00:25:47.783 --> 00:25:49.625
I can't imagine that happening today.

474
00:25:49.645 --> 00:25:58.972
There's no – I mean, a funny story that someone always brings up to me about this is the former Senate Majority Leader, Mitch McConnell, I think when he was younger, had –

475
00:25:59.812 --> 00:26:00.052
Smallpox.

476
00:26:00.072 --> 00:26:00.812
That's crazy.

477
00:26:01.112 --> 00:26:13.295
And, you know, leader of the Republican Party in the United States for a long period of time, he's been very vocal about the defense of like science in this particular isolated use case.

478
00:26:13.335 --> 00:26:14.055
Yeah.

479
00:26:14.215 --> 00:26:20.116
And to your point, imagining any of the current global conflicts going on pausing to do a public health.

480
00:26:20.657 --> 00:26:21.317
No, I can't.

481
00:26:21.637 --> 00:26:22.017
It's hard.

482
00:26:22.117 --> 00:26:22.857
It's hard to envision.

483
00:26:23.057 --> 00:26:24.058
Yeah, it really is.

484
00:26:24.078 --> 00:26:24.738
It's hard to envision.

485
00:26:24.758 --> 00:26:24.918
Yeah.

486
00:26:25.098 --> 00:26:26.019
Which is sad.

487
00:26:26.159 --> 00:26:26.879
It really is sad.

488
00:26:27.099 --> 00:26:30.781
But like those are the advancements that got us this longevity revolution, right?

489
00:26:30.861 --> 00:26:37.225
Where we had an insanely increasing lifespan, average lifespan.

490
00:26:37.405 --> 00:26:38.366
Okay.

491
00:26:39.206 --> 00:26:45.270
And one of the things that happened as a consequence was policymakers just took it for granted.

492
00:26:45.390 --> 00:26:45.550
Yep.

493
00:26:46.030 --> 00:26:46.290
Okay.

494
00:26:46.611 --> 00:26:49.753
Where, yeah, of course, we're going to just keep living longer.

495
00:26:50.633 --> 00:26:50.834
Right.

496
00:26:50.934 --> 00:26:57.899
You look at the, you look at the, the trend of like how average life expectancy is increasing and it looks like a line.

497
00:26:58.319 --> 00:27:02.522
And so like any good high school student, they get three data points.

498
00:27:02.542 --> 00:27:05.064
They're like, oh, it's just going to go that way.

499
00:27:06.425 --> 00:27:06.645
Right.

500
00:27:07.425 --> 00:27:13.790
And so, and, and this, this paper is really putting that claim to rest.

501
00:27:13.810 --> 00:27:13.970
Right.

502
00:27:14.831 --> 00:27:17.252
A lot of people have been sounding the alarm, though.

503
00:27:17.412 --> 00:27:24.397
Because anyone who's reasonable is like, okay, just because it's linear for now doesn't mean it's going to keep staying linear.

504
00:27:24.417 --> 00:27:27.759
There's plenty of systems that go linear and then they plateau off.

505
00:27:29.040 --> 00:27:31.142
But there's been sort of two camps.

506
00:27:31.582 --> 00:27:34.344
One camp has been there is kind of an inherent...

507
00:27:35.044 --> 00:27:37.447
limitation to how long humans can live.

508
00:27:37.768 --> 00:27:37.948
Okay.

509
00:27:37.968 --> 00:27:38.268
Okay.

510
00:27:38.589 --> 00:27:44.956
Based on just evolution, mutation rates, the ability of our cells and our organs to keep up.

511
00:27:45.137 --> 00:27:45.257
Yep.

512
00:27:45.437 --> 00:27:45.637
Right.

513
00:27:45.878 --> 00:27:50.503
And then there's another camp that's like, well, you know, we could have enough science to,

514
00:27:51.204 --> 00:27:52.785
To circumnavigate.

515
00:27:53.005 --> 00:27:53.365
Science.

516
00:27:53.606 --> 00:27:56.587
Like inherent evolution, right?

517
00:27:56.748 --> 00:28:03.412
And just like replace bad cells and do all of this bioengineering to just make us live forever, right?

518
00:28:03.652 --> 00:28:04.693
So those are the two camps.

519
00:28:04.753 --> 00:28:12.457
This is sort of adjacent to the transhumanism kind of movement where they're like, there's going to be the symbiosis between all of our science and stuff.

520
00:28:12.637 --> 00:28:16.240
And then that way we'll be able to be beyond, we'll be this next thing that is...

521
00:28:16.820 --> 00:28:17.580
Yeah, yeah, yeah.

522
00:28:17.620 --> 00:28:21.201
We'll merge with AI or some nonsense, right?

523
00:28:21.521 --> 00:28:25.061
And it's like, okay, yeah, maybe in like 100 years, right?

524
00:28:25.101 --> 00:28:26.582
But not me.

525
00:28:26.742 --> 00:28:27.482
I ain't doing that.

526
00:28:28.102 --> 00:28:29.202
No, ain't no way, boy.

527
00:28:29.482 --> 00:28:30.822
But these are the two camps.

528
00:28:30.842 --> 00:28:32.283
But those are the two camps, right?

529
00:28:32.423 --> 00:28:37.644
And it's been hard to sort of justify one or the other without hard data.

530
00:28:37.924 --> 00:28:41.544
And that's what this group is doing from Max Planck, right?

531
00:28:41.644 --> 00:28:44.825
The paper that's come out in PNAS, it's...

532
00:28:46.085 --> 00:28:49.690
sort of taking a very rigorous approach to this problem.

533
00:28:50.030 --> 00:28:50.271
Okay.

534
00:28:50.291 --> 00:28:55.457
Because this problem is a hard problem to get good data and good analysis on.

535
00:28:55.477 --> 00:28:55.737
Okay.

536
00:28:55.818 --> 00:28:56.018
Okay.

537
00:28:56.298 --> 00:28:59.122
There's, let me, let me give you a sense of why it's hard.

538
00:28:59.342 --> 00:29:03.868
So there's two approaches to how we quantify life expectancy.

539
00:29:04.208 --> 00:29:08.649
Okay, there's the period life expectancy and then there's the cohort life expectancy.

540
00:29:08.830 --> 00:29:14.131
Okay, the period life expectancy is the easier one to quantify.

541
00:29:14.391 --> 00:29:18.833
It's basically the idea that reflects the average mortality risk in a given year.

542
00:29:19.153 --> 00:29:19.333
Right.

543
00:29:19.413 --> 00:29:27.135
Okay, so it's kind of like saying like, okay, like if a newborn is born now and you say the life expectancy is 80 years, but that's a period life expectancy.

544
00:29:27.155 --> 00:29:32.057
What you mean is that like given the climate right now,

545
00:29:33.245 --> 00:29:38.548
This newborn, given the mortality climate right now, this newborn would be expected to live 80 years, right?

546
00:29:38.568 --> 00:29:41.429
Given the chances of dying on a random person.

547
00:29:41.569 --> 00:29:41.729
Yep.

548
00:29:41.769 --> 00:29:41.949
Right?

549
00:29:42.009 --> 00:29:45.911
But that is a very local, local in time measure.

550
00:29:46.131 --> 00:29:46.292
Right.

551
00:29:46.352 --> 00:29:46.552
Right?

552
00:29:46.612 --> 00:29:50.934
Because if the COVID pandemic happens, that period life expectancy goes down.

553
00:29:50.974 --> 00:29:51.114
Right.

554
00:29:51.454 --> 00:29:56.643
If the Spanish flu happens in 1918, the period life expectancy goes down.

555
00:29:56.943 --> 00:30:00.089
But, like, it's kind of a live metric of life expectancy, okay?

556
00:30:00.409 --> 00:30:04.336
Then there's something called the cohort life expectancy, which is, like, saying, okay...

557
00:30:05.905 --> 00:30:12.674
The people born in 1950 in that birth cohort, the average lifespan of that individual is 80 years.

558
00:30:13.174 --> 00:30:13.455
Okay?

559
00:30:14.015 --> 00:30:21.565
That's different because that is taking into account this giant time span of 80, like 1950 plus whatever 80 years.

560
00:30:21.645 --> 00:30:21.845
Yes.

561
00:30:21.865 --> 00:30:22.105
Right?

562
00:30:22.226 --> 00:30:22.786
Yes.

563
00:30:23.467 --> 00:30:28.954
In principle, that's really hard to quantify unless everyone has died from 1950.

564
00:30:28.994 --> 00:30:29.535
Right.

565
00:30:29.715 --> 00:30:29.895
Right?

566
00:30:30.136 --> 00:30:35.282
Then you take everyone who died from 1950, you average it out, and you say that was the cohort life expectancy.

567
00:30:35.302 --> 00:30:40.309
So it can be useful in a retrospective context when you have a generation or cohort that's already...

568
00:30:40.909 --> 00:30:42.230
Yeah, that's already done.

569
00:30:42.491 --> 00:30:43.651
But it's not predicting.

570
00:30:43.731 --> 00:30:44.132
It's hard.

571
00:30:44.152 --> 00:30:48.615
You can't play it forward as like, well, we have a generation that's like halfway.

572
00:30:48.635 --> 00:30:49.095
Yeah.

573
00:30:49.536 --> 00:30:49.716
Yeah.

574
00:30:49.816 --> 00:30:51.097
So like maybe predict.

575
00:30:51.317 --> 00:30:52.318
Yeah, exactly.

576
00:30:52.338 --> 00:30:53.639
These unknowns about the future.

577
00:30:53.759 --> 00:30:54.339
Exactly.

578
00:30:54.479 --> 00:30:54.700
Yeah.

579
00:30:54.760 --> 00:30:55.280
Yeah.

580
00:30:55.460 --> 00:30:57.101
So you see now the struggle, right?

581
00:30:57.121 --> 00:31:02.045
That like scientists have in defining a cohort life expectancy of a generation that's still alive.

582
00:31:02.265 --> 00:31:02.465
Yes.

583
00:31:02.545 --> 00:31:04.447
And that's what these guys are trying to forecast.

584
00:31:04.467 --> 00:31:04.567
Yeah.

585
00:31:04.707 --> 00:31:04.947
Okay.

586
00:31:05.067 --> 00:31:05.308
Okay.

587
00:31:05.628 --> 00:31:07.870
And so the way they did it was really cool.

588
00:31:08.070 --> 00:31:09.931
They got six different types of models.

589
00:31:10.432 --> 00:31:18.419
So, you know, in social science, if you just do like a single model, lots of confounds, you want to do a bunch.

590
00:31:18.439 --> 00:31:19.720
That's why I like this paper, right?

591
00:31:19.960 --> 00:31:21.241
They were pretty rigorous about it.

592
00:31:21.901 --> 00:31:26.005
You do six different, different approaches, and you see if they all give the same answer.

593
00:31:26.225 --> 00:31:28.187
If they do, then it's a pretty robust finding.

594
00:31:28.347 --> 00:31:28.607
Got it.

595
00:31:28.727 --> 00:31:28.907
Right.

596
00:31:29.227 --> 00:31:35.510
On the other hand, if it's a giant spread, that means that the nitty gritties of the model and the initial data that you put in, that's what's messing it up.

597
00:31:35.550 --> 00:31:35.730
Right.

598
00:31:36.091 --> 00:31:44.695
So they did six different models, like the Lee Carter model, the cohort segmented transformation, age at death model, all of these like different kinds of models.

599
00:31:44.715 --> 00:31:50.898
They're basically like fancy data analysis techniques where what you do is you take a time series of data.

600
00:31:51.518 --> 00:31:55.139
of different age groups and how they're doing in terms of mortality.

601
00:31:55.459 --> 00:32:01.161
And then from that you project forward and try to get like a sense of, okay, when is the time series going to end?

602
00:32:01.521 --> 00:32:01.741
Okay.

603
00:32:02.101 --> 00:32:03.721
You try to do this for six different things.

604
00:32:04.041 --> 00:32:12.304
And what they came up with the same, they came up with the same endpoint, which is that this life expectancy is going up, but it's not going to get to a hundred.

605
00:32:12.384 --> 00:32:13.184
It's going to plateau.

606
00:32:13.744 --> 00:32:13.984
Okay.

607
00:32:14.204 --> 00:32:17.145
To validate it, you might be thinking, okay, well, these are just models, right?

608
00:32:17.285 --> 00:32:17.685
It's nice.

609
00:32:17.985 --> 00:32:24.807
To validate it, they forecasted life expectancy for people from 1919 to 1938 who are all dead.

610
00:32:25.267 --> 00:32:25.487
Okay?

611
00:32:25.507 --> 00:32:27.448
Everyone born in that time period are all dead.

612
00:32:27.768 --> 00:32:32.389
So you can take that data, right, as your test data in some sense, right?

613
00:32:32.629 --> 00:32:41.832
And be like, okay, I'm going to do the same analysis for this data from people born in 1919 to 1938 and then try to project when they're going to die.

614
00:32:42.012 --> 00:32:42.632
And it worked.

615
00:32:42.772 --> 00:32:43.732
And we have the answer for that.

616
00:32:43.752 --> 00:32:46.053
And we have the answer for that because we have the data about when they died.

617
00:32:46.133 --> 00:32:50.775
So we built a sort of predictive model to be able to work for current cohorts that are still alive.

618
00:32:50.815 --> 00:32:51.056
Yeah.

619
00:32:51.276 --> 00:32:52.296
And we validated it.

620
00:32:52.376 --> 00:32:54.197
And it worked across six models.

621
00:32:54.637 --> 00:33:03.122
And then we used a historical cohort who, where everyone's dead to put it in to see if it mapped onto the actual results we see saw.

622
00:33:03.442 --> 00:33:05.063
from that cohort that's now all.

623
00:33:05.203 --> 00:33:11.168
Yeah, and it was a little bit off, but not enough to justify what they were seeing with the present cohort.

624
00:33:11.208 --> 00:33:20.435
And it might be a little bit off because like, you know, when you're trying to do, like obviously the environment of the data back then is different from the environment now, right?

625
00:33:20.675 --> 00:33:23.837
But then you can make arguments about how off the results should be.

626
00:33:24.257 --> 00:33:29.201
And the results should not be that off compared to what we're finding now, which is that this thing is definitely plateauing.

627
00:33:30.102 --> 00:33:30.342
Okay?

628
00:33:30.462 --> 00:33:31.363
Yes, yes.

629
00:33:32.123 --> 00:33:38.549
The cooler part of this is they actually did a kind of causation of what is causing this plateau.

630
00:33:38.689 --> 00:33:38.909
Right.

631
00:33:39.289 --> 00:33:48.117
And what's causing this plateau is the fact that we have exhausted all the low-hanging fruits when it comes to longevity of human lifespan.

632
00:33:48.297 --> 00:33:53.220
We've taken all the low branches on the tree for extending life, and we've actualized.

633
00:33:53.300 --> 00:33:56.982
We've actualized them all because young people are now living longer, right?

634
00:33:57.262 --> 00:34:02.005
No one's, I mean, people, obviously people are dying before the age of five, but much less.

635
00:34:02.305 --> 00:34:04.747
Infant mortality is way down, okay?

636
00:34:04.967 --> 00:34:10.610
People dying before the age of 20 is way down because infectious diseases are way down, right?

637
00:34:10.670 --> 00:34:13.752
Like the problems that you have as a young person.

638
00:34:14.710 --> 00:34:18.437
are no longer the kinds of problems that you used to have before.

639
00:34:18.858 --> 00:34:19.159
Okay?

640
00:34:19.580 --> 00:34:26.673
Like, because of vaccinations, because of fundamental science, because of all of the advancement that we've been doing over the past century.

641
00:34:27.768 --> 00:34:31.052
The hard stuff is now the actual aging.

642
00:34:31.893 --> 00:34:32.233
Okay?

643
00:34:32.573 --> 00:34:33.955
That's the cancer.

644
00:34:34.375 --> 00:34:34.536
Right.

645
00:34:34.556 --> 00:34:37.559
That's the Alzheimer's.

646
00:34:37.579 --> 00:34:43.085
That's these higher order things that are notoriously hard.

647
00:34:43.325 --> 00:34:44.266
Because they're no longer...

648
00:34:44.967 --> 00:34:48.149
They're no longer agents that are trying to kill you.

649
00:34:48.550 --> 00:34:48.830
Okay?

650
00:34:48.870 --> 00:34:49.610
They're yourself.

651
00:34:50.211 --> 00:34:55.635
That's just like sort of at a cellular level giving up in some sense.

652
00:34:55.695 --> 00:34:55.875
Right?

653
00:34:55.955 --> 00:34:57.416
Like the neurons are sort of dying.

654
00:34:57.536 --> 00:34:57.656
Right.

655
00:34:57.716 --> 00:35:02.420
And the nervous system and the brain is in some sense deteriorating just because of...

656
00:35:03.100 --> 00:35:03.881
Like time.

657
00:35:04.061 --> 00:35:04.241
Right.

658
00:35:04.401 --> 00:35:04.601
Right.

659
00:35:04.922 --> 00:35:11.827
And in cancer cells, it's like the mechanisms that control the growth of cells is sort of giving up.

660
00:35:11.987 --> 00:35:12.128
Right.

661
00:35:12.148 --> 00:35:13.709
And then now you have these tumors growth.

662
00:35:13.789 --> 00:35:13.989
Right.

663
00:35:14.309 --> 00:35:17.232
So so it's a much more complicated problem now.

664
00:35:17.352 --> 00:35:17.532
Right.

665
00:35:17.972 --> 00:35:18.133
That.

666
00:35:18.833 --> 00:35:30.719
We haven't really had the kind of progress that we have had in cases like polio and smallpox and all of these other sort of, you know, quote, easier problems.

667
00:35:30.959 --> 00:35:32.159
They weren't easy at the time.

668
00:35:32.179 --> 00:35:35.861
But now that we've looking in hindsight, it's like, OK, they were all like agents.

669
00:35:36.201 --> 00:35:38.923
And now we have these like modular mechanisms to like solve that.

670
00:35:38.963 --> 00:35:39.143
Right.

671
00:35:39.203 --> 00:35:40.403
Oh, just make a vaccine.

672
00:35:40.423 --> 00:35:41.044
Coronavirus.

673
00:35:41.084 --> 00:35:42.604
Like within six months, we had a vaccine.

674
00:35:43.145 --> 00:35:43.445
Easy.

675
00:35:43.905 --> 00:35:47.847
Now we have to solve for our own body trying to kill us over time.

676
00:35:48.327 --> 00:36:06.314
And the open-ended issue with this is it's not clear that in the short or midterm that we're going to actually make this sufficient or significant progress that would lead to outcomes that were at the rate of progress.

677
00:36:06.374 --> 00:36:06.854
Exactly.

678
00:36:07.114 --> 00:36:08.575
Yeah, yeah, yeah.

679
00:36:08.695 --> 00:36:16.138
It doesn't seem that like it's not going to – we're not making progress at the speed that would like temper this plateau.

680
00:36:16.378 --> 00:36:25.744
You know what this reminds me of, funny enough, is in technology, there's this, you know, markets and cycles are defined by this S-curve, right?

681
00:36:25.784 --> 00:36:26.024
Yeah.

682
00:36:26.184 --> 00:36:29.906
Where you start, you know, with a low rate of change and then you have this explosive growth.

683
00:36:30.707 --> 00:36:33.469
And with every technology cycle, eventually it tapers off.

684
00:36:33.529 --> 00:36:35.610
Mobile phones is the perfect example, right?

685
00:36:35.650 --> 00:36:39.272
We went through years of Apple innovating and doing new things and combining all this stuff.

686
00:36:39.552 --> 00:36:43.575
And then all of a sudden everyone catches up and then we kind of get to the place where we are now where...

687
00:36:44.175 --> 00:36:50.460
Is a foldable versus, is that really, you know, or are we now just kind of making things for marketing?

688
00:36:50.580 --> 00:36:50.760
Yeah.

689
00:36:50.801 --> 00:36:55.464
And so this similar idea of sort of the S curves that you see in technology cycles.

690
00:36:56.005 --> 00:36:58.387
Yeah, it's happening in fundamental science when it comes to biology.

691
00:36:58.407 --> 00:37:02.290
When it comes to this longevity issue, it's like we kind of have reached the end of the first S curve.

692
00:37:02.430 --> 00:37:02.650
Yes.

693
00:37:02.750 --> 00:37:04.992
And so in technology, the idea is S-curves move, right?

694
00:37:05.012 --> 00:37:07.434
So it's like you first have like vacuum tubes, all this stuff.

695
00:37:07.695 --> 00:37:09.156
Then you get to microprocessors.

696
00:37:09.216 --> 00:37:11.398
Then you move to mobile, PC, mobile, et cetera.

697
00:37:12.038 --> 00:37:16.242
The open question is arguably we just went through the first, let's say, longevity cycle.

698
00:37:16.262 --> 00:37:16.502
Yes.

699
00:37:16.522 --> 00:37:17.103
S-curves cycle.

700
00:37:17.163 --> 00:37:17.383
Yes.

701
00:37:17.783 --> 00:37:18.624
And it's not clear.

702
00:37:19.945 --> 00:37:25.068
what the second S-curve cycle may or may not look like, and whether we can traverse it or not.

703
00:37:25.128 --> 00:37:25.868
Yes, exactly.

704
00:37:26.169 --> 00:37:30.371
And if we get to that S-curve, what is the real upper limit?

705
00:37:30.451 --> 00:37:35.053
Because right now, the upper limit, I think the record is 122 years for someone to live.

706
00:37:35.394 --> 00:37:35.654
Okay?

707
00:37:36.154 --> 00:37:36.354
And...

708
00:37:38.015 --> 00:37:39.196
Is that the real limit?

709
00:37:39.476 --> 00:37:39.756
Right.

710
00:37:39.916 --> 00:37:40.996
Is it at that point?

711
00:37:41.096 --> 00:37:42.297
The body is just like, nah.

712
00:37:42.457 --> 00:37:42.677
Right.

713
00:37:43.097 --> 00:37:43.277
Right.

714
00:37:44.418 --> 00:37:48.980
Evolution has literally programmed us to be like, no, no, no, we're still evolving.

715
00:37:49.100 --> 00:37:49.300
Right.

716
00:37:49.340 --> 00:37:54.122
The human genome has a really high mutation rate compared to like, like, for example, sharks.

717
00:37:54.222 --> 00:37:54.442
Right.

718
00:37:54.522 --> 00:38:01.585
Sharks have like a insanely low mutation rate, which is why their fossils 300, 300 million years ago look exactly the same.

719
00:38:01.665 --> 00:38:01.825
Right.

720
00:38:01.945 --> 00:38:05.167
There's no humans like even 10 million years ago that look like us.

721
00:38:05.528 --> 00:38:07.509
So the mutation rate is really high.

722
00:38:07.869 --> 00:38:12.513
On top of that, we're incredibly agile creatures.

723
00:38:13.273 --> 00:38:17.957
We're mammals at a high body temperature, high metabolism.

724
00:38:18.097 --> 00:38:24.521
So all of these things sort of start contributing to tempering an infinite lifespan.

725
00:38:24.561 --> 00:38:24.742
Right.

726
00:38:24.782 --> 00:38:25.302
The limits.

727
00:38:25.762 --> 00:38:30.626
We have these limits that exist that are fundamental to the existence of our being.

728
00:38:30.646 --> 00:38:30.766
Right.

729
00:38:31.306 --> 00:38:38.291
And it is not clear that they are necessarily surmountable by a human manufactured solution in human time spans.

730
00:38:38.411 --> 00:38:38.851
Exactly.

731
00:38:38.871 --> 00:38:39.071
Yeah.

732
00:38:39.371 --> 00:38:54.321
And what I liked, I mean, I guess not what I like about this story, but what concerns me about this story is that I think we need to, again, start taking this very seriously as a policymaking substrate.

733
00:38:54.401 --> 00:38:54.601
Right.

734
00:38:54.861 --> 00:38:55.101
Right.

735
00:38:55.581 --> 00:39:03.285
As like now a scientific fact that we have learned and urgently need to act on.

736
00:39:03.645 --> 00:39:03.925
Okay.

737
00:39:04.506 --> 00:39:12.990
Because from a policy perspective, I mean, to extrapolate this, there's huge arguments right now over the debt in the U.S. Yeah.

738
00:39:13.390 --> 00:39:15.231
Being at untenable levels.

739
00:39:15.711 --> 00:39:18.933
And one of the biggest line items in that is Social Security.

740
00:39:19.013 --> 00:39:19.233
Yeah.

741
00:39:20.153 --> 00:39:23.735
And it makes assumptions about retirement age and who gets Medicare.

742
00:39:24.155 --> 00:39:29.838
And what happens if you're the line of average life expectancy shifts or moves.

743
00:39:29.858 --> 00:39:30.138
Yeah.

744
00:39:30.478 --> 00:39:31.198
It's no longer...

745
00:39:31.238 --> 00:39:36.921
I mean, everybody has made this policy based on an increasing life expectancy forever.

746
00:39:36.961 --> 00:39:40.943
And now this paper is saying that actually, no, life expectancy has plateaued.

747
00:39:41.343 --> 00:39:44.324
We're at the end of the road when it comes to how old we can get.

748
00:39:44.684 --> 00:39:44.864
Right?

749
00:39:45.064 --> 00:39:45.625
Like all these...

750
00:39:46.885 --> 00:39:57.488
With pension funds, with social security systems, there was a riot in France the other year when the government tried to raise the retirement age.

751
00:39:57.748 --> 00:40:06.030
The reason why they're raising the retirement age is a pretty simple logic, which is if people are living longer, then they can work longer.

752
00:40:06.470 --> 00:40:09.971
And if they can work longer, we should make them work longer before they get their retirement age.

753
00:40:10.111 --> 00:40:11.574
Well, this isn't the case anymore.

754
00:40:11.594 --> 00:40:13.318
People aren't actually living longer.

755
00:40:13.578 --> 00:40:16.605
They're living the same as they were like 10 years ago.

756
00:40:16.725 --> 00:40:16.885
Right.

757
00:40:17.126 --> 00:40:17.366
Right.

758
00:40:17.647 --> 00:40:18.950
We're not actually making it...

759
00:40:21.515 --> 00:40:22.056
Right.

760
00:40:22.276 --> 00:40:32.085
And the worst part about these kinds of policies that sort of raise the retirement age is it actually impacts those of worse socioeconomic background.

761
00:40:32.125 --> 00:40:32.305
Right.

762
00:40:32.505 --> 00:40:34.587
Because the rich people, rich people can just retire.

763
00:40:34.647 --> 00:40:34.787
Right.

764
00:40:34.927 --> 00:40:36.088
You could retire at 50.

765
00:40:36.108 --> 00:40:37.429
You probably already did.

766
00:40:37.469 --> 00:40:37.609
Right.

767
00:40:37.770 --> 00:40:41.173
It's the poor people that are still going to have to work until like 70 now.

768
00:40:41.813 --> 00:40:44.235
And and then what die at like 72.

769
00:40:44.315 --> 00:40:44.516
Right.

770
00:40:44.716 --> 00:40:44.896
Like.

771
00:40:45.136 --> 00:40:45.997
You know, hundred.

772
00:40:46.077 --> 00:40:52.100
No, it's a really, it's a really important point about how a lot of this very fundamental science research.

773
00:40:53.311 --> 00:41:00.914
historically, especially in the United States, has both informed like military and intelligence posture as well as policy-making posture.

774
00:41:00.934 --> 00:41:01.494
That's right, yeah.

775
00:41:01.835 --> 00:41:21.863
And there's just been this disconnect where for whatever reason, there's no longer the through lines from fundamental research and the insights that are gleaned from that into bleeding into these areas that are in the control of nation states in terms of how do you sort of organize society around it.

776
00:41:22.103 --> 00:41:22.643
Yeah, dude.

777
00:41:22.663 --> 00:41:30.887
This is like a meaningful and important point because people do still generally do still think like, oh, yeah, we're just going to live longer than our parents.

778
00:41:31.327 --> 00:41:34.689
Much like every other aspect, at least in the U.S. of life, right?

779
00:41:34.749 --> 00:41:36.890
We are less like the millennials and younger.

780
00:41:37.090 --> 00:41:37.330
Yeah.

781
00:41:37.430 --> 00:41:39.391
We're less well off financially than our parents.

782
00:41:39.411 --> 00:41:39.511
Yeah.

783
00:41:39.871 --> 00:41:42.372
Everything is less affordable than it has been for our parents.

784
00:41:42.392 --> 00:41:43.673
And we're the first generation.

785
00:41:45.153 --> 00:41:47.154
to be less well off than our parents.

786
00:41:47.414 --> 00:41:55.338
And so the idea that we're going to also now live just as long makes it like you can kind of see like all these things are like aligning.

787
00:41:55.378 --> 00:41:55.538
Yeah.

788
00:41:55.618 --> 00:41:58.920
And then and then when they when when, you know, our parents generation retires,

789
00:42:00.607 --> 00:42:05.808
The population pyramid is now basically inverted for all of these developing countries where this is true.

790
00:42:05.908 --> 00:42:06.268
Right.

791
00:42:06.288 --> 00:42:06.949
The Japan problem.

792
00:42:06.989 --> 00:42:07.169
Right.

793
00:42:07.209 --> 00:42:13.170
Like it's it's now there's fewer and fewer young people to support this older population for Social Security.

794
00:42:13.510 --> 00:42:19.911
Turns out, given that population pyramid and the mathematics of just two plus two equals four and it's never five.

795
00:42:20.031 --> 00:42:20.211
Right.

796
00:42:20.271 --> 00:42:22.372
Like this might just be a scam.

797
00:42:23.072 --> 00:42:24.413
Compound on top of that.

798
00:42:24.473 --> 00:42:24.673
Right.

799
00:42:24.713 --> 00:42:30.556
This idea that like actually the fundamental premise that like the younger generation is going to live longer.

800
00:42:30.696 --> 00:42:30.876
Yes.

801
00:42:30.976 --> 00:42:32.637
And so can contribute more.

802
00:42:32.737 --> 00:42:32.917
Right.

803
00:42:33.097 --> 00:42:34.758
To the social security pile.

804
00:42:34.898 --> 00:42:35.038
Yes.

805
00:42:35.158 --> 00:42:36.439
Is now no longer correct.

806
00:42:36.499 --> 00:42:36.659
Right.

807
00:42:36.879 --> 00:42:46.384
And we need to start asking questions like is this like it's there's a fundamental shift now I think in my mind from is this a tenable strategy.

808
00:42:46.644 --> 00:42:46.824
Right.

809
00:42:46.904 --> 00:42:51.546
To is this even a correct way to do business in a developed society.

810
00:42:51.586 --> 00:42:51.766
Right.

811
00:42:52.087 --> 00:42:52.247
Right.

812
00:42:52.327 --> 00:42:52.547
Right.

813
00:42:52.687 --> 00:42:52.827
Right.

814
00:42:52.987 --> 00:42:56.989
I don't want to sound communist because I swear I am not a communist.

815
00:42:57.009 --> 00:42:57.369
Okay.

816
00:42:57.890 --> 00:43:05.473
But like at some point you got to start asking questions about fundamental things like the social construct.

817
00:43:05.873 --> 00:43:06.034
Right.

818
00:43:06.074 --> 00:43:08.135
Like is social contract.

819
00:43:08.175 --> 00:43:08.475
Sorry.

820
00:43:08.535 --> 00:43:08.715
Yes.

821
00:43:08.795 --> 00:43:10.336
Like, you know, is this true?

822
00:43:10.396 --> 00:43:12.617
Like this is the social contract of.

823
00:43:13.317 --> 00:43:16.660
the post-war era still applicable today.

824
00:43:16.820 --> 00:43:16.980
Right.

825
00:43:17.341 --> 00:43:20.904
And the fundamental variables in that equation have changed so drastically.

826
00:43:21.224 --> 00:43:29.051
And we're still saying the outcomes are going to be the same, even though all the variables on the other side of the equation are fundamentally very, very different.

827
00:43:29.091 --> 00:43:29.472
Yeah, yeah.

828
00:43:29.532 --> 00:43:33.235
Like, this is a reality now, and it's, like, going to catch up with us as we grow older.

829
00:43:33.255 --> 00:43:33.355
Yeah.

830
00:43:33.795 --> 00:43:35.495
This is a really, really important story.

831
00:43:35.515 --> 00:44:01.320
I'm sure we're going to come back to the subject of longevity because I think the nexus you brought up from longevity studies to policy is a really important note, especially when we're seeing not only in Western, let's say, quote-unquote, Western developed nations, but any developed nation, this similar issue of generational friction that arises from 50-year-old to 80-year-old people.

832
00:44:02.840 --> 00:44:13.706
policy conceptions that are based on fundamentals that are now statistically, data wise and scientifically are proving to not necessarily be the same.

833
00:44:14.346 --> 00:44:18.148
And yet we've not changed any of our posture.

834
00:44:18.168 --> 00:44:18.768
Yeah, yeah, yeah.

835
00:44:18.788 --> 00:44:20.909
And then we're just going to run into a brick wall at some point.

836
00:44:20.949 --> 00:44:22.550
And this is it's a problem.

837
00:44:22.810 --> 00:44:23.451
Yeah.

838
00:44:23.771 --> 00:44:24.292
Yang gang.

839
00:44:24.352 --> 00:44:27.896
Yeah.

840
00:44:28.076 --> 00:44:31.760
It's interesting because it does dovetail somewhat to our third story.

841
00:44:31.840 --> 00:44:38.508
One aspect we didn't talk about the factors of longevity that changes, you know, climate is a big part of that.

842
00:44:38.648 --> 00:44:39.389
It's getting bad.

843
00:44:39.929 --> 00:44:44.894
And our next story is about a new no-sort plastic recycling being near.

844
00:44:44.934 --> 00:44:51.100
The headline, New Nickel Catalyst Enables No-Sort Mixed Plastic Recycling.

845
00:44:51.420 --> 00:44:56.605
This is a new, brand new breaking research coming out of Northwestern University.

846
00:44:56.905 --> 00:44:58.767
They put their paper out in Nature Chemistry.

847
00:44:59.007 --> 00:44:59.508
Yeah.

848
00:45:00.109 --> 00:45:09.400
I just went to Japan earlier this year, and one of the interesting things about in Japan, unlike in the US, is they have different recycling bins for different types of recycling, glass, cardboard, et cetera.

849
00:45:09.420 --> 00:45:11.062
In the US, we just put it all in the same thing.

850
00:45:11.623 --> 00:45:17.129
So my assumption is this no-sort plastic is saying like there's no need to,

851
00:45:18.130 --> 00:45:18.671
X, Y, Z.

852
00:45:19.211 --> 00:45:26.675
This seems like it's on the sort of pollution and our need to be more effective with our resource users.

853
00:45:26.935 --> 00:45:28.096
This might have impact.

854
00:45:28.116 --> 00:45:29.497
But tell me what we need to know about this story.

855
00:45:29.657 --> 00:45:30.377
Yeah.

856
00:45:30.538 --> 00:45:31.078
Plastic.

857
00:45:31.138 --> 00:45:31.618
We love it.

858
00:45:31.838 --> 00:45:32.659
Yes, we do love it.

859
00:45:32.719 --> 00:45:33.559
We do love it.

860
00:45:33.899 --> 00:45:35.380
As humans, we are cheap.

861
00:45:36.901 --> 00:45:38.482
And we want things.

862
00:45:39.443 --> 00:45:40.924
We want things to be easy.

863
00:45:41.384 --> 00:45:42.545
We want them to be cheap.

864
00:45:43.026 --> 00:45:45.007
And so we love plastic.

865
00:45:45.347 --> 00:45:45.607
Okay.

866
00:45:45.908 --> 00:45:49.310
Plastic is really bad for the environment.

867
00:45:49.510 --> 00:45:52.492
It sticks around for like millions of years or something like that.

868
00:45:52.532 --> 00:45:52.712
Right.

869
00:45:53.873 --> 00:45:55.574
And we are producing a lot of it.

870
00:45:55.614 --> 00:45:57.956
Something like 400 million tons a year.

871
00:45:59.898 --> 00:46:02.679
Something like 11 million tons go into the ocean.

872
00:46:03.300 --> 00:46:05.582
Which, you know, I'm very...

873
00:46:06.402 --> 00:46:07.444
I really don't like that.

874
00:46:07.464 --> 00:46:08.005
Okay.

875
00:46:08.365 --> 00:46:11.670
But I'm also a lazy, selfish American.

876
00:46:12.912 --> 00:46:13.613
And...

877
00:46:15.659 --> 00:46:18.280
I like my things, right?

878
00:46:18.500 --> 00:46:25.563
I like my mustard and ketchup bottles in plastic and my milk jug that comes in a gallon in plastic.

879
00:46:25.603 --> 00:46:29.024
I don't want to go with a glass bottle and refill the same milk jug.

880
00:46:29.064 --> 00:46:38.348
It's just so convenient to just go to the store and grab a new thing and then throw away the old one and just not worry about what the old one is doing, right?

881
00:46:38.988 --> 00:46:42.490
So as Americans, we are very selfish and we're very lazy.

882
00:46:43.411 --> 00:46:46.833
And, you know, it's not just something that Americans are.

883
00:46:46.853 --> 00:46:48.254
It's something that the whole world is.

884
00:46:48.294 --> 00:46:49.475
This is a world problem.

885
00:46:50.055 --> 00:46:54.118
And, you know, everyone who says that, like, we need to start using less.

886
00:46:55.238 --> 00:47:01.502
It's just simply not a practical way to save the planet.

887
00:47:01.843 --> 00:47:02.143
Okay?

888
00:47:02.163 --> 00:47:02.303
Mm-hmm.

889
00:47:02.703 --> 00:47:07.028
Like, we love our planet, but we love ourselves and our comfort way more.

890
00:47:07.168 --> 00:47:08.629
And that is just not going to change.

891
00:47:08.910 --> 00:47:09.190
Okay?

892
00:47:09.610 --> 00:47:13.014
Even the developing, especially the developing countries, right?

893
00:47:13.314 --> 00:47:17.578
Like China and India and the continent of Africa that are...

894
00:47:18.239 --> 00:47:23.146
also contributing immensely to this chaos of plastic and consumables and all this stuff.

895
00:47:23.426 --> 00:47:26.571
They're looking at the developed countries being like, no, you guys did that.

896
00:47:26.871 --> 00:47:29.194
Well, you guys just had 50 years of doing that.

897
00:47:29.294 --> 00:47:30.396
And now we're just starting.

898
00:47:30.516 --> 00:47:32.198
And now it's like, oh, I'm sorry.

899
00:47:32.258 --> 00:47:33.240
I can't use coal.

900
00:47:34.501 --> 00:47:41.726
You know, when that's the one resource that's super cheap to use to make electricity, which is something I need so I can like educate my kids.

901
00:47:41.866 --> 00:47:42.586
Yeah.

902
00:47:42.926 --> 00:47:49.310
So this idea that like, you know, using humans doing less to save the planet is not going to work.

903
00:47:49.410 --> 00:47:49.570
Yeah.

904
00:47:49.810 --> 00:47:50.091
Okay.

905
00:47:50.351 --> 00:47:57.055
The solution that is going to work is scientific innovation to make our current lifestyle better for the planet.

906
00:47:57.075 --> 00:47:57.875
And more sustainable.

907
00:47:57.975 --> 00:47:59.076
More sustainable for the planet.

908
00:47:59.536 --> 00:48:00.757
I know this is a hot take.

909
00:48:00.917 --> 00:48:01.197
I know.

910
00:48:01.337 --> 00:48:05.718
About like how we need to deal with human directed climate change because it is real.

911
00:48:06.038 --> 00:48:06.199
Right.

912
00:48:06.499 --> 00:48:07.939
Humans are changing the planet.

913
00:48:07.999 --> 00:48:08.179
Yes.

914
00:48:08.319 --> 00:48:11.200
But my fundamental take is like we're not going to stop.

915
00:48:11.841 --> 00:48:18.083
I mean, at the end of the day, the idea that, you know, people using paper straws instead of plastic.

916
00:48:18.883 --> 00:48:30.212
But then every corporation and manufacturing entity and these large people who are actually producing the largest volume of the problem is really where the problem is.

917
00:48:30.232 --> 00:48:31.292
Yeah, yeah.

918
00:48:31.432 --> 00:48:31.733
Exactly.

919
00:48:31.753 --> 00:48:34.074
In terms of having the most impact for our efforts.

920
00:48:34.154 --> 00:48:34.955
Yes, yes.

921
00:48:34.995 --> 00:48:36.796
We need a bulk strategy.

922
00:48:36.957 --> 00:48:37.137
Right.

923
00:48:37.197 --> 00:48:45.043
Because even if you made every human, like individual human, like feel bad, corporations who are apparently also people –

924
00:48:46.003 --> 00:48:49.166
They're not going to feel bad because the mob is crazy, right?

925
00:48:49.846 --> 00:48:50.467
No, exactly.

926
00:48:50.587 --> 00:48:53.249
And so, no, they're going to be like, oh, yeah, just dump it right there.

927
00:48:53.269 --> 00:48:53.689
It's fine.

928
00:48:53.790 --> 00:48:53.970
Exactly.

929
00:48:53.990 --> 00:48:54.650
You know?

930
00:48:54.770 --> 00:49:00.935
So we need a strategy from fundamental science perspective that helps us recycle efficiently and in a better way.

931
00:49:01.015 --> 00:49:01.115
Yes.

932
00:49:01.135 --> 00:49:01.316
Okay?

933
00:49:01.616 --> 00:49:06.460
Plastic recycling is a key problem that we as humans, I think we need to solve.

934
00:49:06.580 --> 00:49:06.720
Yes.

935
00:49:06.800 --> 00:49:07.060
Okay?

936
00:49:07.461 --> 00:49:09.822
And this is a really big step in that direction.

937
00:49:09.842 --> 00:49:09.983
Okay.

938
00:49:10.003 --> 00:49:10.103
Okay?

939
00:49:10.123 --> 00:49:10.303
Okay.

940
00:49:10.663 --> 00:49:11.804
So let's talk about plastic.

941
00:49:11.824 --> 00:49:12.705
Plastic is really bad.

942
00:49:12.745 --> 00:49:17.030
I just told you like something like 11 million tons go into the ocean every year.

943
00:49:17.410 --> 00:49:19.432
It's going to just stay there for thousands of years.

944
00:49:19.513 --> 00:49:20.594
It's not going to decompose.

945
00:49:20.954 --> 00:49:23.256
Plastic is essentially chemically.

946
00:49:23.397 --> 00:49:26.620
It's just polymer chains of hydrocarbon.

947
00:49:26.660 --> 00:49:32.506
So carbon, hydrogen, hydrogen, hydrogen, carbon, hydrogen, hydrogen, hydrogen, just chains of carbon and hydrogen.

948
00:49:32.606 --> 00:49:32.847
Okay.

949
00:49:33.307 --> 00:49:34.729
Like that's all plastic is.

950
00:49:34.889 --> 00:49:39.576
It comes from like petroleum, which is also hydrocarbons.

951
00:49:39.976 --> 00:49:44.423
So another like, you know, it's another use of petroleum in that case.

952
00:49:44.623 --> 00:49:48.528
And in order to recycle it, there's basically like a few ways that we do it right now.

953
00:49:48.548 --> 00:49:49.650
And they're all pretty bad.

954
00:49:50.090 --> 00:49:50.291
Okay.

955
00:49:50.731 --> 00:49:58.359
One of them is the thermomechanical method, which is what you got to do is you got to basically materially degrade this thing.

956
00:50:00.141 --> 00:50:02.104
You've got to like heat it up.

957
00:50:02.564 --> 00:50:09.912
You've got to make it into like lower, worse plastic and then use that plastic somehow for some other purpose.

958
00:50:10.052 --> 00:50:10.333
Okay.

959
00:50:10.813 --> 00:50:12.154
It's a pretty bad way of doing things.

960
00:50:13.075 --> 00:50:14.557
But we do it 9%.

961
00:50:14.737 --> 00:50:17.440
9% of the plastic globally is done this way.

962
00:50:18.381 --> 00:50:19.362
It's really expensive.

963
00:50:19.402 --> 00:50:26.449
Nobody really wants to do it because you've got to sort all these plastics for all their different types of plastic.

964
00:50:26.469 --> 00:50:26.590
Right.

965
00:50:27.070 --> 00:50:30.512
Because they break down to different fundamental parts and you don't necessarily want those mixed.

966
00:50:30.552 --> 00:50:30.732
Yeah.

967
00:50:30.752 --> 00:50:34.314
And then if they mix, then they make even worse stuff.

968
00:50:34.334 --> 00:50:35.955
And then they just go to the landfill.

969
00:50:35.975 --> 00:50:37.536
And it's like, okay, what did you even do?

970
00:50:37.676 --> 00:50:38.817
We just wasted a bunch of money.

971
00:50:38.837 --> 00:50:40.698
You just, yeah.

972
00:50:40.718 --> 00:50:41.358
So that's one way.

973
00:50:41.398 --> 00:50:45.461
The other way is pyrolysis, which is just heat it up.

974
00:50:46.081 --> 00:50:47.462
Heat it up, degrade this thing.

975
00:50:47.782 --> 00:50:55.887
And then you get some kind of low value type of GHG, which is like some kind of like low value fuel.

976
00:50:56.067 --> 00:50:56.267
Right?

977
00:50:56.507 --> 00:50:57.587
The problem with that is, okay, fine.

978
00:50:57.607 --> 00:51:00.628
Like, how much energy did you use to, like, heat this thing up?

979
00:51:01.068 --> 00:51:02.808
To get a low value fuel.

980
00:51:03.068 --> 00:51:05.969
You're getting, it's like nuclear fusion, right?

981
00:51:06.049 --> 00:51:07.449
It's like, why are you even doing that?

982
00:51:07.469 --> 00:51:11.350
You're putting more energy in to actually get, like, some kind of fuel out.

983
00:51:11.450 --> 00:51:11.610
Right.

984
00:51:12.310 --> 00:51:16.111
On top of that, if oil prices are super cheap, then, like, why would you even do this?

985
00:51:16.251 --> 00:51:16.391
Right.

986
00:51:16.431 --> 00:51:16.611
Right.

987
00:51:16.891 --> 00:51:17.352
So...

988
00:51:20.973 --> 00:51:26.855
And whenever you're doing anything like this, there's also all of these contaminations that happen, right?

989
00:51:26.875 --> 00:51:28.456
Because plastic has like food products.

990
00:51:28.756 --> 00:51:31.577
Sometimes they have PVC, which is polyvinyl chloride.

991
00:51:32.078 --> 00:51:38.481
And if that gets in here and as it's degrading, so this polyvinyl chloride, it's found in like pipes, flooring.

992
00:51:39.041 --> 00:51:44.103
And if that gets in there, even in a batch, a small thing in a huge batch, right?

993
00:51:44.183 --> 00:51:47.004
That whole batch is done because you get this hydrochloric acid.

994
00:51:48.465 --> 00:51:49.606
All the equipment is done.

995
00:51:50.086 --> 00:51:52.567
And at the end of the day, you're dumping it into the landfill anyway.

996
00:51:53.007 --> 00:51:53.768
Okay.

997
00:51:53.908 --> 00:51:56.649
So it's really kind of bad.

998
00:51:56.829 --> 00:51:58.110
We have a lot of bad options.

999
00:51:58.170 --> 00:52:00.051
We have really a lot of bad options.

1000
00:52:00.091 --> 00:52:02.572
The only good option is to take the plastic.

1001
00:52:02.672 --> 00:52:06.694
Usually it's in this thing called polyolefin, which is a CH2.

1002
00:52:08.115 --> 00:52:15.063
Carbon, two hydrogens attached to a C and a H and then some other thing in our group, a residual group.

1003
00:52:16.464 --> 00:52:22.010
Those are the things that are basically two thirds of the global plastic chain.

1004
00:52:22.030 --> 00:52:27.476
This is like, you know, your milk cartons and plastic spoons and like plastic wrap.

1005
00:52:27.816 --> 00:52:29.437
and all this other kind of stuff.

1006
00:52:30.358 --> 00:52:34.802
Single use, really short lifespan, just gets used, gets thrown.

1007
00:52:35.162 --> 00:52:35.382
Okay?

1008
00:52:35.662 --> 00:52:40.846
That's what we're trying to, at the end, ultimately, we're trying to recycle.

1009
00:52:40.866 --> 00:52:42.428
We're trying to repeal and replace it.

1010
00:52:42.688 --> 00:52:42.868
Yeah.

1011
00:52:43.268 --> 00:52:44.049
Yes, exactly.

1012
00:52:44.429 --> 00:52:48.312
And what I just told you, right, the options are pretty bad.

1013
00:52:48.472 --> 00:52:48.632
Right.

1014
00:52:48.853 --> 00:52:49.093
Okay?

1015
00:52:49.613 --> 00:52:51.795
There's one option that has a little bit of hope.

1016
00:52:53.136 --> 00:52:58.820
And that is this thing called hydrogenolysis.

1017
00:52:59.321 --> 00:53:03.324
It's basically breaking it down in the presence of hydrogen.

1018
00:53:04.104 --> 00:53:10.429
And what it does is the hydrogen sort of comes in along with some kind of catalyst that helps this reaction.

1019
00:53:10.730 --> 00:53:16.054
And that hydrogen comes in and with the carbons and the hydrogens there, it creates hydrocarbons.

1020
00:53:16.154 --> 00:53:16.274
Okay.

1021
00:53:16.434 --> 00:53:26.877
that can then be used for either fuel or like waxes or like these higher sort of lubricants, things like that that you can recycle them to, right?

1022
00:53:27.778 --> 00:53:29.558
So that's one way of doing things.

1023
00:53:29.598 --> 00:53:33.980
That's like kind of a promising way of like recycling this plastic and...

1024
00:53:35.100 --> 00:53:37.461
putting it back into simple hydrocarbons, right?

1025
00:53:37.641 --> 00:53:41.342
And that's something that's like much less worse for the environment, right?

1026
00:53:41.382 --> 00:53:43.942
Like animals aren't choking on it and shit like that.

1027
00:53:44.322 --> 00:53:45.663
So that's what we want to do.

1028
00:53:45.723 --> 00:53:57.266
Now, usually this hydrogenolysis takes catalysts, and those catalysts require like higher order elements, rare earth elements like platinum, palladium,

1029
00:53:58.106 --> 00:53:59.587
We don't got a lot of that stuff.

1030
00:53:59.607 --> 00:54:03.909
Like, the amount of plastic we got compared to the amount of platinum and palladium.

1031
00:54:03.929 --> 00:54:07.831
Like, this is the stuff that Iron Man used to make his, like, new element.

1032
00:54:08.031 --> 00:54:08.272
Okay?

1033
00:54:08.312 --> 00:54:11.093
So, that's how rare it was that Marvel was using it in a movie.

1034
00:54:11.413 --> 00:54:18.277
So, we can't be dependent on platinum and palladium to, like, recycle...

1035
00:54:18.557 --> 00:54:43.080
the tons and tons of plastic right how many mines we gonna get right so so the idea is you want you want to have a catalyst that is not that rare that's what this team is doing this okay this this fundamentally okay yes fundamentally they're they they found a way to catal catalyze hydrogenolysis using a catalyst made out of nickel which is very abundant and

1036
00:54:43.540 --> 00:54:49.821
Furthermore, this catalysis only targets a specific type of this polyolefin.

1037
00:54:50.162 --> 00:54:50.402
Okay?

1038
00:54:50.622 --> 00:54:53.002
Because as I said, this is the CH2, CH2, right?

1039
00:54:53.042 --> 00:54:54.263
And then there's the R group.

1040
00:54:54.583 --> 00:54:57.463
That R group can be sometimes H, which makes it polyethylene.

1041
00:54:58.524 --> 00:55:05.805
Or it can be sometimes polypropylene, which is the other group is another complex thing, like another CH2.

1042
00:55:06.125 --> 00:55:11.166
So you get these like, with polyethylene, you just get this chain of carbons with hydrogens attached.

1043
00:55:11.287 --> 00:55:11.467
Yes.

1044
00:55:11.747 --> 00:55:12.547
Like a caterpillar.

1045
00:55:12.687 --> 00:55:12.887
Yes.

1046
00:55:13.787 --> 00:55:19.452
Or you get for polypropylene, you get the chain of carbons with carbon sticking out.

1047
00:55:19.632 --> 00:55:19.812
Yep.

1048
00:55:19.932 --> 00:55:20.132
Okay?

1049
00:55:20.752 --> 00:55:25.276
And so what you want to do is you don't want to recycle these together.

1050
00:55:25.836 --> 00:55:26.016
Okay.

1051
00:55:26.036 --> 00:55:27.998
And usually that requires separation.

1052
00:55:28.338 --> 00:55:31.420
And that's really, really just a pain.

1053
00:55:31.440 --> 00:55:31.961
Yeah, yeah, yeah.

1054
00:55:32.181 --> 00:55:32.341
Yeah.

1055
00:55:32.521 --> 00:55:32.801
All right.

1056
00:55:32.821 --> 00:55:33.762
Nobody wants to do this.

1057
00:55:33.822 --> 00:55:34.783
Both time and cost prohibitive.

1058
00:55:34.823 --> 00:55:36.284
Both time and cost prohibitive.

1059
00:55:36.404 --> 00:55:38.626
And more importantly, it's cost prohibitive, right?

1060
00:55:38.846 --> 00:55:41.208
And the consumer is not going to do it because, again, we're lazy.

1061
00:55:41.228 --> 00:55:41.548
Yeah, yeah.

1062
00:55:41.648 --> 00:55:42.669
We don't want to do that.

1063
00:55:43.089 --> 00:55:50.334
Yes, I love the earth, but I don't want to check under the bottle what kind of plastic it is before I throw it in the garbage.

1064
00:55:50.354 --> 00:56:01.641
This goes back to the sorting issue, which is like, ideally, this is why you want to have the humans who have the products put it into different buckets so that when it gets to the facility to recycle, it is pre-sorted.

1065
00:56:01.982 --> 00:56:02.222
Yeah.

1066
00:56:02.762 --> 00:56:03.303
But yeah.

1067
00:56:03.443 --> 00:56:06.325
And it's like, even if you got the humans to do it, though...

1068
00:56:07.045 --> 00:56:08.547
Do you really trust that human?

1069
00:56:08.587 --> 00:56:09.167
That's the point.

1070
00:56:09.588 --> 00:56:12.210
Like, I don't know if this human can read.

1071
00:56:12.490 --> 00:56:12.671
Right.

1072
00:56:12.771 --> 00:56:12.991
Right.

1073
00:56:13.131 --> 00:56:14.592
So it's like, I don't know.

1074
00:56:14.673 --> 00:56:15.233
Right.

1075
00:56:15.253 --> 00:56:19.918
So you would rather have a way of doing it within the chemical reaction.

1076
00:56:20.258 --> 00:56:20.418
Right.

1077
00:56:20.658 --> 00:56:23.081
That targets one type of plastic and not the other.

1078
00:56:23.701 --> 00:56:23.882
Right.

1079
00:56:23.942 --> 00:56:25.203
And that's what this thing is doing.

1080
00:56:25.483 --> 00:56:28.326
So they developed a catalyst out of nickel.

1081
00:56:28.866 --> 00:56:30.108
It's got aluminum, nickel.

1082
00:56:30.688 --> 00:56:32.529
the usual carbon and all this other kind of stuff.

1083
00:56:32.989 --> 00:56:38.791
And one of the cool things they did was this catalyst, you can imagine it as, again, a Lego block, okay?

1084
00:56:39.211 --> 00:56:46.434
But this Lego block has only a single piece of, like, active site, okay?

1085
00:56:46.454 --> 00:56:46.674
Okay.

1086
00:56:46.934 --> 00:56:55.417
Whenever you have a catalyst, you have something called an active site, which is the part that does the locking and the moving around of the atoms, okay?

1087
00:56:55.837 --> 00:56:57.918
This thing has a single active site, okay?

1088
00:56:58.198 --> 00:56:59.139
And what that gives it is...

1089
00:57:00.359 --> 00:57:08.242
the benefit of being extremely specific to what shape it wants to actually attach to, right?

1090
00:57:08.522 --> 00:57:17.385
A lot of catalysts have different shapes, which means that when you try to put it into something like polypropylene or polyethylene, they're going to attack every single carbon-carbon bond, okay?

1091
00:57:17.626 --> 00:57:22.047
And then you're going to get a bunch of methane, which is fine, I guess, but like it's not like great.

1092
00:57:22.127 --> 00:57:23.148
Methane is like super cheap.

1093
00:57:23.168 --> 00:57:23.928
It's like this low-

1094
00:57:24.989 --> 00:57:30.634
What you'd like to do is with polypropylene, I told you, right, there's these carbon and then there's these carbons that are sticking out.

1095
00:57:30.874 --> 00:57:33.076
You want to just take out the carbon that's sticking out.

1096
00:57:33.256 --> 00:57:36.179
And then this thing just becomes like polyethylene.

1097
00:57:36.319 --> 00:57:40.222
And then you can then do those two together in some other sort of process.

1098
00:57:41.864 --> 00:57:46.327
So you want to take those out and then only target specific bonds.

1099
00:57:46.387 --> 00:57:48.109
And that's what this new catalyst does.

1100
00:57:48.129 --> 00:57:48.990
Very, very clever.

1101
00:57:49.230 --> 00:57:52.232
So it's doing the sorting within the chemistry.

1102
00:57:52.693 --> 00:57:57.297
At the facility where they're going to be doing the recycling, regardless of the input,

1103
00:57:57.977 --> 00:58:03.999
The chemical process itself is doing the job of separation.

1104
00:58:04.239 --> 00:58:04.699
Exactly.

1105
00:58:04.739 --> 00:58:09.981
That we otherwise are, in some cases, in some places, depending on humans, to put things in different buckets.

1106
00:58:10.101 --> 00:58:10.941
Yeah, exactly.

1107
00:58:11.281 --> 00:58:13.782
And the results are, they speak for themselves.

1108
00:58:14.062 --> 00:58:22.665
So they tested this catalyst with isotactic polypropylene, which is this thing that I was telling you about with the branches coming out.

1109
00:58:23.065 --> 00:58:23.625
And 99.5% conversion.

1110
00:58:25.426 --> 00:58:28.067
into lighter compounds in 20 minutes.

1111
00:58:28.547 --> 00:58:34.390
And then when they did the same thing, they did the same thing with the polyethylene.

1112
00:58:34.530 --> 00:58:34.730
Yes.

1113
00:58:35.390 --> 00:58:36.311
Nothing happened.

1114
00:58:37.031 --> 00:58:37.231
Right?

1115
00:58:37.311 --> 00:58:37.952
Yes.

1116
00:58:38.192 --> 00:58:39.132
And it took four hours.

1117
00:58:39.192 --> 00:58:40.153
Yeah.

1118
00:58:40.353 --> 00:58:41.993
And then they finally got a little bit of signal.

1119
00:58:42.274 --> 00:58:44.374
So what you can do is you can have a mixture of this stuff.

1120
00:58:44.395 --> 00:58:45.555
You can apply this catalyst.

1121
00:58:45.615 --> 00:58:45.815
Yes.

1122
00:58:45.955 --> 00:58:48.076
And then you can just let it run.

1123
00:58:48.196 --> 00:58:48.396
Yes.

1124
00:58:48.456 --> 00:58:52.138
And then within a short amount of time, all the polypropylene will be,

1125
00:58:52.438 --> 00:58:56.301
decomposed, but the polyethylene will be in this native state, and then you can go forward.

1126
00:58:56.321 --> 00:59:07.208
And now you have these two buckets of distilled compounds that you then have functional use for in any number of other downstream or upstream, depending on what way you want to look at it, industrial processes.

1127
00:59:07.228 --> 00:59:07.889
That's right, yeah.

1128
00:59:08.149 --> 00:59:11.191
And then the best part was that, you know that PVC that I was talking about?

1129
00:59:11.331 --> 00:59:11.491
Yes.

1130
00:59:11.551 --> 00:59:20.857
The polyvinyl chloride, which is a contaminant for all of these plastics, that basically makes plastic unrecyclable in these previous strategies?

1131
00:59:20.877 --> 00:59:21.698
Yes.

1132
00:59:22.218 --> 00:59:25.439
Somehow, and I don't think they fully understand how this works.

1133
00:59:25.479 --> 00:59:25.860
We love that.

1134
00:59:25.880 --> 00:59:26.560
This made it better.

1135
00:59:27.360 --> 00:59:27.700
Really?

1136
00:59:27.720 --> 00:59:28.401
Yeah.

1137
00:59:28.461 --> 00:59:32.582
The introduction of PVC made the catalysis reaction better.

1138
00:59:32.902 --> 00:59:35.003
No friggin' way.

1139
00:59:35.043 --> 00:59:35.143
Yeah.

1140
00:59:35.163 --> 00:59:36.804
So then they were like, oh, that's free.

1141
00:59:37.144 --> 00:59:37.644
That's great.

1142
00:59:38.185 --> 00:59:39.425
Just like free upside.

1143
00:59:39.445 --> 00:59:40.706
It's like, oh, that's great.

1144
00:59:40.946 --> 00:59:41.206
Yeah.

1145
00:59:41.266 --> 00:59:44.187
Which then removes the cross-contamination issue.

1146
00:59:44.367 --> 00:59:44.587
Yeah.

1147
00:59:44.687 --> 00:59:46.288
Which goes back to this no sort.

1148
00:59:46.308 --> 00:59:47.108
Like, yeah.

1149
00:59:47.128 --> 00:59:48.128
So then we don't have to sort it.

1150
00:59:48.148 --> 00:59:49.669
We don't even have to worry about the PVC.

1151
00:59:49.689 --> 00:59:51.850
The PVC could be in there and it's just going to make it better.

1152
00:59:51.890 --> 00:59:52.090
Yeah.

1153
00:59:52.210 --> 00:59:53.010
It's just like fine.

1154
00:59:53.270 --> 01:00:01.173
So I think this is a really cool use of like fundamental chemistry to try to like tackle a very, very real problem.

1155
01:00:01.213 --> 01:00:02.173
This is really good.

1156
01:00:02.253 --> 01:00:09.176
That's actually a great explanation because I actually had no context for how the recycling process really worked.

1157
01:00:09.396 --> 01:00:11.257
Yeah, I didn't either until I started researching this.

1158
01:00:11.457 --> 01:00:14.698
And this is like a clear like it's a clear problem set.

1159
01:00:14.718 --> 01:00:17.019
We don't have good solutions currently implemented at scale.

1160
01:00:17.239 --> 01:00:17.460
No.

1161
01:00:18.220 --> 01:00:18.840
And we.

1162
01:00:19.280 --> 01:00:20.921
But it's something that I really do care about.

1163
01:00:21.061 --> 01:00:21.221
Right.

1164
01:00:21.261 --> 01:00:22.062
Just, you know.

1165
01:00:22.562 --> 01:00:22.722
Right.

1166
01:00:22.782 --> 01:00:23.662
But do it at the factory.

1167
01:00:23.702 --> 01:00:23.942
Yeah.

1168
01:00:24.082 --> 01:00:24.322
Yeah.

1169
01:00:26.103 --> 01:00:26.723
I'm lazy.

1170
01:00:26.783 --> 01:00:27.324
I'm sorry.

1171
01:00:27.964 --> 01:00:29.124
As are the majority.

1172
01:00:29.384 --> 01:00:32.045
Again, it also should not be incumbent.

1173
01:00:32.145 --> 01:00:33.506
The commons is a concept.

1174
01:00:33.526 --> 01:00:36.927
We all should invest in the commons and trying to upkeep.

1175
01:00:37.267 --> 01:00:42.629
But at the end of the day, the biggest kind of issue is the industrial grade, industrial level.

1176
01:00:42.789 --> 01:00:43.209
Exactly.

1177
01:00:43.229 --> 01:00:46.790
It's hard for me to do anything.

1178
01:00:47.291 --> 01:00:48.871
And we have to address it at that level.

1179
01:00:48.891 --> 01:00:53.953
And one of the ways you can do so is when they bring the stuff to the facility, it can actually repurpose it in a way that creates...

1180
01:00:54.533 --> 01:01:04.724
functional, valuable output that is not the current state of affairs where it's either the output sucks and we wasted a bunch of energy, time, and money.

1181
01:01:04.964 --> 01:01:05.184
Yeah.

1182
01:01:05.384 --> 01:01:06.706
Or it just went to the landfill.

1183
01:01:06.726 --> 01:01:07.807
It just went to the landfill anyway.

1184
01:01:08.127 --> 01:01:15.575
Or the output doesn't have real value in terms of the concept of recycling it back into society.

1185
01:01:15.635 --> 01:01:16.495
Exactly, yeah.

1186
01:01:16.675 --> 01:01:19.256
And your point about the commons, I think, is important, right?

1187
01:01:19.296 --> 01:01:21.817
There is a commons that I am heavily invested in.

1188
01:01:22.237 --> 01:01:27.539
I pay taxes, and I want those commons, my tax dollars, to go into this kind of research.

1189
01:01:27.639 --> 01:01:30.160
Because that is the future of saving this planet.

1190
01:01:30.601 --> 01:01:36.343
Saving the planet is not making humans somehow be less comfortable, okay?

1191
01:01:36.683 --> 01:01:38.123
That's not going to work.

1192
01:01:38.263 --> 01:01:38.944
I know humans.

1193
01:01:39.064 --> 01:01:39.944
I know several humans.

1194
01:01:40.464 --> 01:01:42.425
They're all like me, okay?

1195
01:01:43.627 --> 01:01:50.552
And what we want is solutions that actually make sense, that don't compromise my state of life.

1196
01:01:51.112 --> 01:01:55.815
And yeah, we need fundamental science research to do this.

1197
01:01:56.296 --> 01:01:58.637
I think we've had enough innovation on our mobile phones.

1198
01:01:58.677 --> 01:02:00.678
Yeah, I'm totally fine.

1199
01:02:00.719 --> 01:02:03.901
Yeah, with the current state of my iPhone, it's sufficient.

1200
01:02:03.961 --> 01:02:10.185
Maybe let's not try to iterate on an incremental version for next year and spend that brain power, time, and energy.

1201
01:02:10.685 --> 01:02:17.009
on things that will create a better commons environment, which then drives the consumerism that capitalism loves.

1202
01:02:17.449 --> 01:02:18.830
Great, great story.

1203
01:02:18.850 --> 01:02:20.011
Yeah, I like the story.

1204
01:02:20.391 --> 01:02:29.916
Speaking of the private sector and multi-trillion dollar companies, we're going to end today with a story about analog AI coming out of Microsoft.

1205
01:02:29.936 --> 01:02:31.898
We almost cannot avoid AI every week.

1206
01:02:32.158 --> 01:02:32.338
Nope.

1207
01:02:32.738 --> 01:02:40.164
Headline on the story, Microsoft's analog optical computer cracks two practical problems and shows AI promise.

1208
01:02:40.184 --> 01:02:42.706
This was published on Microsoft News' blog.

1209
01:02:42.987 --> 01:02:46.590
But they did put out a research paper associated with this in nature.

1210
01:02:47.010 --> 01:02:49.152
So it's not just marketing.

1211
01:02:49.772 --> 01:02:51.614
They did sort of go through the peer review.

1212
01:02:52.114 --> 01:02:57.403
And so I don't really actually even understand what they mean when they say analog optical computer.

1213
01:02:57.503 --> 01:02:58.124
Yeah.

1214
01:02:58.144 --> 01:03:00.969
And there's several problems to crack, but they cracked two of them.

1215
01:03:00.989 --> 01:03:01.510
So let me know.

1216
01:03:01.630 --> 01:03:02.732
What are we talking about here?

1217
01:03:02.920 --> 01:03:03.320
That's right.

1218
01:03:04.301 --> 01:03:08.345
It's an analog computer instead of a digital one, which is very cool to me.

1219
01:03:09.366 --> 01:03:12.528
Digital, you got your zeros and ones, right?

1220
01:03:13.169 --> 01:03:17.292
And we talked a lot about digital computers in the past and some of their problems, right?

1221
01:03:17.993 --> 01:03:22.497
The fact that Moore's Law is slowing down because we're sort of reaching this limit of

1222
01:03:23.137 --> 01:03:35.583
of how small we can go before the capacitors and the transistors that we're using start leaking off so much power that we just have to constantly keep powering it in order for the data to actually stay alive.

1223
01:03:35.983 --> 01:03:40.305
We also talked about the von Neumann gap, right?

1224
01:03:40.325 --> 01:03:49.049
The von Neumann bottleneck, which is the separation between your processor and the memory there, which is your RAM,

1225
01:03:49.689 --> 01:04:10.212
and your hard disk or your actual memory yes right and the where you're storing stuff is different from where you are doing the computation yes right this is a thing that plagues cpus but it also plagues gpus okay and um as ai becomes increasingly big

1226
01:04:11.610 --> 01:04:19.651
There's going to come a point where we're going to really have to start having the conversation of, again, you know, what we were talking about last in the last story.

1227
01:04:20.492 --> 01:04:22.972
Is this worth the environmental impact?

1228
01:04:23.052 --> 01:04:23.232
Right.

1229
01:04:23.392 --> 01:04:23.612
Right.

1230
01:04:24.152 --> 01:04:26.333
And of course, the corporations are going to be like, yes.

1231
01:04:26.773 --> 01:04:29.093
And they're going to make bigger and bigger data centers.

1232
01:04:29.493 --> 01:04:33.014
For example, the energy usage has tripled from 2014 to 2023.

1233
01:04:33.094 --> 01:04:36.274
And it's projected to be 20% of global electricity by 2030.

1234
01:04:39.955 --> 01:04:46.618
Okay, the data centers and the AI, all that AI computations for you making a cat that wears a hat or whatever.

1235
01:04:47.078 --> 01:04:49.779
Like that's 20% of global electricity.

1236
01:04:50.140 --> 01:04:52.321
Now, I'm not saying that we shouldn't be using AI.

1237
01:04:52.641 --> 01:04:54.862
Like because I use AI, AI is great.

1238
01:04:56.682 --> 01:05:04.346
But I think all of these companies, including Microsoft, are starting to realize that it's no longer going to become tenable, right?

1239
01:05:04.366 --> 01:05:04.506
100%.

1240
01:05:06.067 --> 01:05:11.073
there needs to be an alternative way to do the computation that they are seeking.

1241
01:05:11.534 --> 01:05:19.543
I'll just make a brief note here, which is that every single one of the either Frontier Model Builders

1242
01:05:20.444 --> 01:05:37.194
or sort of hyperscaler cloud computing platforms, Google Cloud Product, AWS, they're all spending a ton of their capital on directly investing into power infrastructure.

1243
01:05:37.334 --> 01:05:40.155
Yeah, because they know that that's going to be a bottleneck.

1244
01:05:40.316 --> 01:05:47.500
Every single hyperscaler or frontier model company is investing, if not hundreds of millions, billions of dollars in...

1245
01:05:48.340 --> 01:05:52.944
into partnerships with these mini nuclear reactor companies.

1246
01:05:53.044 --> 01:05:58.128
They're literally returning on the big power facility that's just north of us in California.

1247
01:05:58.648 --> 01:06:00.790
The nuclear one, Facebook has a little alcove there.

1248
01:06:01.150 --> 01:06:03.252
So they all know the story.

1249
01:06:03.312 --> 01:06:03.512
Yeah.

1250
01:06:03.692 --> 01:06:04.133
Like it's not.

1251
01:06:04.153 --> 01:06:04.613
They see it.

1252
01:06:04.653 --> 01:06:05.494
They see the writing on the wall.

1253
01:06:05.514 --> 01:06:06.855
It's not controversial for any of them.

1254
01:06:07.859 --> 01:06:08.620
Yeah.

1255
01:06:08.780 --> 01:06:14.945
And on top of that, there's a bunch of these companies that are investing in alternative computing technology, right?

1256
01:06:14.965 --> 01:06:15.145
100%.

1257
01:06:15.285 --> 01:06:17.267
Like we were talking about Spintronics the other day.

1258
01:06:17.287 --> 01:06:18.147
Yes.

1259
01:06:18.348 --> 01:06:19.909
This is an analog computer.

1260
01:06:20.149 --> 01:06:22.371
It's doing sort of the same type of deal.

1261
01:06:22.731 --> 01:06:31.038
Not the same physics, but the same sense of let's go away from transistors and physical stuff.

1262
01:06:31.378 --> 01:06:36.182
And let's try to find a different, completely different way to store data and manipulate data.

1263
01:06:36.202 --> 01:06:36.443
Okay.

1264
01:06:36.563 --> 01:06:36.803
Okay.

1265
01:06:38.263 --> 01:06:49.628
If we go back to the von Neumann bottleneck, right, from John von Neumann, he championed this idea of that early computer where there was going to be a separation between the processor and memory, right?

1266
01:06:49.888 --> 01:06:52.769
That's become sort of the bedrock for modern computation.

1267
01:06:53.189 --> 01:06:58.871
But it's really become a huge handicap for AI, okay?

1268
01:06:59.191 --> 01:07:03.053
Because when we want to do AI, there's two core operations that we have to do.

1269
01:07:03.333 --> 01:07:15.142
OK, one is matrix multiplication, right, where we take the weights of your neural network and you multiply it by some vector and then you multiply by another matrix and another matrix.

1270
01:07:15.162 --> 01:07:17.424
And all of these are weights that you've stored.

1271
01:07:18.245 --> 01:07:29.093
And the other thing that you have to do is you have to do a nonlinearity, which is this idea that if if my input is above a certain threshold, only then do I let it through.

1272
01:07:29.433 --> 01:07:31.495
If it's below a certain threshold, then it's a zero.

1273
01:07:31.855 --> 01:07:32.095
Okay.

1274
01:07:33.096 --> 01:07:35.297
That's one of the types of non-linearity that you have to do.

1275
01:07:35.578 --> 01:07:41.622
Those are the two fundamental things that you have to do as someone who wants to implement a neural network.

1276
01:07:41.962 --> 01:07:42.182
Okay.

1277
01:07:42.783 --> 01:07:46.345
Then there's other things like momentum and annealing and all this other stuff that you have to do.

1278
01:07:46.365 --> 01:07:53.310
But fundamentally, if you really want to do neural network stuff, it's matrix multiplication and implementing a non-linearity.

1279
01:07:53.610 --> 01:08:01.175
Okay, so now in order to do these things in a traditional computer, what you have to do is you have to store your weights in a certain spot, okay?

1280
01:08:01.435 --> 01:08:03.336
And then you got to load them up in your processor.

1281
01:08:03.396 --> 01:08:03.556
Right.

1282
01:08:03.716 --> 01:08:06.758
And then you got to multiply the weights to this thing.

1283
01:08:07.118 --> 01:08:08.679
And then you got to load up new weights.

1284
01:08:08.739 --> 01:08:10.260
And then you got to multiply to this thing.

1285
01:08:10.601 --> 01:08:17.845
And, you know, now we've gotten clever where maybe we can load up all of the weights to this thing and, you know, do a forward pass and things like that.

1286
01:08:18.105 --> 01:08:28.758
But fundamentally, a lot of the source of latency that comes from like you waiting while you've typed something into ChatGPT and it's like thinking is doing this, right?

1287
01:08:28.818 --> 01:08:29.699
It's like loading up.

1288
01:08:30.060 --> 01:08:35.466
And then there's like copper wires in whatever server that is like transferring data.

1289
01:08:36.327 --> 01:08:38.951
And the latency is a huge part of it too, right?

1290
01:08:39.211 --> 01:08:42.055
So there's a lot of power that's being dissipated for no reason.

1291
01:08:42.115 --> 01:08:43.817
And then there's also a lot of latency.

1292
01:08:44.158 --> 01:08:45.540
This thing is solving both of those.

1293
01:08:45.580 --> 01:08:46.401
That's very interesting.

1294
01:08:46.441 --> 01:08:46.581
Okay.

1295
01:08:47.002 --> 01:08:54.071
This thing is solving both of those by creating a computer out of light, lenses, and filters.

1296
01:08:54.712 --> 01:08:55.293
Very interesting.

1297
01:08:55.313 --> 01:08:55.453
Okay?

1298
01:08:55.653 --> 01:08:55.913
Okay.

1299
01:08:55.993 --> 01:08:56.934
It's very cool to me.

1300
01:08:56.954 --> 01:08:57.295
Yeah, yeah, yeah.

1301
01:08:57.315 --> 01:09:00.598
Because it's like fundamental physics used in a whole new way.

1302
01:09:00.698 --> 01:09:00.978
Okay.

1303
01:09:01.119 --> 01:09:01.339
Okay?

1304
01:09:02.300 --> 01:09:11.749
So, to understand this, let's look at the two things that we need to do in order to do a neural network implementation.

1305
01:09:11.869 --> 01:09:12.090
Okay?

1306
01:09:12.110 --> 01:09:14.332
The first one I talked about was matrix multiplication.

1307
01:09:14.352 --> 01:09:14.532
Okay?

1308
01:09:14.772 --> 01:09:20.637
That's this idea of you take a vector of numbers, which is, let's say, a row of numbers, and then you apply a matrix to it.

1309
01:09:20.657 --> 01:09:24.039
So this this number gets applied, gets multiplied to all of these.

1310
01:09:24.079 --> 01:09:25.741
This number gets multiplied to all of these.

1311
01:09:26.021 --> 01:09:28.683
And then you sum it up and you get a new vector, kind of like things like that.

1312
01:09:28.743 --> 01:09:28.923
Right.

1313
01:09:30.124 --> 01:09:31.205
What these guys figured is.

1314
01:09:33.497 --> 01:09:37.383
We could do that with like light and filters.

1315
01:09:37.803 --> 01:09:38.184
Okay.

1316
01:09:38.204 --> 01:09:38.444
Okay.

1317
01:09:39.005 --> 01:09:41.749
So the vector is an array of numbers, right?

1318
01:09:42.871 --> 01:09:45.775
We could represent that with the brightness of LEDs.

1319
01:09:48.355 --> 01:09:50.096
Do you see where I'm going now with the whole thing?

1320
01:09:50.156 --> 01:09:51.077
I already see where you're going.

1321
01:09:51.157 --> 01:09:51.417
Okay.

1322
01:09:51.857 --> 01:09:53.999
I could represent that with a brightness of LEDs.

1323
01:09:54.479 --> 01:09:54.759
Okay.

1324
01:09:55.439 --> 01:09:58.862
Then I've got a matrix that I need to multiply it with, right?

1325
01:09:59.342 --> 01:10:01.283
I could represent that with a bunch of filters.

1326
01:10:01.303 --> 01:10:02.104
Oh, no.

1327
01:10:02.184 --> 01:10:02.464
Right?

1328
01:10:03.324 --> 01:10:04.725
That each light goes through.

1329
01:10:04.745 --> 01:10:05.546
I'm so mad.

1330
01:10:06.106 --> 01:10:06.786
It's so nice.

1331
01:10:06.867 --> 01:10:07.067
Yeah.

1332
01:10:07.527 --> 01:10:08.207
It's so nice.

1333
01:10:08.768 --> 01:10:08.968
Right?

1334
01:10:09.308 --> 01:10:11.770
And then I've got an output vector, right?

1335
01:10:12.050 --> 01:10:17.233
That output vector could be a CCD that senses how much light went through.

1336
01:10:19.384 --> 01:10:19.644
Okay?

1337
01:10:20.325 --> 01:10:22.026
So completely analog.

1338
01:10:22.086 --> 01:10:22.286
Right.

1339
01:10:22.907 --> 01:10:23.067
Right.

1340
01:10:23.527 --> 01:10:23.687
Right?

1341
01:10:23.747 --> 01:10:25.088
There's no zeros and ones.

1342
01:10:25.208 --> 01:10:33.714
It's like something, there's a continuous value here of the brightness, there's a continuous value of the filters, and then there's a CCD that captures whatever comes out.

1343
01:10:34.014 --> 01:10:34.175
Right?

1344
01:10:34.215 --> 01:10:37.217
And then now, okay, the second thing I have to do is implement a nonlinearity.

1345
01:10:37.497 --> 01:10:37.657
Right?

1346
01:10:37.917 --> 01:10:39.959
In order to implement the nonlinearity, they've got...

1347
01:10:40.660 --> 01:10:44.521
like a bunch of filters that they can use, like the bipolar difference filter.

1348
01:10:44.541 --> 01:10:46.422
So they've got, so now they've got these CCDs, right?

1349
01:10:46.442 --> 01:10:47.483
That converts it into current.

1350
01:10:47.823 --> 01:10:49.623
The current now goes through analog.

1351
01:10:49.643 --> 01:10:50.644
It's still analog.

1352
01:10:51.004 --> 01:10:55.345
A lot of these implementations, what they've done is they've done a hybrid approach of digital and analog.

1353
01:10:55.685 --> 01:10:57.466
This is all analog.

1354
01:10:57.746 --> 01:10:58.947
Pure analog, it's wonderful.

1355
01:10:59.227 --> 01:11:02.248
You got these, you got the light that's coming out like this.

1356
01:11:03.572 --> 01:11:05.794
It is now going to go through this sort of filter block.

1357
01:11:06.294 --> 01:11:06.534
Okay.

1358
01:11:06.734 --> 01:11:11.517
It's going to go through a bipolar differential pair, which is going to do the non-linearity.

1359
01:11:11.877 --> 01:11:16.100
So if the light, if the current is below a certain amount, it's going to give it zero.

1360
01:11:16.360 --> 01:11:19.742
But if it's above a certain amount, then it's going to let through a bunch of current.

1361
01:11:19.862 --> 01:11:23.665
You also want to do this thing called annealing, which is this idea of like...

1362
01:11:24.045 --> 01:11:37.457
Taking the current representation and sort of changing it a tiny bit so that you don't get stuck in like a state of the neural network where you're not like moving and you're getting the wrong answer, so to speak.

1363
01:11:37.597 --> 01:11:40.559
So you can do that using these things called variable gain amplifiers.

1364
01:11:41.520 --> 01:11:44.942
Which are VGAs used a lot all over tech.

1365
01:11:45.342 --> 01:11:49.224
You can also do this thing called momentum, which is something that you use in neural networks.

1366
01:11:49.264 --> 01:11:53.786
If you're like converging on an answer, you want to keep going in that direction.

1367
01:11:53.846 --> 01:11:57.948
That's something that you can implement with VGAs, which is, again, something that everybody uses.

1368
01:11:58.249 --> 01:12:02.211
And then you can feed it back into the thing and have this loop.

1369
01:12:02.591 --> 01:12:07.233
Now, this is a fundamentally different architecture, not just for the physics of it, right?

1370
01:12:07.253 --> 01:12:08.494
Because we're not using GPUs.

1371
01:12:08.514 --> 01:12:10.355
We're using light and matrices.

1372
01:12:10.395 --> 01:12:10.515
Right.

1373
01:12:11.275 --> 01:12:12.796
in the form of filters and things like that.

1374
01:12:12.816 --> 01:12:19.240
But also you can imagine it as like, there's a single sort of neural network that we're like passing over and over again.

1375
01:12:19.280 --> 01:12:19.520
Right.

1376
01:12:20.041 --> 01:12:25.244
So what these things are implementing is something called deep equilibrium models.

1377
01:12:25.665 --> 01:12:36.392
There are different types of neural network where you've got a neural network and instead of a forward pass that gives you a right answer, what these models do is they call the model over and over again,

1378
01:12:37.272 --> 01:12:42.717
until the model doesn't change its output.

1379
01:12:43.077 --> 01:12:55.328
So what it's doing is, it's in the space and it's going down, kind of like a gradient descent algorithm, where it's going down this sort of energy landscape, until it gets to this minima, where as it moves, it's not really moving anywhere.

1380
01:12:55.368 --> 01:12:58.370
Before it was moving here, then it's moving here, it's trying to find the answer.

1381
01:12:58.390 --> 01:12:59.951
But once it gets to the answer,

1382
01:13:00.732 --> 01:13:04.153
the more you apply it all the way to infinity, it's not really going to move.

1383
01:13:04.213 --> 01:13:04.373
Right.

1384
01:13:04.573 --> 01:13:04.753
Right.

1385
01:13:05.013 --> 01:13:06.113
That's what this thing is doing.

1386
01:13:06.373 --> 01:13:12.915
The problem with deep equilibrium models on GPUs was, it was like extremely, like how many times am I going to, right.

1387
01:13:13.115 --> 01:13:18.456
But this, the physics itself is geared towards these deep equilibrium models.

1388
01:13:18.556 --> 01:13:18.776
Yes.

1389
01:13:18.796 --> 01:13:19.496
Right.

1390
01:13:19.636 --> 01:13:27.378
And so you've got this, you've got this entire like system that is a digital analog.

1391
01:13:27.998 --> 01:13:30.219
I mean, sorry, it's an analog computer.

1392
01:13:30.399 --> 01:13:30.539
Right.

1393
01:13:30.719 --> 01:13:30.939
Right.

1394
01:13:31.180 --> 01:13:31.340
Right.

1395
01:13:31.500 --> 01:13:33.382
Which is which is, I think, very cool.

1396
01:13:33.502 --> 01:13:46.458
And because part of the this goes back to like the energy and cost issue with, you know, GPUs are both expensive on their face to produce at the scale we need to to do these like chain of thought reasoning level models, these frontier models.

1397
01:13:47.559 --> 01:13:57.948
And then the power usage, not only to acquire the physical hardware is cost prohibitive, but then to run them to be able to get to that level is also cost prohibitive.

1398
01:13:58.229 --> 01:14:00.271
And this sort of solves both issues.

1399
01:14:01.292 --> 01:14:06.777
Again, accounting for the fact that this is obviously not an industrial manufacturing scale and all these other things.

1400
01:14:08.628 --> 01:14:12.832
This is not industrial manufacturing, but they're not using like crazy tech.

1401
01:14:12.872 --> 01:14:13.653
That's okay.

1402
01:14:13.673 --> 01:14:14.894
They're using LEDs.

1403
01:14:15.355 --> 01:14:18.578
They're using the matrix that they use for the multiplication.

1404
01:14:18.878 --> 01:14:20.700
That's just like the filter in your projector.

1405
01:14:20.720 --> 01:14:21.701
That's very interesting.

1406
01:14:21.721 --> 01:14:21.982
Okay.

1407
01:14:22.042 --> 01:14:25.085
And then the CCD is like the sensor in your phone.

1408
01:14:25.565 --> 01:14:29.009
And then you got VGAs, which everybody knows how to, like everybody has.

1409
01:14:29.289 --> 01:14:31.411
We know how to make it super cheap at this point.

1410
01:14:31.692 --> 01:14:34.895
You don't need like an insane fab to do any of this stuff.

1411
01:14:35.035 --> 01:14:46.327
Which this becomes very interesting because of the, I mean, there's so many implications here, not only from just from the fundamental perspective, but if you extrapolate both like geopolitically and economically.

1412
01:14:46.927 --> 01:15:01.857
You know, one of the issues with the whole AI debate is NVIDIA has now become a trillion dollar company in record speed because they're the only ones that can build these chips, these GPUs, sorry, these GPUs that can really run the frontier models effectively.

1413
01:15:02.897 --> 01:15:09.541
And to be able to get to NVIDIA scale because of the complexity of the manufacturing process and the amount of proprietary knowledge necessary.

1414
01:15:09.761 --> 01:15:11.302
There's like no IBM has tried.

1415
01:15:11.342 --> 01:15:12.262
They can't.

1416
01:15:12.342 --> 01:15:13.463
And there's no one else.

1417
01:15:13.563 --> 01:15:17.064
Yeah, because they've done 20 years of work to get to where they are.

1418
01:15:17.124 --> 01:15:17.305
Right.

1419
01:15:17.365 --> 01:15:17.825
Right.

1420
01:15:17.965 --> 01:15:27.810
And what's interesting is it sort of upsets the apple cart in the sense of if there is a low cost, high efficacy hardware substrate.

1421
01:15:28.490 --> 01:15:33.174
that you can utilize to do the same type of work that you're seeing out of a GT5.

1422
01:15:33.474 --> 01:15:35.816
Faster, cheaper, with less power.

1423
01:15:36.256 --> 01:15:39.799
Like one caveat is this thing is only going to work for inference.

1424
01:15:40.220 --> 01:15:40.660
Okay.

1425
01:15:40.720 --> 01:15:41.681
It's not for training.

1426
01:15:41.801 --> 01:15:45.384
Which is... Because in training, you'd have to change the weights and things like that, right?

1427
01:15:45.664 --> 01:15:50.348
So in order to train, what they did was they created a digital twin, which they then trained.

1428
01:15:50.508 --> 01:15:50.628
Yep.

1429
01:15:50.889 --> 01:15:53.591
And the digital twin matched the hardware's... Yep.

1430
01:15:53.951 --> 01:15:54.471
Performance.

1431
01:15:54.511 --> 01:16:06.239
So then when the digital twin said, okay, you want to set your filters this way and you want to set your nonlinearities this way, they could go and implement that with the hardware and then they could like do the inference time stuff on this.

1432
01:16:06.539 --> 01:16:09.942
But to be fair, like all of the...

1433
01:16:10.944 --> 01:16:15.565
Like the reason why all of these companies are building all of these data centers is not to train.

1434
01:16:15.625 --> 01:16:16.126
It's for inference.

1435
01:16:16.146 --> 01:16:17.106
It's for inference, right?

1436
01:16:17.126 --> 01:16:22.107
The training happens in like San Francisco in their facility or, you know, whatever.

1437
01:16:22.147 --> 01:16:29.169
In like some like tiny, like the hundreds of millions of users of ChatGPT and Gemini are just using it for inference.

1438
01:16:29.209 --> 01:16:29.830
It's all inference.

1439
01:16:29.850 --> 01:16:32.150
I was going to say like where the usage is going.

1440
01:16:32.350 --> 01:16:33.991
Like the power is going for inference.

1441
01:16:34.071 --> 01:16:34.351
Correct.

1442
01:16:34.731 --> 01:16:47.218
As more users start using, you know, LLMs and other versions of AI in everyday life, and we get to the order of billions, billions, billions of users doing so, what they're using is the inference.

1443
01:16:47.238 --> 01:16:53.681
And so this can be a whole – you can still have your proprietary training stacks with all these fancy GPUs and all that stuff.

1444
01:16:54.001 --> 01:16:56.883
But now you don't need to have that high-cost infrastructure.

1445
01:16:57.303 --> 01:17:00.605
in order to serve the end user of the models.

1446
01:17:00.625 --> 01:17:04.587
Yeah, and if you can miniaturize this thing, then you're good to go.

1447
01:17:05.167 --> 01:17:08.789
I mean, again, we're extrapolating here, but you can imagine I have my own little analog.

1448
01:17:08.969 --> 01:17:13.472
Yeah, yeah, and it's just, and the great thing is this uses LEDs, it's not using lasers.

1449
01:17:13.972 --> 01:17:14.152
Right.

1450
01:17:14.252 --> 01:17:15.513
So I don't need coherent light.

1451
01:17:15.533 --> 01:17:17.534
This is just like normal LEDs.

1452
01:17:17.655 --> 01:17:19.316
So that's so clever.

1453
01:17:19.376 --> 01:17:20.897
Yeah, that's so I love.

1454
01:17:20.937 --> 01:17:21.537
I love that.

1455
01:17:21.617 --> 01:17:23.859
Yeah, I think I think they did a really good job.

1456
01:17:23.899 --> 01:17:24.519
I quite like it.

1457
01:17:24.599 --> 01:17:26.581
Creativity is alive and well.

1458
01:17:26.601 --> 01:17:26.841
Yeah.

1459
01:17:26.881 --> 01:17:28.122
Fundamental science research.

1460
01:17:28.282 --> 01:17:31.344
One of the ways you get breakthroughs is by being creative.

1461
01:17:31.404 --> 01:17:31.604
Yeah.

1462
01:17:31.624 --> 01:17:32.124
Dogmatic.

1463
01:17:32.164 --> 01:17:32.384
Yeah.

1464
01:17:32.805 --> 01:17:32.965
Um.

1465
01:17:33.345 --> 01:17:35.365
That is really, really.

1466
01:17:35.385 --> 01:17:36.086
Yeah.

1467
01:17:36.206 --> 01:17:37.746
Coming from the tech space, this obviously.

1468
01:17:37.766 --> 01:17:37.866
Yeah.

1469
01:17:37.906 --> 01:17:39.346
And then my light bulbs are going off.

1470
01:17:39.386 --> 01:17:40.026
Yeah, dude.

1471
01:17:40.106 --> 01:17:45.908
And it's like 20 nanoseconds per search per iteration of this thing.

1472
01:17:46.888 --> 01:17:51.849
The idea is called a fixed point search because you're searching for a fixed point to get to the answer, right?

1473
01:17:51.869 --> 01:17:53.929
What I was telling you earlier about.

1474
01:17:54.049 --> 01:17:54.529
Yes.

1475
01:17:54.889 --> 01:17:55.329
Yeah.

1476
01:17:55.389 --> 01:17:57.130
20 nanoseconds per fixed point search.

1477
01:17:57.530 --> 01:18:02.491
And this thing is projected to be 100 times more efficient than leading GPUs.

1478
01:18:03.031 --> 01:18:07.453
Going at 4.5 tera operations per second per watt.

1479
01:18:07.933 --> 01:18:08.093
Right?

1480
01:18:08.113 --> 01:18:10.074
That's 100 times more efficient than GPUs.

1481
01:18:10.134 --> 01:18:11.214
That's actually crazy.

1482
01:18:11.294 --> 01:18:11.494
Yeah.

1483
01:18:11.654 --> 01:18:12.895
No, that's actually crazy.

1484
01:18:12.995 --> 01:18:13.635
It's a lot.

1485
01:18:14.916 --> 01:18:17.857
Like, in terms of per watt, like, yeah.

1486
01:18:17.897 --> 01:18:20.798
Having 100x decrease in watt, like, usage.

1487
01:18:21.558 --> 01:18:22.258
That's not trivial.

1488
01:18:22.278 --> 01:18:25.519
I mean, like, that's not so much.

1489
01:18:25.679 --> 01:18:26.259
Yeah, yeah.

1490
01:18:26.299 --> 01:18:28.160
That's so sick, actually.

1491
01:18:28.200 --> 01:18:28.660
Yeah, yeah.

1492
01:18:28.740 --> 01:18:31.841
And they proved it, you know, with their validation.

1493
01:18:31.861 --> 01:18:33.121
Obviously, you make something like this.

1494
01:18:33.181 --> 01:18:35.982
Okay, how are you validating that something like this works?

1495
01:18:36.222 --> 01:18:45.704
So they got their digital twin to train on the MNIST data set, which is the classic handwritten digits data set to identify, like, two, three, four.

1496
01:18:45.904 --> 01:18:46.505
Yes.

1497
01:18:46.665 --> 01:18:47.706
Did really well on that.

1498
01:18:48.047 --> 01:18:49.889
Train it on the MNIST fashion data set.

1499
01:18:49.929 --> 01:18:50.870
Did really well on that.

1500
01:18:51.190 --> 01:18:53.873
The other thing they did was they can really do.

1501
01:18:53.954 --> 01:18:55.015
So this this was A.I.

1502
01:18:55.055 --> 01:18:55.455
inference.

1503
01:18:55.535 --> 01:19:04.426
The other thing they can really do is combinatorial optimization, which is this like this class of NP hard problems like the traveling salesman problem and things like that.

1504
01:19:04.806 --> 01:19:13.431
that they can efficiently, they're not solving P equals NP, but they can efficiently solve NP hard problems using this technique.

1505
01:19:13.511 --> 01:19:24.076
And one of the things that they were saying was, as I was reading, was like, you know, in MRI scans, the way the MRI scans work is like the patient has to go in and be there for like half hour to an hour.

1506
01:19:24.496 --> 01:19:24.917
Because...

1507
01:19:25.837 --> 01:19:40.042
What you got to do is you got to take a bunch of different little puzzle pieces of effectively what you're doing is you're taking like scans in frequency space because of the way that the physics of the MRI works.

1508
01:19:40.322 --> 01:19:44.203
You're trying to look at the precession of hydrogen nuclei.

1509
01:19:45.384 --> 01:19:47.626
the patient as you like disturb it.

1510
01:19:47.646 --> 01:19:48.767
And then there's a magnetic field.

1511
01:19:48.807 --> 01:19:51.170
So then the proton is going to relax.

1512
01:19:51.210 --> 01:19:53.552
And as it relaxes, it releases light.

1513
01:19:53.613 --> 01:19:54.974
And you were trying to gather this light.

1514
01:19:55.034 --> 01:20:00.019
And that'll tell you some, it'll give you like a frequency map of the patient.

1515
01:20:00.039 --> 01:20:03.483
And then from that frequency map, you go into like a positional map.

1516
01:20:03.603 --> 01:20:03.803
Right.

1517
01:20:04.904 --> 01:20:10.426
So in order to do that, it takes a lot of scans in this Fourier space to then reconstruct the actual thing.

1518
01:20:10.527 --> 01:20:10.867
Right.

1519
01:20:10.887 --> 01:20:11.067
Right.

1520
01:20:11.367 --> 01:20:16.750
But there's a way to do it with something called sensing.

1521
01:20:17.070 --> 01:20:17.290
Right.

1522
01:20:17.430 --> 01:20:20.792
Like and what you can do is you can.

1523
01:20:22.148 --> 01:20:26.933
very selectively probe this frequency space.

1524
01:20:27.373 --> 01:20:30.656
And do it in a very smart way with this compressed sensing.

1525
01:20:30.676 --> 01:20:32.518
It's like a zip file for sensing.

1526
01:20:33.138 --> 01:20:35.821
And you can then reconstruct

1527
01:20:37.302 --> 01:20:37.742
The image.

1528
01:20:37.903 --> 01:20:38.083
Yes.

1529
01:20:38.263 --> 01:20:38.463
Right?

1530
01:20:38.983 --> 01:20:40.825
So they could do that.

1531
01:20:40.945 --> 01:20:42.066
That's an optimization problem.

1532
01:20:42.106 --> 01:20:42.927
Yeah, I get what you're saying.

1533
01:20:42.947 --> 01:20:43.627
At the end of the day, right?

1534
01:20:43.647 --> 01:20:44.428
I get what you're saying.

1535
01:20:44.448 --> 01:20:48.511
And the compressed sensing is an optimization problem that they could then figure out.

1536
01:20:48.771 --> 01:20:54.736
And they projected that, you know, instead of a 30-minute to an hour-long MRI, it'll take a patient five minutes.

1537
01:20:55.116 --> 01:20:57.818
And then this thing can just, like, reconstruct it pretty quickly.

1538
01:20:57.999 --> 01:21:01.461
What's an important point here is that the analog...

1539
01:21:02.382 --> 01:21:14.268
infrastructure, the analog computer infrastructure that they've built, has this application as sort of a replacement for GPUs at inference time in the AI use case.

1540
01:21:14.348 --> 01:21:24.153
Because it's good at one aspect, but it's also good at this combinatorial piece, which gives it implications in the medical field, for example, in decreasing MRI times.

1541
01:21:24.573 --> 01:21:28.275
Because it's fundamentally so much faster and better at this very specific...

1542
01:21:29.155 --> 01:21:52.286
function that is a part of the mri process that currently is using digital computers and if you just swap out the hardware infrastructure and a variety of other details to get it integrated into the mri workflow you're now just able to actually just do more with less yeah yeah and that's like super huge right yeah like now i can do like 12 patients instead of one right right that's

1543
01:21:53.813 --> 01:21:54.373
I love.

1544
01:21:54.994 --> 01:21:58.296
Yeah, so innovation is happening.

1545
01:21:58.336 --> 01:22:05.061
For everyone who said innovation is dead, watch From First Principles because we will dispel that rumor quickly.

1546
01:22:05.602 --> 01:22:09.464
We touched on four very, very fascinating stories this week.

1547
01:22:09.484 --> 01:22:10.665
We started with 3i Atlas.

1548
01:22:10.885 --> 01:22:10.986
Yep.

1549
01:22:11.006 --> 01:22:11.966
We had some new imagery.

1550
01:22:13.007 --> 01:22:13.868
It's not aliens.

1551
01:22:14.048 --> 01:22:15.229
It's still very interesting.

1552
01:22:16.109 --> 01:22:19.672
We may not live to 100, so Brian Johnson, watch out.

1553
01:22:20.052 --> 01:22:26.676
But longevity studies are showing that there is a real plateau being reached in this sort of increase in life expectancy.

1554
01:22:26.896 --> 01:22:39.425
There's huge implications for policy and economics related to how we shift the zeitgeist opinion about that to be connected to the fundamental frontier research in that area.

1555
01:22:39.845 --> 01:22:41.225
We don't have to sort plastic anymore.

1556
01:22:41.846 --> 01:22:45.107
We talked about this new nickel catalyst that enables this.

1557
01:22:45.147 --> 01:22:46.227
That was a fascinating one.

1558
01:22:46.467 --> 01:22:55.610
And last but not least, analog AI from Microsoft, which has potential to solve for the energy issue with AI.

1559
01:22:55.810 --> 01:22:55.990
Yeah.

1560
01:22:56.010 --> 01:22:57.450
But also has other applications.

1561
01:22:58.050 --> 01:22:58.190
Yeah.

1562
01:22:58.210 --> 01:22:59.371
We want to make AI better.

1563
01:22:59.771 --> 01:22:59.931
Right.

1564
01:23:00.271 --> 01:23:01.391
It's good.

1565
01:23:01.511 --> 01:23:02.012
And cheaper.

1566
01:23:02.112 --> 01:23:02.992
And cheaper, faster.

1567
01:23:03.612 --> 01:23:04.632
Better for the environment.

1568
01:23:04.672 --> 01:23:05.152
Yeah, who knows?

1569
01:23:05.232 --> 01:23:08.573
Maybe it'll solve, like, the problems.

1570
01:23:08.793 --> 01:23:08.973
Right.

1571
01:23:09.053 --> 01:23:09.453
Right.

1572
01:23:09.473 --> 01:23:13.734
If it also solves the NP problem, it might solve other problem types that we didn't have the tools for before.

1573
01:23:14.214 --> 01:23:17.835
This is what we love to do here at From First Principles.

1574
01:23:18.175 --> 01:23:20.175
Talk about this breaking science research.

1575
01:23:20.575 --> 01:23:23.156
Break it down so you don't need a PhD to understand it.

1576
01:23:23.496 --> 01:23:24.896
The implications here are crazy.

1577
01:23:25.336 --> 01:23:32.458
My name is Lester Nare, joined, as always, by my co-host and our resident PhD, Krishna Chowdhury.

1578
01:23:33.661 --> 01:23:34.859
all next week.
