Latest / From First Principles / JWST’s “Little Red Dots,” TimeVaults, and the Dawn of Math (EP. 23)

Transcript

JWST’s “Little Red Dots,” TimeVaults, and the Dawn of Math (EP. 23) · From First Principles

0:00
Hello, Internet.
0:00
This is your captain speaking, Lester Nare, joined as always by my co-host and our resident PhD, Krishna Chowdhury.
0:08
We have three great stories lined up for you today, along with the rundown.
0:13
We're going to hit astrophysics, genomics, ancient mathematics, couple of housekeeping items at the top of the show.
0:22
Our new collaboration episode with
0:30
The episode is focused on the future of intelligence.
0:32
It's part one of two.
0:33
It's a great conversation.
0:35
Definitely worth a watch.
0:37
And at the end of this story, we're going to do a follow-up from episode 22 as it relates to the Cloud 9 story that we covered, the gas cloud.
0:46
Yeah.
0:46
Because we got some interesting insight about the naming conventions and some of the background from the scientists who put the paper out.
0:53
Yeah, that was really cool.
0:55
We sent out the video to the scientists, the authors of that study, and they were gracious enough to give us some of the backstory.
1:02
So it's really interesting.
1:05
As always, you're going to learn something today because this is from First Principles.
1:25
For our first story, we're going to dive into the James Webb Space Telescope in an astrophysics story that was a research paper published in Nature about the little red dots as young, supermassive black holes in dense ionized cocoons.
1:44
This is a paper out of the University of Manchester, University of Copenhagen, and the Cosmic Cove.
1:50
Dawn Center.
1:52
And it's a nature study that reveals that these young supermassive black holes are hiding in dense cosmic cocoons that we were able to sort of get more insight from because of this high fidelity, high quality JWST spectra.
2:07
Yeah.
2:07
So what's this story all about?
2:09
Yeah.
2:09
Yeah, so I don't know if you remember, but back in July 2022, JWST came online.
2:15
I was traveling in Europe, and it was a whole frenzy.
2:17
And within two weeks of when the data was released, we had headlines like this from the Scientific American that said, the first glimpses of early galaxies could break cosmology.
2:29
You know, JWST was seeing universe breakers.
2:32
And it was like...
2:34
All of astrophysics was, you know, oh my God, what's going to happen?
2:39
What's going to happen?
2:40
And what it was all about was there were these spots of red light everywhere at really high red shifts.
2:46
And if they're really far away, that means that they're really old.
2:49
But that also means that they're very young for the universe, right?
2:53
Because the universe was only less than a billion years old and we were seeing galaxies.
2:58
And in order to make galaxies, you need a lot of time, right?
3:01
And so...
3:02
Even there was like creationists that were like going gangbusters saying that, oh, this is proof of God created the universe and all of that stuff.
3:12
So the focus of our study is something called little red dots.
3:17
And that is the technical term.
3:19
Is it really?
3:20
It is literally the technical term.
3:22
These are little red dots in the JWST deep field surveys.
3:27
And they're at a really high redshift, which means they're very young in terms of when the universe was born.
3:33
So this is about a billion years, less than a billion years after the Big Bang.
3:38
And the problem is they're really compact.
3:41
And the debate is if these are star-filled galaxies...
3:45
The galaxies shouldn't have had enough time to become galaxies.
3:52
Right.
3:52
Okay?
3:53
Because it takes time to form stars and then from those stars to amass and become a clustered, gravitationally attracted object like a galaxy.
4:01
And this idea is based off of our current cosmological models of how the early universe, the mechanics and the dynamics of the early universe.
4:10
Yes, yes.
4:10
Based on how much we know dark matter is there, how much we know dark energy is there.
4:15
In the beginning, there was not a lot of dark energy because the universe was small and dark energy scales with how big the universe is.
4:23
So in the beginning, there was really just a lot of dark energy.
4:26
Sorry, there was a lot of dark matter and a lot of normal matter, but more dark matter than normal matter.
4:32
And those...
4:34
models of our universe tend to not have galaxies so young.
4:40
Right.
4:41
Okay.
4:41
So these became these universe breakers.
4:43
And then the debate was, okay, maybe they're not star filled galaxies, but maybe they're just weird, massive black holes.
4:48
Okay.
4:49
And you would think, okay, like maybe that's going to solve this universe breaking thing.
4:53
Okay.
4:54
No, it still doesn't because those black holes are, again, too big for how young the universe was.
4:59
Right.
5:00
So it was a real conundrum for a while, and this paper might have the answer.
5:04
Okay.
5:04
That's the idea.
5:05
Beautiful.
5:05
And so it's quite a big deal if it, you know, holds up to scientific scrutiny.
5:10
Because it's both, it's going to solve this new problem that has arisen as we've gotten higher quality data that's giving us...
5:16
experimentally in the real world information that doesn't quite map onto our existing models of the early universe.
5:23
Exactly.
5:24
And that's why, I mean, people were so excited when the James Webb Space Telescope went online.
5:30
This is exactly what it is.
5:31
Because we knew there was going to be all sorts of random stuff that we were going to see that we'd never seen before.
5:36
And, you know, we're at the forefront of technology.
5:39
It was launched in 2021, late 2021.
5:42
It came online in July 2022.
5:46
And these deep space fields were crucial.
5:49
This is a photo of the James Webb.
5:51
It was folded up.
5:53
Incredible engineering feat, by the way, to even get such a scientific instrument to be folded up inside a rocket and then have it go to a Lagrange point.
6:02
So not even in Earth orbit, really far away from Earth orbit.
6:07
Get it parked there.
6:08
Have it unfold like origami.
6:11
No issues.
6:12
And Biden made an announcement being like, we got this amazing data.
6:17
And two weeks after that amazing data was published, everyone was like, what's going on?
6:22
Right?
6:24
So one of the crucial things about the James Webb Space Telescope is it is an infrared telescope.
6:29
Unlike the Hubble, which is optical, the James Webb is an infrared telescope.
6:33
And the reason why it is sensitive in infrared is because infrared is wavelengths of light that are longer than optical.
6:41
And we want it to be that way because...
6:44
We want to look at the very depths of the universe, okay, really, really far back.
6:50
And because the universe is expanding, you have something called a gravitational red shift.
6:55
Actually, no, it's called a cosmological red shift.
6:59
What ends up happening is light that is coming from really far away...
7:04
As the universe stretches, you can imagine a little sine wave, and then I stretch that sine wave like a spring.
7:10
So the wavelength is going to stretch out.
7:12
So if I'm in infrared, I can catch these highly red-shifted light.
7:19
And that's exactly what we see in this photo.
7:21
And we see as we get...
7:23
farther and farther into the universe, closer and closer to the Big Bang, the spectra, these unique fingerprints where oxygen should be emitting light or where hydrogen should be emitting light, that stuff gets stretched more and more into the infrared.
7:37
Mm-hmm.
7:38
The James Webb being an infrared telescope means that we are sensitive to that infrared.
7:44
Got it.
7:44
Okay?
7:45
Right.
7:45
So we sort of have, instead of seeing a fuzzier image of things kind of blending, there's more discreteness in what we're looking at because it's...
7:54
We're not limited to light has to travel and things are getting squished at these scales.
7:59
Stretched.
8:00
Sorry, stretched at these scales.
8:02
And so we need to expand what we're looking at in order to capture just more clarity.
8:07
Yes.
8:07
And understanding what it is we're looking at.
8:08
Yeah, we want to be sensitive to that correct part of the spectrum, right?
8:13
That makes sense.
8:13
And so that's where these little red dots come in.
8:15
And they're literally called little red dots, okay?
8:18
Yeah.
8:18
Weird thing about them is they show up all over the place, about 600 million years to 800 million years after the Big Bang.
8:27
But then they vanish out of existence 1.5 billion years after the Big Bang.
8:31
And we don't see any of them.
8:33
It's kind of like dinosaurs where dinosaurs were alive for 300 million years to 65 million years before us.
8:40
And then they're just gone.
8:42
So these little red dots are like dinosaurs from the early universe that are no longer there in any nearby galaxies.
8:50
So it's very weird.
8:51
Yes.
8:52
Right?
8:52
Right.
8:52
That is curious.
8:53
The other thing about them is they're really small.
8:56
Okay?
8:56
They're really, really small.
8:57
If you look on the right-hand side, on the left-hand side, we've got one of the JWST deep fields.
9:02
And on the right-hand side, we have a close-up of one of these red dots.
9:06
And if you look really closely, that red dot has a sort of hexagonal hue around it.
9:12
That hexagonal hue is an artifact of the James Webb Space Telescope's design because the
9:19
The James Webb has a primary mirror and then a secondary mirror that sort of the light comes in primary, goes to the secondary, and then goes to the detector.
9:27
And the scaffolding that is holding up that secondary mirror has three sort of...
9:36
rods that hold it holds it up so those rods are going to cause something called diffraction spikes okay they're just artifacts of the fact that the secondary mirror has these three rods holding it up in the hubble there's four and so you see for the hubble hubble photographs the diffraction spikes are yeah right in diamond shape but in the james webb they'll be in hexagonal shape makes sense okay so
10:02
But the fact that we're seeing those diffraction spikes for those little red dots means that they're localized point sources.
10:08
And we can actually calculate how big they are, and they'd be about 150 to 500 light years across.
10:13
That's light years across.
10:15
Light years.
10:16
Right?
10:17
The Milky Way is on the order of 100,000 light years.
10:21
Okay.
10:21
Okay?
10:22
The Milky Way is much bigger than these things by a factor of 100 to 1,000.
10:26
Okay.
10:26
Okay.
10:27
So perhaps they are just like black holes.
10:29
This is the first clue that maybe they're black holes.
10:32
They're really compact objects, right?
10:35
They have a lot of activity and a lot of mass.
10:37
So, yeah, seems like a black hole.
10:39
It tracks.
10:39
Okay, so if it's a black hole, we'd like to calculate its size.
10:44
Right.
10:45
Okay, how do we calculate a black hole size?
10:47
Well, in the Milky Way, the way we calculated the black hole size is we had Andrea Ghez from UCLA.
10:58
path of stars that were nearby and then you just do newton's laws right and you figure out how big is the compact object on the inside now if we've got a black hole that is at the edge of the universe 13 billion light years away how are we going to do it well we use something called the doppler broadening of spectral lines okay so
11:19
Effectively, what's happening is this.
11:21
We look at a certain spectra, let's say H-alpha, which is an intrinsic piece of light that comes out of hydrogen.
11:28
Now, if the hydrogen is not moving at all and it's static, that line is going to be really sharp.
11:34
But in a black hole, there's going to be stuff moving around in an accretion disk.
11:38
Right.
11:38
So there's going to be hydrogen that's moving away from us.
11:41
Yes.
11:41
And hydrogen that move that's moving towards us.
11:43
Yes.
11:44
So there's going to be some bits of H alpha, the light that is going to be blue shifted.
11:49
So the wavelength will be shorter.
11:50
And then there's going to be some bits that are red shifted because it's moving away and the wavelength is going to be longer.
11:56
And what's that going to what that's going to cause is my H alpha line, which used to be just a really well defined frequency.
12:02
It's going to broaden.
12:03
I see.
12:03
Right.
12:04
And that's called Doppler broadening.
12:05
This is what we're seeing here.
12:06
Yeah.
12:07
And that's what we're seeing here.
12:08
We get a nice Gaussian profile of our spectral line.
12:11
And that Gaussian profile from there, we can then back calculate how fast does the gas need to be moving around to create that broadening.
12:21
Right.
12:21
And then you can apply the Virial Theorem, which is something in basic physics.
12:25
The mass is proportional to the velocity squared.
12:28
And from that, you get something like a number that is 10 to the 7 to 10 to the 9 times the mass of the sun.
12:35
So it's like, you know, a billion to 100 million, tens of million times the mass of the sun.
12:40
This is for something that is moving at 2,000 kilometers per second.
12:44
That's what we can calculate back from the broadening is the velocity.
12:47
And from the velocity, we calculate back the mass.
12:51
Got it.
12:51
Okay.
12:52
So we're getting 10 to the 7 times or 10 to the 9 mass of the sun.
12:57
As the scale of these.
12:59
Of how big these black holes should be.
13:00
Should be.
13:01
Right.
13:01
How big they should be.
13:03
Okay.
13:04
So you would think that, okay, we're good, right?
13:06
Like, they're not galaxies.
13:07
They're just black holes.
13:09
Not so fast.
13:10
Okay.
13:11
There's two problems with these masses.
13:14
For one thing, these black holes are way too big for...
13:19
whatever galaxy this little red dot is.
13:21
If this little red dot is a galaxy and at the center is this active galactic nuclei that's, you know, churning up a bunch of mass and then spewing it out as radiation, then those black holes are about 10 to 100 percent the host stellar mass.
13:37
Okay?
13:38
In contrast, for our Milky Way, the center...
13:41
of the galaxy where the black hole is that black hole is only 0.1 percent of the entire mass of the milky way so this is the math is not math the math is not mathing okay sagittarius a star at the center of our milky way is only 0.1 percent of the total mass of the milky way and here i'm getting 10 to 100 the total mass like the whole thing is a black hole they're way too massive yeah
14:05
It doesn't make any sense.
14:05
Okay.
14:06
Okay.
14:06
Yeah.
14:07
The second thing is, okay, so you're telling me that these black holes are 100 million times the mass of the sun, right?
14:14
How did it get that big?
14:15
Well, the way that black holes get big is through accretion, right?
14:19
It's like taking in mass and it's eating up mass.
14:23
And as it eats up all this matter, it gets bigger and bigger.
14:27
There's a caveat to it, though.
14:28
Okay?
14:29
It can't just, like, eat endlessly.
14:31
It's not like those video games where you just eat the whole city around you.
14:34
Yeah, in like a single moment, you can't do that.
14:38
You've got to take your time with your meal.
14:41
That's the point.
14:41
There's something called the Eddington Limit, okay?
14:44
And it says there's a maximum luminosity to accretion rate.
14:48
Before I'm eating too much, and when I eat all that much, I'm going to spew out radiation.
14:55
That radiation is going to shock all of the rest of my food away.
15:00
You see what I'm saying?
15:01
It's the same argument of hydrostatic equilibrium that we've been visiting a lot on this podcast.
15:07
The idea is if I eat too fast...
15:11
Then the radiation from all of that eating, because the black hole is going to accrete all this mass, the mass is going to start heating up.
15:19
That's going to cause radiation.
15:21
That radiation is going to be enough to shoo all of the rest of my mass away.
15:26
So there's a maximum limit at which I can grow.
15:29
3,000 calories a day max.
15:31
Max, right?
15:32
And if you do too much, then there's no food around you.
15:34
Right, right.
15:35
Okay, that's the idea.
15:36
Look, that's very interesting.
15:37
Okay, so that's something called the Eddington Limit.
15:39
It's named after Arthur Eddington, who's an amazing man of science in history.
15:45
Also a bit problematic.
15:47
And by a bit, I mean a lot.
15:49
That's for another episode.
15:52
In any case, he finds this sort of limit to how how much stars can grow.
15:57
What's the luminosity of these stars?
15:59
And you can apply that same logic to black holes.
16:01
Got it.
16:02
OK, so from that, we get an accretion rate.
16:06
It's like, how much am I eating?
16:07
Yes.
16:08
And from that, I can have a characteristic growth timescale.
16:11
of like how how much time do i need to grow by a certain factor because you know what your limit on eating is and so you'll know if i'm trying to gain 10 pounds or lose 10 pounds yeah how long is that going to take right because you're let you have a max yeah and and for this by a factor of e because everything is an exponential exponential to the e so if i want to grow by a factor of e which is like 2.7 ish i need to wait 50 million years yeah all right sure 50 million years to
16:41
All right.
16:42
Now, these red dots that we see in the universe, they're at a red shift of seven, which means they're approximately 800 million years after the Big Bang.
16:51
800 million years is not a lot of time.
16:54
No.
16:55
Even especially for growing at only 50 million years per factor of 2.7.
17:00
Right.
17:01
OK.
17:01
Right.
17:02
So there's two possibilities.
17:04
One is you've got these light seeds.
17:06
Light as in like light instead of heavy.
17:09
Okay.
17:09
You have the first stars.
17:11
These are population three stars.
17:12
They're the remnants of...
17:14
the very first stars, when the very first stars that are, you know, about, let's say a hundred mass, a hundred times the mass of the sun or a thousand times the mass of the sun, very, very old stars.
17:24
When they blow up, they're going to make a black hole.
17:26
And that black hole is going to be about a hundred times the mass of the sun.
17:30
And so in order to reach 10 to the nine times the mass of the sun in 800 million years, you got to be continuously eating at the Eddington limit.
17:40
I see.
17:41
Okay.
17:41
Stuffing your face, stuffing your face, uh,
17:44
For like hundreds of millions of years.
17:47
You got to be going like the hot dog eating competitions.
17:51
Have you seen those?
17:52
Kobayashi.
17:53
Yeah, dude.
17:55
So you got to be in the zone for hundreds of millions of years.
18:00
Constantly.
18:00
Constantly.
18:01
Constantly.
18:02
very improbable seems yeah seems improbable to just be going at it yes right okay so the second possibility is there's heavy seeds which is this direct collapse you don't make a star you just got a gas cloud the gas cloud maybe runs into another gas cloud reaches critical mass and then the whole thing just falls inside the schwarzludge radius you get a black hole no star needed for that
18:27
The advantage is you can reach that high mass, 10 to the 9 mass of the sun.
18:33
You can reach that.
18:34
But the condition's got to be perfect.
18:36
You got to have really pristine gas.
18:38
It's got to be like sort of spherically symmetric.
18:41
And you need a lot of UV background so that the gas doesn't cool down to form stars.
18:47
Right.
18:47
It has to like really just be focused on making a black hole.
18:50
Which is...
18:51
Improbable.
18:52
Right, right.
18:53
So the idea is the light seed and heavy seed options don't pass the sniff test.
18:59
Yeah, they really don't.
19:00
It's like both are like super improbable.
19:02
On the one hand, you have like a black hole that is just eating max out.
19:07
like in the zone.
19:08
And then the second, the conditions to create that first black hole are like perfect.
19:14
Okay.
19:15
And the number of these little rod dots that we're seeing, it just doesn't make sense.
19:18
There's too many.
19:19
There's too many for all of them to have just this perfect, uh,
19:23
like condition the idea is we're seeing so many of them so you're saying this thing that's improbable is happening like frequently yeah which is the whole point yeah so what are we doing all right so so this was the big conundrum right yes so what can we do well let's point an x-ray telescope at it so they had the chandra x-ray telescope pointed at it no x-rays that's weird okay black holes
19:45
spew a lot of x-rays.
19:47
Okay.
19:47
Because stuff gets really hot, and if stuff gets really hot, it's going to, like, spew out very high-energy radiation.
19:54
It also...
19:54
There's no radio emissions.
19:56
You look at... You look at it from...
19:58
the earth with really nice radio telescopes that can localize that point source, no radio emissions.
20:04
Black holes should have radio emissions because there's high magnetic fields.
20:07
And if charged particles are spinning around in that high magnetic fields, it's going to shoot off.
20:12
They're going to shoot off radio waves.
20:15
So no x-rays, no radio waves.
20:17
What do we have?
20:17
What do we have?
20:18
Well, let's look at the photo.
20:20
What do we have?
20:20
Well, the James Webb Space Telescope has a really nice spectrograph, right?
20:24
So we can look at the photometric spectra, and this is from a particular little red dot called the cliff.
20:30
The cliff.
20:31
And the cliff, you can see the cliff, right?
20:33
Yeah.
20:34
So at shorter wavelengths, you've got a little bit of light coming through in the ultraviolet.
20:38
Yeah.
20:38
And then right at about 300 and, like, 380 degrees,
20:45
nanometers, you get a really high spike.
20:48
Everything that's over that 365 nanometers...
20:54
You get a lot of light coming in.
20:56
Yeah.
20:56
Okay?
20:56
So you get this, like, distinct V-shape.
20:59
Yeah.
20:59
This particular one is called the cliff because it's such a nice spectra.
21:03
Yeah.
21:03
Okay?
21:04
So the question is, what would actually cause a spectra like this, where you've got a little bit of ultraviolet, but right at 365, there's nothing, almost nothing.
21:15
And then for any wavelength longer than 365, you've got a lot of light coming through.
21:20
Well, 365 nanometers, and this is in the rest frame, so they had to shift everything to suppose this thing wasn't moving away.
21:29
Because in reality, that light is going to be somewhere in the infrared, right?
21:33
Because it's moving away.
21:34
But you can shift it back into the rest frame.
21:36
And you can say, okay, at 365, there's nothing.
21:39
And then anything higher than 365 nanometers, which means lower in energy because higher wavelength, lower energy.
21:48
We're getting a lot of light.
21:49
Let there be light at 365.
21:51
So 365 nanometers is something called the Balmer limit.
21:55
Okay.
21:56
This is what happens because any piece of light that is higher energy than the 365 nanometer wavelength...
22:06
it ionizes hydrogen.
22:08
That photon has enough energy to kick out an electron if the electron is in the second shell of hydrogen.
22:15
Okay?
22:16
So any photon that is in lower energy, meaning longer wavelength,
22:21
That's not going to be able to kick out that hydrogen, and it's not going to be able to ionize.
22:26
And this is actually something that we see a lot in stellar atmospheres, because stellar atmospheres have a lot of ionized hydrogen.
22:33
This is a blueprint of a blue supergiant spectrum.
22:37
And there you can see, right at about that Balmer limit, you're going to get a massive discontinuity.
22:43
Yes, absolutely.
22:44
Where there's a bunch of stuff at a higher wavelength because those photons don't have enough energy to ionize hydrogen.
22:51
But the stuff that's with a smaller wavelength, so higher energy, those things ionize hydrogen no problem.
22:58
And the electron goes off and does whatever it wants to do.
23:01
Right?
23:02
So that's the key.
23:03
The key here is that the ionized hydrogen around a star could be causing this.
23:11
Okay.
23:12
Okay?
23:12
Yeah, yeah, yeah.
23:13
So the key inside of this paper is to say, all right, suppose I were to take that ionized hydrogen idea and I were to put that ionized hydrogen envelope around a black hole.
23:26
I see.
23:27
What would that mean?
23:28
What would that mean?
23:29
Okay.
23:30
And that's the key inside of this paper.
23:33
It's funny because we've been talking about reionization and hydrogen in astrophysics, cosmology, and space in a couple of our previous episodes.
23:44
So folks will have a good grounding for this conversation based on several of our past stories.
23:50
Yeah.
23:51
And this paper came out in, I think it was Nature.
23:54
We have the...
23:55
Yes.
23:55
It was Nature, January 13th, 2026.
23:59
Right.
23:59
So very recently, and this is photo 13 just to show our folks what the paper looks like.
24:05
Little red dots as young supermassive black holes in dense ionized cocoons.
24:10
That's where the ionized cocoons comes in.
24:12
It's an ionized cocoon of stuff around my black hole.
24:17
And the key is that that hypothesis...
24:20
then lays to rest all of the different conundrums that we've had.
24:24
I see.
24:25
Okay.
24:25
I see.
24:26
Like when we talked about earlier about it's overmassive.
24:29
Yeah.
24:29
And it's big.
24:30
The thing is too big.
24:32
The Eddington limit.
24:33
So we can't accrete enough to like be able to get to that size.
24:37
But the note here is that this, the ionized hydrogen is,
24:43
envelope that's around it is what is making what we're measuring or seeing look the way it is and so perhaps these black holes aren't as big as we thought they were i see i see that's the idea okay okay okay so the
24:57
The inside of the ionized gas cloud means we can now recalculate the mass of these black holes.
25:03
Because the line broadening, before we thought the line broadening just had to do with the hydrogen spinning around the black hole.
25:11
And then from that we calculate what the mass should be for it to spin around at that rate.
25:16
Okay.
25:18
Well, if there's ionized hydrogen, then what we're going to see is not the actual photons from the accretion disk, but we're going to see those photons get released, and then they're going to bounce around in the ionized hydrogen, because the ionized hydrogen is just plasma.
25:32
And plasma means that light doesn't really have a free path to travel all the way to us.
25:38
It's going from here to there, and so the photons undergo this random walk,
25:44
of free electrons in a highly ionized environment.
25:48
So you've got the black hole in the center, and it releases a Gaussian profile because of the accretion.
25:54
But by the time it gets to us, by the time it gets to the James Webb, it's going to be bouncing around this hydrogen atom, then this electron, then this hydrogen, then this electron, electron, blah, blah, blah.
26:06
And you're going to get this Thompson scattering.
26:09
the free electrons are going to scatter around.
26:12
Yes.
26:13
Okay?
26:13
Yes.
26:14
And when that happens, the Gaussian profile is actually going to resemble more of an exponential decay on both sides.
26:22
So the way to test this...
26:24
And in physics and in science, really, a lot of times what it comes down to is can you get good enough data to discern one theory from another?
26:34
So the two theories that are competing in this case are it's a Gaussian profile, so it's a bell curve, and the other one is it's a...
26:42
Exponential decay.
26:43
Yes.
26:44
The James Webb telescope is giving us good enough data where we can discern between those two theories.
26:51
And that's why it's so important to get really good data.
26:55
Which means we have to build really great instruments.
26:57
Yeah.
26:58
And the James Webb is just a phenomenal instrument.
27:02
And so here you can see from their paper, they...
27:05
fit an exponential and then they fit a Gaussian.
27:08
And the errors when it comes to fitting an exponential are basically zero.
27:12
That's the teal color.
27:13
And then if you look at the purple, there's plenty of error.
27:16
Yeah.
27:17
Yes.
27:17
The exponential fits this profile a lot better.
27:21
The model, the exponential model of the theory to explain why we're seeing this
27:28
what we believe to be these black holes look the way they do, this ionization of hydrogen, which then ends up having this exponential decay on this planet.
27:40
That...
27:41
That we did theoretically on paper.
27:44
Yeah.
27:45
Yeah.
27:45
We're like, okay, if there's a cocoon of hydrogen that's ionized, so a bunch of plasma that's like diffuse.
27:51
Yes.
27:51
What would it look like?
27:52
Well, it would look like an exponential.
27:53
So let's fit an exponential.
27:54
And then we've got the experimental data of high fidelity from JWST that now it's like, oh yeah, that's exactly what we're seeing.
28:00
That's exactly right.
28:01
Yeah.
28:01
And now instead of 2000 kilometers per second.
28:05
Yes.
28:05
That the accretion disk is going around.
28:07
Yes.
28:08
Instead...
28:09
it's something like 300 kilometers per second.
28:13
So actually, the accretion disk isn't moving all that fast.
28:16
The broadening is happening because of the plasma that's all around.
28:20
That's surrounding it, not the actual internal structure of the black hole.
28:23
Exactly.
28:24
So I guess the intuition is then that means the black holes are much smaller and much younger because they're not super massive and they're not going very, very quickly.
28:34
That's all being sort of obfuscated by the amount of...
28:38
plasma ionized hydrogen that's around this really young small black hole that's exactly right okay yeah and and so when you when you redo the estimate for the mass you get something like 10 to the 5 which is you know several orders of magnitude but below right what we had earlier right
28:55
And now this makes sense.
28:56
Yeah, now it fits.
28:57
Now it fits.
28:58
It's like now you can have that light seat of the population three stars.
29:01
But this guy doesn't have to constantly eat like a hot dog contest for hundreds of millionaires.
29:07
He can take breaks.
29:08
He can go sub-Eddington limit, and he'll be fine.
29:13
He'll make it to this far.
29:14
You know?
29:16
And given the volume of these that we're seeing, it then also matches the frequency of its appearance.
29:24
This is probable now.
29:25
Yes.
29:26
And then now it kind of explains why there's so many of them.
29:29
The missing X-rays also make sense because all of these X-rays are now going to get absorbed by the plasma.
29:35
Right.
29:35
So we don't see the X-rays.
29:36
The red color makes sense because the cocoon absorbs the high energy light and then re-emits it in the lighter part of the spectrum, the redder part of the spectrum.
29:45
I was going to ask earlier.
29:46
I'm like, I thought black holes were like black holes of light, but I get the accretion disk and then stuff happens.
29:51
I'm like, why is it?
29:52
Why was it red?
29:53
Yeah.
29:54
Yeah.
29:54
But it's just, I mean, the stuff around the black hole is so hot that it's releasing a bunch of light.
30:00
Right.
30:00
Which makes sense.
30:01
And then it also explains the black hole mass problem because now...
30:04
You don't need this accretion.
30:06
And the host mass, you know, before we used to think, oh, this thing is like 10% to 100% of the host galaxy.
30:13
As opposed to the 0.1 we see.
30:14
Yeah, yeah.
30:15
And now you can see the final figure in their paper is actually showing.
30:19
So the...
30:21
The gray circles are what it would be without their hypothesis.
30:27
And you can see the mass is decreasing because they use this exponential fit.
30:31
And they're saying, actually, these black holes are not that massive.
30:34
And the blue line is what the theory says.
30:36
Right.
30:37
And it's like lining up very nicely.
30:39
Yeah, yeah, yeah.
30:40
Right?
30:40
Yep.
30:41
Yes.
30:41
Now, there's caveats.
30:43
They only analyzed 12 of these.
30:45
And we've seen, you know, several hundred.
30:48
Okay.
30:48
So it could be that there's a bunch of different things.
30:51
Okay.
30:52
And it's not like a one answer fits all little red dots, right?
30:58
Some of them could be young galaxies, and then there's always more questions.
31:02
But this at least puts a lot of the black hole theories.
31:05
Yes.
31:07
To rest.
31:07
To rest.
31:08
Like, it's making sense now.
31:10
Yes.
31:10
Okay?
31:11
The universe did...
31:12
The Big Bang did start 13.6 billion years ago.
31:16
And right after the Big Bang, within a billion years, there were black holes, but they weren't that big.
31:21
It's fine.
31:21
It's fine.
31:22
But because there was so much gas around at the time, you could have these ionized cocoons.
31:27
Nowadays, you can't because the gas is so sort of diffuse.
31:30
Yeah, we have Pepto-Bismol now, so it's fine.
31:32
Exactly.
31:33
The universe is 13.6 billion years old.
31:36
The universe needs some Pepto-Bismol, you know?
31:39
But I thought that was really cool.
31:40
Just the way that, you know...
31:42
It's a recurring theme in astrophysics research, actually, that the closer you measure something, the better you can discern between two competing theories.
31:53
Right.
31:53
Right.
31:54
Right.
31:54
And, like, it goes all the way back to Kepler and Tycho Brahe.
31:58
Tycho Brahe was this...
32:01
really rich guy who had a very nice observatory, not with lenses and stuff, but an observatory to point out locations of planets.
32:10
And he tracked the locations of planets to a really high degree of accuracy.
32:15
And then Kepler, who was the mathematician, took that data and he was trying to fit Copernicus's model of circles.
32:23
And he's like, the circles don't fit, but the ellipses do fit.
32:26
Yeah.
32:27
And the difference between the ellipse and the circle was really small because the orbits of the planets are basically circles.
32:34
For sure.
32:35
But they're more ellipses than they are circles, right?
32:37
Yeah.
32:38
But there's a very famous quote by Tycho Brahe.
32:42
I'm going to paraphrase.
32:42
He said that this error is too big for Tycho Brahe to make.
32:47
Yeah.
32:48
Yeah.
32:49
Right?
32:49
Like the error between the circle and the ellipse is big enough that I'm pretty sure Tycho Brahe's error bar is smaller than the difference between these theories.
32:58
So I'm pretty sure these things are ellipses.
33:00
Yeah.
33:01
And then he came up with Kepler's three laws and then, you know, Newton ran with it.
33:05
So it's an amazing like sort of it's the tradition holds true.
33:09
The better data you get, the more you can discern between one or the other theory.
33:13
This is why we need to continue to support and invest in frontier research.
33:19
It's not necessarily tied to commercial impact.
33:22
No.
33:23
Because the better instruments will eventually somehow, someway, always is dual use.
33:28
Yeah.
33:28
It's always dual use.
33:30
The better we understand the world around us, the better we can get all the conveniences we love every day.
33:34
Yeah.
33:35
And this is...
33:36
I mean, it was just so cool that now we know that there were black holes...
33:41
800 million, like an infant universe had just a bunch of black holes that were just going gangbusters.
33:49
Yes, and then started spitting out and creating all of this.
33:52
And again, we can see it.
33:54
We can literally...
33:56
See it now I still get so My mind still gets so weirded out By the idea that Because of the speed of light when we look far enough It's like looking in the past Oh dude it's still weird to me It's very hard for my brain It makes logical sense Yeah of course But then you're like Are you sure Yeah
34:20
That's our astrophysics story of the day.
34:22
I think we're almost averaging one per week at this point.
34:27
Great story one.
34:29
We're going to move now into the rundown before we get to story two.
34:33
We can't cover every story happening in the frontiers of research labs, not only in the U.S., but across the globe.
34:42
It's an incredible time, and it's really important that we kind of keep track of what's going on because these things are eventually going to impact our lives one way together.
34:51
or another.
34:53
And we're going to start the rundown.
34:54
These are some fun stories this week covering a couple different areas.
34:58
Our first story in the rundown is about a prehistoric wolf's gut.
35:03
Yeah.
35:04
Frozen in time and revealed now.
35:07
An Ice Age giant.
35:09
This is from the Center of Paleogenetics in Stockholm, Sweden and Stockholm University that was published in Genome Biology and Evolution.
35:19
The summary here is that a wolf pup that died 14,000 years ago, close to when the woolly rhino went extinct.
35:29
Went extinct.
35:29
And there's some hair and stuff left.
35:31
Yeah, that's still in the stomach.
35:33
It's still in the gut, which means that the DNA apparently is well-preserved.
35:38
Yeah, because apparently the wolf died pretty soon after having dinner, this wolf pup.
35:45
And they don't think that it died from predation or something like that.
35:48
They think a landslide just came and knocked its home.
35:51
And so it had just had dinner, and then the wolf pup died along with its sister.
35:55
And it was found right next to each other.
35:57
And what's really cool is the...
36:00
the hair and all of this biological material was actually intact in the gut of this wolf pup.
36:08
So from there, we can do genetic sequencing and find the genetic remnants of this woolly rhino.
36:17
And the key thing is, if the woolly rhino went out because the population dwindled, and there were not that many viable populations,
36:27
individuals in the population then what you would expect is a lot of inbreeding and that would leave markers in the genomics but what they found was there was no evidence of inbreeding which means that there was a healthy population even as recent as 14 000 years ago and that's right around the time when all of them died off and that's also right around the time when the ice age ended right so it's probably due to climate change that
36:53
This massive, massive species that used to roam Siberia and Canada is now gone.
36:59
I didn't even know the woolly rhino was a thing.
37:02
Yeah, it's pretty crazy.
37:03
It's a rhino, but like... Like hair.
37:06
With hair.
37:06
I mean, it's a mammal, so it's got hair, but this has a lot of hair.
37:09
Because it lives in the winter.
37:12
I just found...
37:14
This might be totally nonsense.
37:15
Mm-hmm.
37:16
a pig can become a wild boar if you leave it in the wild like like like what like there's it's it's about the environment that transitions them from being like a generic domesticated domesticated pig versus like what we think of as a wild boar like they're not necessarily different geni or oh really i this look someone in the comments please correct me if i'm wrong
37:42
I saw it scrolling, and I was like, this makes... That's pretty crazy.
37:46
That means that certain genes turn on and off based on environment.
37:49
Right.
37:50
Wow.
37:50
Which, story for another day.
37:52
Yeah.
37:53
Number two in the rundown.
37:55
And we're going to be doing a lot of this Artemis story because we're coming up on our return to the moon.
38:01
NASA rolled out an 11 million pound moon rocket in preparations for the astronauts launch.
38:08
There's no journal or institution here.
38:11
One of our favorite government agencies, NASA moved a 322-foot-tall rocket out to the launch pad on Saturday.
38:19
A key step as the agency prepares for its long-awaited mission to send four astronauts around the moon on the widest orbit.
38:27
It's been around the moon.
38:29
Not on the moon.
38:30
We have been to the moon, but not at this widest orbit.
38:33
There were so many comments.
38:34
Oh, yeah.
38:35
Guys, it's fine.
38:37
It's fine.
38:39
We've been to the moon, but now we're going really far away.
38:42
That's the idea.
38:43
That's the idea here.
38:45
We did, I think, have someone in our comments that was mentioning that they were a rocket scientist and they were actually driving down.
38:53
Yes.
38:54
To basically get ready for the launch.
38:56
It was really cool.
38:57
Some of them, like they worked at Huntsville, Alabama, which is involved with the Artemis program.
39:02
Yes.
39:03
It's very cool to get feedback like that.
39:05
Um, sorry for not immediately being able to pull the name, but comment again and we'll make sure to get your name next time.
39:12
Uh, we will have a deep dive on, uh, the Artemis launch and everything soon.
39:16
So we will take care of all the moon people.
39:18
Want more moon info.
39:20
Story number three is about teenagers.
39:24
Yeah.
39:24
Scientists have found hidden Synap hotspots in the teen brain.
39:30
This is out of Kyushu University in Japan.
39:34
It was published in Science Advances.
39:36
Scientists have discovered that the adolescent brain does more than prune old connections.
39:41
During the teen years, it actively builds dense new clusters of synapses in specific parts of neurons.
39:47
These clusters emerge only in adolescence and may help shape higher level thinking.
39:52
When the process is disrupted, it could play a role in conditions like schizophrenia.
39:59
Yeah, this was very exciting to me because, you know, by normal...
40:04
neurological dogma.
40:06
We grow a lot of synapses as we get into our teenage years.
40:10
So like when we're children, when we're toddlers, and then during teenage years, the sort of main understanding is the synapses that don't have a lot of traffic, synapses are ways to connect from one neuron to another.
40:21
And when those synapses don't have a lot of traffic, those synapses get deleted, right?
40:27
This is saying that actually in certain parts of the brain,
40:31
new synapses form.
40:33
Not new neurons, but neurons are making new connections while we're teenagers.
40:38
This kind of
40:40
tracks with how i was as a team you know um specifically they looked at layer five neurons and the cerebral cortex cerebral cortex is the the neocortex part of our brain that's the newest evolved for us mammals the experiments were done in mice and layer five neurons are very cool because these are the neurons that collect a lot of data from the other layers and then send it out to other different parts of the brain so this is very important for trafficking information
41:07
and making sense of different sensory stimuli or different ideas and so on and so forth.
41:12
The authors do mention that this is experiments in mice.
41:16
And, you know, not everything that happens in mice translates to primates and humans, but it's something to look into.
41:22
Even in mice, I think it's a pretty big deal because this kind of stuff, I would assume, is conserved across the mammalian lineage, right?
41:30
Something is...
41:31
deep as making new synapses yeah i wouldn't be surprised if we found it now in primates but this this is you know something that we can study further on i think it's i think it's very cool it definitely maps on to the behavior we see in teenagers yeah that's for sure uh from a from a purely non-scientific perspective it tracks yeah
41:52
Story number four is from Penn, University of Pennsylvania and University of Michigan.
41:58
Penn and Michigan create the world's smallest programmable autonomous robots because we need more robots.
42:10
This was published in Science Robotics and Proceedings of the National Academy of Sciences.
42:16
Researchers at UPenn and University of Michigan have created the world's smallest fully programmable autonomous robots, microscopic swimming machines, little Michael Phelps is running around, that can independently sense and respond to their surroundings, independently, and operate for months and cost just a penny each.
42:35
Not going to take off in the U.S.
42:37
No way.
42:38
We don't got pennies anymore.
42:39
We don't even have.
42:41
Yeah.
42:41
Yeah.
42:41
You got to buy in bulk now.
42:44
But, okay.
42:45
All jokes aside, this is pretty crazy, okay?
42:47
This is very crazy.
42:48
You saw the photo, right?
42:49
Yes.
42:49
These things are smaller than a grain of salt.
42:52
They're powered by light, and they've got a little computer that can let it move in sophisticated ways.
42:57
Okay.
42:57
This thing is insane.
42:59
And they operate at the biological level.
43:01
So, you know, they say you can monitor individual cells.
43:04
You can have it go inside your body.
43:07
You can help with small-scale manufacturing.
43:09
Like, imagine a laboratory that's the size of, like, you know, microns by microns.
43:14
I think...
43:15
It's definitely look, it's it's a cool technological advance, but all those people who are like, there's a chip in my body like this is a weird.
43:26
It's a little bit weird.
43:27
Now there's some experimental research.
43:29
Yeah.
43:29
Now there's.
43:29
Yeah.
43:30
And they operate, you know, with just light.
43:34
So they've got little tiny solar panels and then they've got a computer that lets them figure out where they want to go and things like that.
43:41
One thing that I thought was really cool is how do you make something that small move around in water?
43:46
Because you run into the low Reynolds numbers, right?
43:50
The Navier-Stokes equations, when you go down to that scale, water is basically like honey.
43:56
And so it's really hard to move around.
43:57
So how are they actually doing it?
43:59
Well, if you have movable parts, like little flippers and stuff...
44:02
One, that's a really inefficient way to move around in honey because, like, as I move this way and then I retract my flippers, I'm actually going to go backwards.
44:10
The other thing is movable parts are really bad because they're not durable.
44:14
Yes.
44:14
They can break, especially if you're moving around in honey.
44:16
There's no nanorobotic autozone.
44:19
Yeah, yeah, yeah.
44:20
And these things are lasting for months on end, right?
44:22
The way they do it is they generate electric fields around their chip, and then the electric fields then move around ions—
44:31
Because our body has a bunch of ions.
44:33
And then that moves around water, which propels them forward.
44:37
So there's no moving parts.
44:39
I just am so mad at how clever that is.
44:41
It's pretty clever, I gotta say.
44:43
It's a bit weird.
44:45
That's what we're going for.
44:46
But at the end, it is pretty clever.
44:48
And I think for the right use cases, it can be huge.
44:52
For delivering, let's say, medical cargo to one specific site of the body versus another.
44:57
Things like that, it could be very useful.
44:59
we've been hearing about the potential of nanotechnology since I think I was like in middle school, how it's the next frontier and all these things.
45:08
This is a real practical, you know, it reminds me of the programmable protein story we covered from Yale in last season.
45:16
We're doing greater and greater things at smaller and smaller scales.
45:21
And again, the, the applications, you know,
45:25
Eat your heart out.
45:26
Yeah, it's it's it's endless.
45:28
That's a really, really good story.
45:30
Our last story for the rundown today is about Leonardo DiCaprio.
45:37
Genetic clues.
45:37
Da Vinci.
45:39
Oh, excuse me.
45:41
It's Golden Globes.
45:42
Excuse me.
45:43
I saw that clip of him.
45:44
Yeah.
45:47
Da Vinci, the one from the 1500s.
45:49
Yeah.
45:50
Genetic clues may be hiding in his artwork.
45:54
There were many institutions involved in this one.
45:57
The Venter Institute, Vanderbilt, Vanderbilt University Medical Center, and the University of Maryland.
46:03
This was out in bioarchive.
46:06
And the idea here is that
46:08
The artist, inventor, and all-around genius, Da Vinci, was sort of the definition of the Renaissance Man.
46:16
Obviously, we have the Da Vinci Code, all these books, all this stuff.
46:18
And scientists have been aiming to unlock the secrets of his genius on a genetic level.
46:22
But there's just one small problem more than 500 years after his death.
46:28
His DNA has proved virtually impossible to locate.
46:32
Yeah.
46:33
No kids.
46:33
His grave was destroyed during the French Revolution.
46:36
Yeah.
46:36
That's not the only thing that was destroyed in the French Revolution.
46:39
Let me tell you that.
46:40
But, you know, that's my jab at the French.
46:43
There's going to be another one later in the episode.
46:46
In any case...
46:47
I don't know why, but there's this thing called the Leonardo Project, and they're obsessed with all things Leonardo, and they really want his DNA.
46:55
They want to sequence his DNA.
46:58
What the project's team did was they swabbed letters that were written by a distant Leonardo relative.
47:04
And they also swabbed something called the Holy Child, which is a drawing that was possibly created by Leonardo.
47:10
And they found a Y chromosome, which is the chromosome that is in all males, and there's not a lot of mutations.
47:19
So if there's a distant Leonardo relative, and then this swab from the letter is actually matching, then it's very highly likely that it is Leonardo.
47:28
And this kind of reminded me of there's this...
47:31
documentary called the lost leonardo it was about this painting salvatore mundi that was apparently the lost leonardo da vinci painting he he only made a handful of paintings something like 10 or 15 less than 15 and each one of his paintings are masterpieces the first painting that i ever saw was madonna with child in the um gallery at london in london and it is it is like
47:56
Quite amazing, just the way that he shades paint together, and the faces that he makes are just, like, incredible.
48:06
This particular painting, the Salvatore Mundi, it's sold for $450 million.
48:11
It's the most expensive painting to have ever been sold by Christie's, and it is currently owned by Mohammed bin Salman.
48:18
of Saudi Arabia, he's trying to open a new art gallery in Saudi Arabia, and this is going to be the big masterpiece there.
48:27
It's debated whether Salvatore Mundi is from Leonardo.
48:31
Okay.
48:31
Okay?
48:32
But if we could figure out, like, a scientific way to say this is Leonardo or not, then, you know, you could actually see whether he made the bet was correct or not.
48:44
We are giving new meaning to the common phrase, the Da Vinci code.
48:49
Yeah.
48:50
Oh, yeah.
48:51
Nice one.
48:53
I do this professionally.
48:56
Great five stories on the rundown today.
49:01
Prehistoric woolly rhino, the rollout of an 11 million pound rocket in preparation for Artemis II launch.
49:11
We figured out why teens get to communicate better in their brain because of all these new synaptic connections.
49:17
nanobots are here yeah and we may finally have the dna of leonardo da vinci so that we can genetically replicate him and have a bunch of geniuses running around we are now going to move into our next story which is about something very fascinating that i've never heard of which is time vaults yeah
49:41
So in a new paper out in science from both Harvard and the Broad Institute of MIT is this new genetically encoded device for transcriptome storage, information storage, in mammalian cells.
49:58
Right.
50:00
This story is super interesting.
50:02
Yeah, this is pretty incredible because I think it is going to, dare I say, if this is what it purports it to be, this might be something that is Nobel Prize worthy.
50:12
Okay, wow.
50:13
Okay, I'm going to start there.
50:15
There's something called a measurement problem.
50:17
In biology.
50:17
In physics, we have the measurement problem, right?
50:19
Which is like you measure something and then that destroys the system.
50:22
Well, in biology, we've got something very similar, which is the destructive nature of observation.
50:27
If I want to measure what is happening inside of a cell, I basically have to kill the cell, get all of its contents out, and then do RNA sequencing to figure out what kind of RNA is inside and all that other kind of stuff.
50:39
So we can never observe the same cell twice.
50:43
And what we have to do is we have to infer biological dynamics using static snapshots of stuff, right?
50:49
Interesting.
50:50
You've got to break down, and then you've got to break down here, and then you've got to sort of connect the dots in time.
50:55
I actually did not know that.
50:57
That's interesting.
50:57
But if you think about it, that kind of makes sense, right?
50:59
How am I going to observe a cell?
51:00
Well, I've got to break all of its contents.
51:02
I've got to get all the proteins and then sequence the proteins.
51:05
Oh, there's this much amount of this protein.
51:06
There's this much amount of this.
51:08
This is the mRNA profile.
51:10
Right.
51:11
Which is the transcriptome.
51:12
So it's like super nonlinear.
51:14
And if you have stuff like cancer and embryonic development, which is nonlinear and non-ergotic, these systems, you can't really piece together a time sequence that confidently.
51:25
You can't really understand what's happening in time.
51:29
You can understand what's happening in snapshots of time.
51:32
Exactly.
51:32
So there's a lot of information loss, right?
51:34
There's cellular decisions that happen, let's say now.
51:37
Yes.
51:37
And then the result of that cellular decision is only going to come forward in a few days.
51:43
Yes.
51:43
Right?
51:44
So how do I connect those two things?
51:46
Right.
51:47
And the existing solutions are something like DNA editing.
51:50
In 2014, there was this CRISPR-based tape recorder.
51:54
Right.
51:54
OK, that what you could do is you could say, OK, I want to watch these specific genes and how they're transcribed.
52:00
And there was a tape recorder that you could literally like be like, oh, OK, this thing was transcribed at this time.
52:05
I'm going to like keep it and save it in some sort of memory.
52:08
But you had to choose beforehand what to record.
52:12
It wasn't OK.
52:13
Right.
52:14
Yeah.
52:14
And there's only a few events that you could record before the memory went out.
52:19
Okay, and that's where the breakthrough comes in here.
52:22
It's called Time Vault, okay?
52:23
Genetically Encoded Molecular Archives.
52:26
It's a time capsule for cells that stores the secret experiences of their past.
52:33
It's very, very cool.
52:34
And it's repurposing something called the Vault Particle to physically encapsulate mRNA and store it as the cell just goes about its daily business.
52:43
We created almost a cellular backup system like Time Machine on your Mac.
52:47
Yeah.
52:48
Dude, it's exactly that.
52:49
It's a Time Machine on your Mac, but now at the cellular level.
52:54
That's incredible.
52:54
It's insane.
52:55
No, that's incredible.
52:57
Especially given before we were only getting snapped.
52:59
Like the scale of difference of capability.
53:03
Yeah.
53:04
okay it's very cool okay so let's talk about like why this is important right yeah so in physics we like to think of systems as state vectors like all of the particles in a room are going to have position and momenta and if we know those position and momenta then we'll know everything about that system well in biology the state vector is really something called the transcriptome which is the mrna counts there's about 20 000 genes
53:30
that are getting expressed in the human body.
53:33
And if you could count which genes and how much of that gene is getting expressed in a cell, you could have something like a state vector for that cell.
53:42
It could be like, what is the cell state in?
53:44
The cells in our skin will have a bunch of genes that express for the proteins that are responsible for our skin.
53:52
The cells in our blood will have the blood genes expressed.
53:56
The DNA is the same, but...
53:58
But how much of which gene is expressed is what matters in terms of what that cell is doing.
54:03
It's what differentiates how cells have these different functions.
54:08
Exactly, yeah.
54:09
And so in nature, you can think of DNA as like a stable read-only memory.
54:14
So that's your ROM.
54:15
And then your mRNA is a volatile random access memory, right?
54:18
That's your RAM.
54:20
Now, mRNA has a half-life that's about five to nine hours, which means...
54:25
In about every five to nine hours, half of the amount of mRNA that was there is going to be gone.
54:30
It's going to be degraded.
54:31
And this is a feature and not a bug because the cell can then quickly respond to an environment and forget what it was doing five hours ago.
54:39
You want to be able to quickly change what you're doing in case there's some outside stress that comes along, right?
54:45
It's sort of this adaptive property.
54:47
Exactly.
54:48
And there were ways that we could tag the new mRNA to be like, okay, I want to check how much of this new mRNA is going through.
54:55
You could tag it with something called 4-thiouridine, which is a specific type of nucleotide that goes into the mRNA and it kind of saves stuff.
55:03
But that signal was limited by about 17 hours.
55:07
So it's like undetectable after 24 hours.
55:09
So your memory was only really 24 hours because you've got these things called exosomes in your cell that are basically the trash collectors.
55:18
And they collect all the trash and then they put it out.
55:21
Okay.
55:22
So, okay.
55:23
How do we extend the recording window?
55:25
Well, one naive way to do it is just be like delete all the exosomes, all the trash collectors.
55:29
Okay.
55:29
That's a really bad idea.
55:32
Imagine a city without trash collectors.
55:34
We don't want that.
55:34
We don't want that.
55:35
We don't want that.
55:35
Some people say that's what L.A. is today.
55:38
I will have you know we still have trash and recycling running.
55:41
Yeah, and it's not that bad.
55:42
It's not that bad.
55:43
But you can imagine that's a really bad way of doing things.
55:46
You can't stop degradation because that stuff is essential for life to keep moving around.
55:50
So instead...
55:52
What if we could spatially segregate the mRNA and keep it away from the trash collectors?
55:58
Okay.
55:58
Right?
55:58
We could put it in a vault.
56:00
Aha.
56:02
I see.
56:02
Although the name vault is actually not for, like, you know, a vault to store stuff.
56:07
It actually comes from this 1986 paper by Karadasha and Rome at UCLA in the Journal of Cell Biology.
56:14
This was 1986.
56:14
Okay?
56:16
What they were doing was studying coded vesicles, which are, like, just transport spheres that, like—
56:22
Package a bunch of stuff and then move it around the cell.
56:25
And they kept coming up with this contaminant.
56:28
Okay?
56:28
There was a contaminant in their sample all the time.
56:30
And when they looked really closely at this contaminant, there were these large barrel-shaped organelles.
56:36
Okay.
56:36
Okay?
56:37
And when you went into an electron microscope and took a look at them, these organelles resembled vaulted ceilings of Gothic cathedrals.
56:47
So they named them vaults.
56:49
And these things are massive.
56:52
If you look at the scale bar there, that's 100 nanometers.
56:55
That's like the size of some cells.
56:57
And these organelles are that big.
56:59
And they're the largest non-viral cells.
57:02
ribonucleic protein complex.
57:05
They've got a mass of approximately 13 megadaltons, which is three times the size of a ribosome.
57:10
The ribosome is the factory that makes proteins out of mRNA, and they're pretty big for just like single units of stuff, right?
57:21
Not really single units.
57:22
There's like subunits that form together to create a ribosome.
57:25
But at the end of the day, it's not a fully-fledged organelle with like a membrane and like a full like, you know,
57:31
Inside and outside.
57:33
This thing...
57:35
13 megadolsons at the dimensions of tens of nanometers.
57:39
And there's a bunch of them in our cells.
57:41
10,000 to 100,000 per cell.
57:43
Jesus.
57:44
Okay.
57:45
A lot of our immune system cells have an overexpression of this stuff.
57:49
Okay.
57:49
And what's weird is we still don't really know what they're there for.
57:54
And they're not essential because if you knock them out in mice, like you just get rid of the protein that makes these vaults, the mice are fine.
58:03
So there's some... You know, there's a variety of things.
58:06
It's either, you know, evolutionarily over time, we no longer need these things, but they're still around.
58:12
Yeah.
58:12
Or there's some functionality... That we just haven't figured out.
58:14
We don't really quite know yet.
58:15
And it's more likely the latter, because there's a lot of stuff that...
58:19
Our DNA is made up of introns and exons, so there's stuff that the DNA keeps inside the nucleus, and then there's stuff that gets expressed.
58:26
This stuff is getting expressed, right?
58:28
If you have 10,000 to 100,000 per cell, clearly evolution by now should have figured out if there's no use for this, why waste time and energy making this stuff?
58:39
But it's being expressed.
58:40
So there is some reason.
58:41
We just haven't quite...
58:43
Figured out what that is.
58:44
But the advantage is because it's not critical for function, you can use it as this kind of ideal, robust, non-toxic chassis for stuff.
58:54
Okay?
58:55
So let's get into this vault protein.
58:56
What does it look like?
58:58
There's a single protein that makes it up that's called a major vault protein.
59:03
You get 78 copies of this protein, and it self-assembles.
59:08
It doesn't need like something to come in and assemble it.
59:11
It's like a bunch of tiny little magnets.
59:13
78 of them come together and make this barrel shape.
59:16
Beautiful structure, right?
59:18
This very beautiful symmetric barrel-like structure.
59:22
They're all just coming together.
59:23
There's these two subunits that come together, one on each side, 39 on each of them.
59:30
And...
59:31
The Young's modulus of this thing, like the stress and strain, like how hard it is, it's about as hard as like some plastics or like viral capsids.
59:37
Oh, wow.
59:37
Which is very good because it's, you know, it's quite sturdy.
59:41
Yeah, yeah.
59:42
And the other thing is, unlike viral capsids, when you like poke them, the viral capsids are kind of...
59:48
They're going to go all over the place, and they're not really going to be able to recover.
59:52
With these major vault proteins, there's interactions with that protein, the way it sort of just magnetically locks into place with all of the others, that ensures that if I were to, like, poke it with, let's say, a little tip, like an AFM tip or something like that, with atomic force microscopy tip, the vault protein will break.
1:00:16
And then when I remove the stress, it'll just form back up.
1:00:20
Just naturally, biology has figured out this exact design to make this beautiful barrel-like structure that is resistant to stress.
1:00:29
It's incredible.
1:00:30
And we still don't know what it does.
1:00:31
What it does.
1:00:32
Like 40 years later.
1:00:34
1986, 40 years later.
1:00:35
We still quite don't know what it does.
1:00:37
But it is clearly fantastically engineered through evolutionary biology.
1:00:42
Yeah, it's absolutely amazing, right?
1:00:45
So how do we get from vaults to now time vaults?
1:00:49
This is a really funny story, okay?
1:00:50
So one of these students at the Harvard Lab found Leonard Rome's YouTube channel.
1:00:56
He's called The Vault Guy on YouTube, and he makes a bunch of YouTube videos about vault proteins and just molecular biology in general.
1:01:05
Shout out to The Vault Guy.
1:01:06
And yeah, I guess one of these grad students in the Harvard Lab found the YouTube channel and was like, yo, this vault guy is making some pretty cool stuff.
1:01:13
Maybe we could use him.
1:01:15
For our purposes, right?
1:01:17
And that's where they got the idea of time vaults, and then this paper has come out in Science.
1:01:22
A genetically encoded device for transcriptome storage in mammalian cells.
1:01:27
The idea is we're going to use these vaults as our Mac time machine.
1:01:32
Yes.
1:01:32
Right?
1:01:32
Yes.
1:01:33
Yes.
1:01:34
Very cool.
1:01:34
So how do we do that?
1:01:36
How exactly mechanistically does it work?
1:01:38
Because, you know, on this podcast, we like to focus on just the Lego blocks of biology.
1:01:44
And how you put together a bunch of Lego blocks that actually work the way that they're supposed to.
1:01:49
Because the concept of what they're trying to accomplish and the reason why makes total sense.
1:01:56
Yeah.
1:01:56
Yeah.
1:01:58
And then the entry point of, you know, these vaults being a possible conduit to an engineered kind of functionality is interesting.
1:02:10
But there's a big gap between those two things.
1:02:12
Yeah.
1:02:12
Like, how am I going to get the mRNA inside?
1:02:14
Yes.
1:02:15
And then how does it stay inside and all this other kind of stuff?
1:02:18
Right.
1:02:18
This is, again, at the molecular level.
1:02:20
Yes.
1:02:20
OK. Well, at the molecular level, everything is Lego blocks.
1:02:23
So let's talk about mRNA.
1:02:25
OK.
1:02:26
One of the properties of mRNA is when it comes out, it has something called a poly-A tail, which means at the end of the mRNA, after it gets transcribed from the DNA, there's a bunch of AAA, the adenine, there's just repetition of adenine, and that poly-A tail...
1:02:42
is kind of like a handle for figuring out where to take this mRNA and things like that.
1:02:48
You need to be able to grab it without messing with the genetic information that's later, right?
1:02:53
So you need some kind of handle for the sword.
1:02:56
It's like on a knife.
1:02:58
Yeah.
1:02:59
You don't want the end to be a blade that you're grabbing.
1:03:01
Yeah, because then you'll mess with the blade or you'll kill yourself.
1:03:05
So the poly-8 tail is there.
1:03:07
And the stuff that interacts with the poly-A tail is something called the poly-A binding protein, PABP.
1:03:14
And what these guys figured is we could have a container and a bait.
1:03:19
Okay.
1:03:20
The container is my vault protein shell.
1:03:23
And my bait would be this protein that interacts with the poly tail and also has a little domain on the protein that interacts with the vault.
1:03:33
Okay.
1:03:33
So the PABP, that binding protein, is going to grab the poly A tail.
1:03:38
And then part of it is going to get stuck to the vault.
1:03:42
Yes.
1:03:43
Okay.
1:03:43
And all of this we can put in a single piece of DNA.
1:03:45
Mm-hmm.
1:03:46
Such that when it gets transcribed, the vault protein gets transcribed to make the vault.
1:03:51
And also the poly-A binding protein gets transcribed.
1:03:56
You see?
1:03:56
That is so good.
1:03:57
And so when I turn this thing on, it makes the whole package.
1:04:01
It makes my vault.
1:04:02
And it makes the thing that grabs the mRNA and puts it inside the vault.
1:04:05
We've created a vertically integrated stack at this molecular level to effectively go through the creation of the individual component parts necessary.
1:04:17
We don't have to hijack something existing.
1:04:20
We're literally able to say, no, we want the time vault.
1:04:22
We want the grab protein, the transporter.
1:04:24
Yeah.
1:04:25
And so it's fully self-contained as a process.
1:04:28
Yeah.
1:04:28
Fully self-contained.
1:04:29
Fascinating.
1:04:30
Amazing, right?
1:04:31
And then what you can do is this entire genetic circuit, you can turn on and off based on this tet on tet off mechanism, which is effectively what you do is you introduce an antibiotic called doxycycline.
1:04:42
And then when you introduce it on, this genetic pathway gets triggered.
1:04:45
So it makes the vault protein and it makes the interaction protein.
1:04:49
Right.
1:04:50
And then and then there's another called docs off where I can introduce docs off and then it's going to stop stop the vault protein.
1:04:58
And so now I have a controlled time window where I can make this time machine.
1:05:05
I can trigger it.
1:05:06
It can turn on.
1:05:06
It can make all this stuff and it'll just grab a bunch of mRNA, stick it inside the vault protein and then I can make it stop.
1:05:14
For many listeners may have heard of Doxycycline before because it's like a it's an antibiotic you'll get for like a variety of use cases.
1:05:22
So I'm like, it's simply just Doxy triggers the initialization of the storage.
1:05:27
That's again, it's you're not inventing like we're taking things that are already there.
1:05:31
Yeah.
1:05:31
But just putting these Lego blocks together in a really clever way.
1:05:35
Right.
1:05:35
Yeah.
1:05:36
And then now you can see if this time vault stuff works, right?
1:05:39
Yes.
1:05:40
So you can show that there's no new capture.
1:05:42
This time vault, the half-life of the mRNA in the time vault is now 132 hours, which is five days instead of the five hours that we had before.
1:05:52
So the mRNA is sticking around because it's been captured.
1:05:56
And if I now lyse the cells, it's still in there.
1:05:59
Mm-hmm.
1:06:00
Mm hmm.
1:06:01
Right.
1:06:01
Yes, this is good.
1:06:03
So you can now you can now basically tell a cell.
1:06:06
Yes.
1:06:07
I want you to save all your memory right now.
1:06:09
Yes.
1:06:10
Yes.
1:06:10
Simply by introducing.
1:06:11
Yeah.
1:06:12
Yeah.
1:06:12
It'll save all the memory.
1:06:14
And then you can just have the cell go on its merry way because all of the stuff is inside the vault protein.
1:06:18
Right.
1:06:19
It's not interacting with the stuff outside.
1:06:20
And we know that the vault protein is not toxic.
1:06:23
Yes.
1:06:24
To the cells themselves because it was already there in the first place.
1:06:26
So it can just hang around doing the function we want it to because it's non-toxic and it's self-contained, meaning like the mRNA inside cannot escape.
1:06:35
And also if it gets poked, it'll reconstruct.
1:06:37
So it sort of has security in that context.
1:06:40
So the point is you want to be careful when you start doing things like this to not introduce...
1:06:46
you know, new variables that the system will attack or are not prepared for.
1:06:49
But the structure we've already walked through really identifies that this is not going to interrupt regularly scheduled programming.
1:06:57
Exactly.
1:06:57
While still enabling the capability of capturing the memory of a cell over this long, much longer period of time.
1:07:03
Exactly.
1:07:04
And so now you can ask, the mRNA that's inside the vault protein, right?
1:07:08
How long does that last?
1:07:10
Well, the thing is lasting for about 17 days.
1:07:12
Right.
1:07:14
Okay?
1:07:15
The stuff that's inside the vault protein.
1:07:16
And you ask, okay, why is it lasting that long?
1:07:18
Well, that's actually the inherent physics of RNA.
1:07:22
Okay.
1:07:22
Okay?
1:07:23
RNA is ribonucleic acid versus DNA is deoxyribonucleic acid.
1:07:28
One of the big differences is, of course, the DNA is double-stranded.
1:07:32
And RNA is single-stranded.
1:07:34
So DNA is a bit more stable, obviously, because you've got, you know, a fully-fledged ladder that's sort of protecting the inner bases.
1:07:41
The other big thing is that the ribose sugar, which RNA has, and DNA has deoxyribose sugar, the ribose sugar has an extra OH.
1:07:50
Okay, it's got an extra hydroxyl group on the two of that pentagon.
1:07:54
Yeah, okay.
1:07:54
Right?
1:07:55
Yep.
1:07:55
And that...
1:07:56
Extra OH, that oxygen, oxygen is always just bad.
1:08:00
Okay, it just attacks random crap.
1:08:02
And so the RNA, that extra oxygen goes and attacks the phosphate backbone.
1:08:07
And that process has a characteristic timescale of 17 days.
1:08:12
So we're reaching the theoretical limit of how long RNA can just hang around before the chemical RNA itself starts attacking itself.
1:08:23
We're literally at the, what did we call the Eddington limit?
1:08:27
Yeah, but in this case it would be the thermodynamic limit of RNA stability.
1:08:33
But what it's showing is that there's no enzymatic degradation of mRNA.
1:08:38
Yes.
1:08:38
There's no other stuff.
1:08:40
There's no exosomes that are coming in trying to, like, take it out.
1:08:43
This is just pure physics.
1:08:45
It's going to last about 17 to 18 days.
1:08:47
It's locked down but for the fact that the structure of ribose itself will degrade.
1:08:53
Will degrade the iron.
1:08:55
Which we – that's a bigger – that's a whole other thing.
1:08:57
Yeah, that's like – I don't know what to do with that.
1:08:59
Yeah, yeah, yeah.
1:09:00
Yeah.
1:09:01
So here you can show that, like, you know, this –
1:09:05
This is really happening, right?
1:09:07
So the critical step is once you have the RNA that's present in the cytosol, let's say, versus RNA that's present in the time vault, how do you tell the two apart?
1:09:18
Right, okay.
1:09:19
Because if I were to now take the cell and I were to say, okay, I want you to record now, and then in five days I want to read it, right?
1:09:27
How do I...
1:09:29
tell what's already there versus what's in the vault.
1:09:31
Well, I can introduce something called RNAs, which is an enzyme that eats RNA.
1:09:37
If I make it eat RNA, the stuff that's inside the time vault is protected from this enzymatic degradation.
1:09:44
Oh my God, bro.
1:09:46
This is so good.
1:09:47
It's so good.
1:09:48
This is so good.
1:09:49
And so now if you add RNAs, the only stuff that we're going to be able to read afterwards is the stuff that's inside the vault protein.
1:09:58
It's so good.
1:09:59
This is, this is, this is... You're watching masters at work here.
1:10:04
Yeah, this is great.
1:10:06
I can see why you started off with such a high bar.
1:10:09
High bar, right?
1:10:11
There are some caveats.
1:10:13
For example, it's not going to take in mitochondrial RNA because the mitochondrial RNA is inside the mitochondria, right?
1:10:20
And it's not going to take in...
1:10:23
any RNAs that lack a poly-A tail.
1:10:25
There's some.
1:10:25
There's always some that are exceptions to the rule.
1:10:28
And so, you know, the interaction protein is not going to be able to grab that mRNA because there's no poly-A tail to actually grab.
1:10:36
The capture efficiency is also quite low.
1:10:38
It only takes in about 3% of the total mRNA, right?
1:10:41
Because there's so much mRNA.
1:10:43
So it's only going to take in a small sample.
1:10:45
And so that low capture rate means that you really need to pool together
1:10:53
statistical bias right where if you were to do just one cell how do you know that you got all of it if it's only taking three percent but if you do a thousand cells then you're pretty sure that you've got all of the mrna covered that's well that's a solvable we'll solve that yeah and and there's there's ideas later that'll that'll help you solve it but i just want to let you know that in at least in the first iteration of this this is still pretty no this is this is right now that this base is there like i i i'm just shocked at how clever that is
1:11:20
Yeah, it's really nice, okay?
1:11:22
And so how do we know that it works?
1:11:25
Well, they did a really simple experiment.
1:11:27
What they did was heat shock these cells after... And then right after the heat shock, they had...
1:11:37
the RNA go into the vault proteins.
1:11:39
Okay.
1:11:40
And then they compared the vault proteins right after the heat shock to when you, and then, you know, you wait five days and you sequence the mRNA that's in the vault proteins.
1:11:49
And then you compare that to the heat shock.
1:11:51
And then right after you, you lysate and then you do it right.
1:11:54
If this whole thing worked, then the mRNA that's in the vault proteins right after the heat shock should match normal cells right after the heat shock.
1:12:03
The transcriptones should match.
1:12:04
Yes.
1:12:04
And that's exactly what they found.
1:12:06
right is that the present transcriptome if you look at five days later the present transcriptome forgot that it was heat shock but the stuff that's inside the vault remembers that it got heat shot because you did the time machine thing you did the backup and you did the backup right then right very beautiful and the figures i gotta say the aesthetics of the figures are also really nice a plus the color coordination of the blue and the red i liked it i don't know look look at
1:12:30
We like design and aesthetic, both experimental and visual.
1:12:34
Yeah.
1:12:35
Yeah.
1:12:36
So I thought this was really cool.
1:12:37
So now applications, right?
1:12:39
They actually went further and went into applications.
1:12:41
This is the one where people talk about, you know, real world applications.
1:12:44
So Harvard, when it came out with the story in their press, they talked about how did that cancer cell become drug resistant?
1:12:51
One of the first things you can do is apply this thing to cancer cells.
1:12:54
Okay.
1:12:56
Whenever you do cancer cells and you treat the cancer cells with a drug, there's always 1% of cancer cells that are drug-tolerant persisters.
1:13:04
You know how you have the soap that kills 99% of germs?
1:13:08
Well, there's always 1% of germs that get away, right?
1:13:11
And the same thing happens with cancer.
1:13:13
If you...
1:13:15
Target therapy with something called osimertinib, which is a drug for cancer.
1:13:21
1% of these cancer cells survive.
1:13:23
Now, the debate is the following.
1:13:25
Were these cancer cells pre-adapted or was the fact that I introduced the drug causing them to adapt?
1:13:32
This is a debate that goes all the way back to Charles Darwin and Jean-Baptiste Lamarck.
1:13:39
There was the idea of how do organisms gain traits?
1:13:42
The Darwinian model was the one at the top that you see, which is, let's take a population of giraffes, right?
1:13:48
Why do giraffes have long necks?
1:13:51
Well, you could have a population that has a bunch of different neck sizes, and then the longer necks could reach the trees and eat the trees.
1:14:02
And then the subsequent populations got longer and longer necks.
1:14:07
That's actually the one at the bottom.
1:14:09
Okay.
1:14:09
Now, the Lamarckian hypothesis is the hereditary nature of acquired characteristics, meaning the giraffes kept stretching their necks, and then that stretching of the necks made the subsequent population stretch their necks.
1:14:28
And so now that's why we have longer neck giraffes.
1:14:31
Would it be almost like an epigenetic kind of theory?
1:14:34
Yeah, yeah, yeah.
1:14:34
Like more of an epigenetic theory.
1:14:36
Exactly, yeah.
1:14:37
And for what we know from large-scale populations, it's really a Darwinian thing.
1:14:41
But the Lamarck hypothesis is not completely off-ground, right?
1:14:44
At the cellular level, we have examples of that.
1:14:47
For example, with bacteria, when we heat shock bacteria, the bacteria just start taking in DNA from their environment, right?
1:14:54
So that's kind of Lamarckian because I've been introduced to stress, and then that stress is causing the genetics to change.
1:15:00
Yes.
1:15:00
So the question is, what's happening with these cancer cells, right?
1:15:03
And currently, we can only record the bulk population, which is mostly dead, and then we can record the survivors, which is after the fact.
1:15:11
So to answer this question in cancer biology, you have to be able to record what's happening before the drug was introduced.
1:15:18
Yes.
1:15:18
And then...
1:15:20
Probe what the survivors were doing.
1:15:23
Yeah.
1:15:23
Yes.
1:15:23
Does that make sense?
1:15:24
Yes.
1:15:24
And this is perfect for the time vaults.
1:15:26
Yes.
1:15:26
So what they did was they had a protocol where you record for 24 hours with no drug.
1:15:31
And then you introduce this cancer drug, Osimertinib.
1:15:35
Yes.
1:15:36
You record this cancer drug for four days, 99% die.
1:15:40
Yes.
1:15:40
And then the ones that survive, let's read out what's inside their vaults.
1:15:45
Yes.
1:15:45
Yes.
1:15:46
And it was definitive evidence for the selection model, meaning that these cancer cells had previously adapted.
1:15:51
Yes.
1:15:52
The cancer cells have a very high mutation rate.
1:15:54
And so they're exploring all of these different possibilities.
1:15:57
And the 1% of cancer cells that survived had a distinct transcriptional signature before the drug treatment.
1:16:04
And so now the focus for therapeutics can shift from adaptation to targeting that preexisting state.
1:16:10
And what they did was they looked at some of the genes that were expressed in those vaults, specifically FN1.
1:16:16
They knocked that out.
1:16:18
And then now none of the cancer cells survived that drug.
1:16:20
So it's like literally working.
1:16:22
This is actually unreal.
1:16:25
I want people to understand how unreal this is.
1:16:27
Yeah.
1:16:27
Because now you're... Because we can now...
1:16:31
DVR ourselves.
1:16:35
We can Mac Time Machine backup ourselves.
1:16:38
It means we can actually see what happens before the
1:16:43
Like we can understand after you introduce some external factor, like what the changes are.
1:16:50
And particularly in the cancer research example you just brought up, this is so fascinating.
1:16:54
You have all of them.
1:16:55
You do the recording in the time vault.
1:16:57
You're doing it over 24 hours.
1:16:59
You have all of them there.
1:17:00
Then you introduce the current treatment.
1:17:02
Yeah, 99% die.
1:17:03
99 are gone.
1:17:04
1% are left over.
1:17:05
You can look at the 1% that are left over and say, what is it about these guys that's different?
1:17:10
And then now we can attack the thing like the FN1.
1:17:13
That's different because we are able to now actually see over time do it again.
1:17:18
And now they all die.
1:17:19
Yeah.
1:17:19
That's incredible.
1:17:21
Yeah.
1:17:21
Yeah.
1:17:22
That's incredible.
1:17:23
Do you understand what we can do with that?
1:17:24
Yeah, that's so cool.
1:17:26
You can use this for like, and the fact that these vault proteins are just ubiquitous in mammalian cells.
1:17:31
Yes.
1:17:32
Means that you can just apply this all over the place.
1:17:34
Exactly.
1:17:35
This almost feels, conceptually, as I try to put it into my brain, it feels CRISPR-esque as being a platform versus a single point solution.
1:17:45
Exactly, yeah.
1:17:47
This is a really big deal.
1:17:48
It's a really big deal.
1:17:50
Once it percolates and people poke, hold, reviewer two, poke the hole.
1:17:54
Yeah.
1:17:55
No, but people are very excited.
1:17:56
Excited about this.
1:17:57
From the press coverage that I've seen, people are very excited.
1:17:59
This is, I mean, holy moly.
1:18:02
Yeah.
1:18:02
It's very cool.
1:18:03
And what a great name.
1:18:04
Time Vaults?
1:18:04
Yeah.
1:18:05
Great name.
1:18:05
Well done.
1:18:06
Great name.
1:18:07
Well done on that one.
1:18:07
Whoever was doing the branding, right?
1:18:10
Yeah.
1:18:11
The branding on that one was great.
1:18:12
So.
1:18:12
What's the future?
1:18:13
Well, we want to overcome this 3% capture efficiency.
1:18:16
Sure.
1:18:16
Right?
1:18:16
So one thing that I thought was really cool, the idea was you can put in inside the vault protein, you can put in something called reverse transcriptase.
1:18:24
And what that's going to do is take the mRNA that's getting stuck inside the vault protein and reverse transcribe that into DNA.
1:18:31
Now the DNA is going to last much longer than the 17 days that's the thermodynamic limit for RNA.
1:18:37
And so you could like have weeks of time vaults, right?
1:18:43
So it's almost like there's a short-term memory in the mRNA storage that gets – and then sort of like how we record.
1:18:49
We record onto a small hard drive, and then we dump it onto our archive drive for long-term storage.
1:18:53
Yeah.
1:18:54
It's exactly that.
1:18:55
Yeah.
1:18:55
The other thing could be like you could have molecular identifiers per vault per cell.
1:19:01
And then it's kind of like a barcode for each cell.
1:19:03
And then you could be like these mRNA came from this cell and these mRNA came from this cell.
1:19:08
So you get the aggregate, but you also get single cell resolution.
1:19:11
Mm-hmm.
1:19:12
Kind of cool, right?
1:19:13
To think about.
1:19:15
You could combine this time vaults with super resolution and expansion microscopy.
1:19:19
So then you could locate where the time vaults were in a cell.
1:19:22
Right now, they're just sort of lysed and in a test tube.
1:19:25
But imagine, you know, you'd stick it under a microscope.
1:19:27
You could locate where in the cell it is.
1:19:29
Is it next to the nucleus?
1:19:30
Is it next to the endoplasmic reticulum?
1:19:33
So on and so forth.
1:19:33
Is it next to the cell membrane?
1:19:36
Synthetic organelles, you could have programmable delivery vehicles that will just go through.
1:19:42
I think it's very, very cool.
1:19:44
My mind is just racing.
1:19:46
Yeah, there's so much that you can do with this.
1:19:48
This was a go.
1:19:50
Again, we don't always like to give credit to two of our arch rivals in Harvard and MIT.
1:19:57
But kudos, guys.
1:19:59
This was a good one.
1:20:00
Still not Princeton, but we'll give credit where credit's due.
1:20:05
This one's a good one.
1:20:07
I thought that was really cool.
1:20:08
That was fantastic.
1:20:09
Time volts storing my brain.
1:20:13
Every week, I know for those of you who watch and listen, my brain is constantly being, my mind is constantly being blown.
1:20:19
But the implications of some stuff we talk about is just insane.
1:20:22
We're going to go to our last main story of the day.
1:20:25
Yeah.
1:20:26
which is going to be about ancient mathematics, ancient maths, as they say, across the pond, which right now we're having some sibling rivalry with.
1:20:39
This story is in the Journal of World Prehistory.
1:20:44
The dawn of art and mathematics, ancient Mesopotamian pottery, reveals early mathematical thinking.
1:20:53
Yes.
1:20:53
So we've been thinking about math, the maths, for quite a while.
1:20:56
For quite a while.
1:20:57
And it used to be that we thought math from the fourth millennium BC, right?
1:21:01
The Sumerian city-states, they had these, like, tablets that had stuff like the Pythagorean theorem.
1:21:06
They even had the quadratic formula, like, figured out.
1:21:10
Which...
1:21:11
I think it's very, very cool, right?
1:21:13
And they had a base 60 system, which we still use today when we talk about, like, minutes to an hour, seconds to a minute, angles, things like that.
1:21:22
We still use base 60.
1:21:24
This is a new discovery that says that Halifian pottery, which is 6,000 BC, it's in northern Mesopotamia.
1:21:33
This might contain geometric sequences.
1:21:37
and the first evidence of humans doing math.
1:21:39
This is 1,000 to 2,000 years before the Sumerian city-states.
1:21:42
Okay?
1:21:43
And it's a pretty cool study, and we're going to get into it.
1:21:47
I have some...
1:21:49
I don't know much about anthropology, and so maybe I'm...
1:21:52
Maybe I'm not getting it, but I think it's still a really cool study.
1:21:56
And we'll get into some of the qualms that I have later on, but it's not much.
1:21:59
Okay.
1:22:00
So this time period is 6th millennium BC, late Neolithic revolution, right?
1:22:05
The Halifians were this culture that were widely dispersed in a network in northern Iraq and Syria.
1:22:12
So this is above the Euphrates and the Tigris rivers.
1:22:15
Mm-hmm.
1:22:16
Sixth century, sixth millennium BC, way, way back.
1:22:20
Okay.
1:22:20
So this is right when we're coming out of the post ice age stabilization, the climate is becoming sort of stabilized and the world population is about 40 million.
1:22:29
It's quadrupling.
1:22:30
It's like growing really fast.
1:22:32
Humans are, you know, getting better.
1:22:34
They used to live in these distinctive circular toloid dwellings, mud brick dwellings, agricultural villages.
1:22:42
They used to have dry farming, so no irrigation as of yet.
1:22:45
They were just relying on the rains to make wheat and barley.
1:22:50
They used to have herding for sheeps and goats.
1:22:53
And...
1:22:54
Before the Halafians, most of the art was dominated by animals.
1:23:00
Okay?
1:23:01
This is an example of cave paintings from 40,000 years ago from Laskaux, France.
1:23:08
I believe that's how they pronounce it over there.
1:23:11
That's how it's spelled anyways.
1:23:12
So I don't know any better.
1:23:14
But 40,000 years ago, they were making a bunch of cave paintings that were a bunch of really beautiful animals and things like that.
1:23:20
Right?
1:23:21
Halafian pottery comes in.
1:23:23
6000 BC.
1:23:25
Very beautiful, first of all.
1:23:26
The pottery itself is like really, really nice.
1:23:29
Okay.
1:23:30
And this pottery...
1:23:32
contains a shift between just focusing on animals to now also focusing on the botanical world.
1:23:39
Okay.
1:23:39
Okay.
1:23:40
So if you go to the next slide, there's examples of plants on pottery, plant art.
1:23:47
And this is a really fundamental cognitive shift from this hunter-gatherer, animal-centered view to now farmers and plant-centered.
1:23:56
The vegetarians, man.
1:23:57
Yeah.
1:23:58
I mean, because farming becomes the way of life, right?
1:24:01
To create civilization.
1:24:02
100%.
1:24:03
To create population.
1:24:04
So...
1:24:06
The idea with this paper is maybe this change is what's creating now mathematics and the need for mathematics.
1:24:13
Because you're shifting into this system that naturally benefits from some foundational mathematics.
1:24:21
Mathematics, right?
1:24:22
Because now you have communal stuff that you need to split up.
1:24:26
Yeah, yeah.
1:24:26
Okay?
1:24:27
So let's get into the history of human mathematical thought, right?
1:24:31
Paleolithic, there was this thing called the Ishano bone.
1:24:34
which was discovered in the Congo in 1960, it's thought that the first example of mathematics, now it's in a Belgian museum because, of course, the Belgians... Boo!
1:24:45
Yeah, the Belgians did great things for the Congo, right?
1:24:48
And for those who are listening, my face is very sarcastic.
1:24:53
Okay, the Belgians were probably the worst colonizers of all the colonizers.
1:24:58
So...
1:24:58
But they got the bone and now it's in one of their museums.
1:25:01
Great.
1:25:02
So why do we think this is a mathematical tool?
1:25:05
Well, it's got all these like notches on it that happen to be prime numbers, which is very cool.
1:25:11
Yeah.
1:25:12
Yeah.
1:25:12
Right.
1:25:12
Yeah.
1:25:13
Like 20,000 years ago, they're putting in these notches that are prime numbers.
1:25:16
It could be like a counting tool.
1:25:17
This is very advanced.
1:25:19
Yeah.
1:25:19
Right.
1:25:20
So for 20 anyway, 20,000 years ago.
1:25:23
Right.
1:25:23
Right.
1:25:24
There's some criticism that maybe it's just a tally stick or like the notches are for better grip.
1:25:29
But the prime numbers are kind of specific.
1:25:32
Specific.
1:25:33
Right.
1:25:34
Yeah.
1:25:34
So I like that one.
1:25:37
OK.
1:25:37
I'm just going to go ahead and say I like that one.
1:25:40
Then there's an understanding about Neanderthal yarn.
1:25:43
This came out, and what they were doing was microscopic analysis of yarn that they found in Neanderthal cave, and it showed evidence of an intertwined three-ply cord.
1:25:54
Okay.
1:25:55
And if it's intertwined like that, you know, where it's like you're doing braids, then there's some understanding of mathematics, right?
1:26:01
You got to do three and then one under and then two and then one under, so on and so forth.
1:26:05
Okay.
1:26:06
So maybe the Neanderthals had...
1:26:08
basic mathematical concepts, right?
1:26:10
Then we get into about 8,000 BC.
1:26:13
There's something called the token system, which is, again, in Mesopotamia.
1:26:16
These early farmers used geometric clay tokens like triangles, cones, spheres, for one-to-one accounting.
1:26:24
And there's evidence that three cones would mean three measures of barley and so on and so forth, right?
1:26:29
And amidst this is where we get the Halafian Gap, right?
1:26:33
which is where this paper sort of fits in.
1:26:35
Okay.
1:26:36
It's a paper that was out very recently, and it's showing that the Halafian period, which is about 6,000 to 5,000 BC, it predates the Sumerian writing, and perhaps...
1:26:48
The pottery is evidence that it's not just aesthetics that they were going for, but they were going for some kind of mathematical codification for their agricultural society.
1:26:58
Okay.
1:26:58
Okay?
1:26:59
And this is—we're getting into something called cognitive archaeology, which is reconstructing ancient thought.
1:27:05
It's always very hard to do.
1:27:06
Yes.
1:27:06
Yes.
1:27:07
ethnomathematics which is inferring math from indirect indications like we did with that bone yes from the congo so it's very hard to do all right so this is not an exact science but i still think it's incredibly cool because i love thinking about how we got here how do we get out of the cradle of civilization yeah yeah and how do we form this kind of amazing society right yes to think about something that happened 8 000 years ago is is i think very very cool yes so
1:27:34
Let's get into the paper.
1:27:35
They did a survey of pottery from 29 different agricultural sites across northern Mesopotamia.
1:27:41
Archaeological sites.
1:27:42
Yes.
1:27:43
Yes.
1:27:43
What did I say?
1:27:44
Agriculture.
1:27:45
Oh, yeah.
1:27:45
Yeah.
1:27:45
Archaeological sites.
1:27:46
I guess they were agricultural, but you're right.
1:27:49
This is archaeological sites where they analyze tens of thousands of painted shards of pottery.
1:27:56
So they have one thing that's very important is they have a very large statistical sample.
1:28:00
This is quite rare.
1:28:01
I was going to say that's I was literally going to say I can't imagine there's a lot of archaeological sites where we have that volume.
1:28:06
Yes.
1:28:07
Of stuff.
1:28:07
Yes.
1:28:08
So this is a lot of data from tens of thousands of shards from all over this geography.
1:28:14
And what we find, what they find is vegetal motifs, so motifs of plants that are a shared cultural language.
1:28:21
The entire Halafian territory had this language.
1:28:26
If you look at pottery from one end all the way to the other—
1:28:29
You've got plants.
1:28:30
So that means this isn't like a one-off thing.
1:28:32
Right.
1:28:33
Okay?
1:28:33
This is a kind of cultural language that goes across the entire civilization.
1:28:39
Right?
1:28:39
And there's an average frequency of these motifs, which is pretty significant.
1:28:43
About 4% to 6% of all of the pottery has this.
1:28:46
And given the amount of data that we have, we can actually be very confident that this is there.
1:28:51
Right?
1:28:51
Mm-hmm.
1:28:52
Now, the motifs are classified into four basic categories.
1:28:55
And there's a critical subgroup.
1:28:57
There's a bunch of large flowers, okay?
1:29:00
And these flowers are on big bowls and the bases of these bowls, and they happen to be in radial patterns.
1:29:07
Yes.
1:29:07
Here you can see there's four, then you have eight, so you get this doubling, right?
1:29:11
Yes.
1:29:12
You get four, which is a cross, and then you half each of the crosses, so you get a doubling and you get eight.
1:29:17
Yes.
1:29:17
And then these radial encodings don't just stop at eight.
1:29:21
They keep going on and on.
1:29:22
So then you get 16, you get 32.
1:29:25
And the one that's really crazy for me is all the way to 64.
1:29:29
Yeah.
1:29:29
Okay.
1:29:31
Getting 32 pedals.
1:29:32
I can imagine you do four.
1:29:34
Then you, you know, you do the pizza cut for eight and then you do another pizza cut for 16.
1:29:38
Then you do another pizza cut for 32.
1:29:40
But if you've ever cut pizzas for like a large group of people and you've only got one pizza, which happened to me once, it was the worst party ever.
1:29:47
You can't get to 64 because then at that point it's like you're doing like a lot of
1:29:51
It's like atoms of bread, you know?
1:29:53
And so what they did with the 16, with the 64, is they divided it into four quadrants.
1:30:00
And then they had 16 in each, a 4x4 sort of, a 4x4 grid in each of the four quadrants to give 16 times 4, which is 64.
1:30:10
Yes.
1:30:10
This is pretty convincing that there is some mathematics.
1:30:14
Yes.
1:30:14
Yes.
1:30:14
That's happening, right?
1:30:15
Yeah.
1:30:16
That level, the complexity, once you start getting to that size.
1:30:18
I mean, also, these designs are beautiful.
1:30:20
These designs are beautiful.
1:30:21
For 8,000 years ago.
1:30:22
Yeah.
1:30:22
They're not just like.
1:30:23
Like, the patterns are amazing.
1:30:26
Yeah.
1:30:26
The checkerboard.
1:30:27
Yes.
1:30:27
The radially outward.
1:30:29
Yes.
1:30:29
Like, even in terms of symmetry.
1:30:32
Yes.
1:30:32
8,000 years ago, this is incredible symmetry.
1:30:35
I totally agree.
1:30:37
I mean, I can barely get that out of chat GPT today.
1:30:40
Yeah.
1:30:43
So the authors argue that there's some kind of consistent use of these two to the power of two, two to the power of three.
1:30:50
So you get four, eight, 16, 32, and 64.
1:30:54
And these are not –
1:30:58
By coincidence.
1:30:59
There's some kind of having logic, right?
1:31:01
Where the sequence emerges because of some intuitive geometric operation.
1:31:05
We are dividing a circle into a bunch of parts, right?
1:31:09
And now comes the theoretical side of things.
1:31:12
These Halifian villages used to have communal sharing economy.
1:31:17
And then the crops, like grain or barley, had to be shared out fairly to everyone in the population.
1:31:23
So they had the sharing economy.
1:31:25
What do we call it now?
1:31:26
With the...
1:31:27
the uber and the instacart oh yeah what are those called it's not sharing a con it's is it uber pool no the the creator economy anyway yeah they used to have that they used to have that yeah all the way back then right and there's there's more um there's more evidence that there used to be complex geometric subdivisions because if you look at the stamp seals from the halafian period the stamp seals also have this very nice geometric pattern and these were seals that were used to mark property and secure trade items so
1:31:56
So it supports this idea that there's some sophisticated economic administration.
1:32:00
Yes.
1:32:00
And perhaps this kind of economic administration would require basic mathematics of two to the power of whatever.
1:32:07
Yes.
1:32:08
Right.
1:32:08
Okay.
1:32:09
Now, there are examples of other numbers.
1:32:14
Okay.
1:32:15
Okay.
1:32:15
Like, there are examples of six or seven numbers.
1:32:19
6, 7.
1:32:20
Or like 12, 13.
1:32:22
And what the authors argue is that these show less meticulous craftsmanship.
1:32:27
This is where I'm like, I don't know.
1:32:29
That still looks pretty good to me.
1:32:32
Right?
1:32:32
6, 7, 12, 13.
1:32:35
These still look very beautiful.
1:32:36
So...
1:32:37
I don't know.
1:32:38
I'm not an anthropologist.
1:32:40
Maybe there's some way to quantify that these are less beautiful than the ones that were 16 and 32.
1:32:48
There's also some pushback in the CNN article.
1:32:50
There was a professor, Jens Harup, from Roskedeik University in Denmark.
1:32:56
He specializes in Mesopotamian mathematics, and he was not convinced.
1:33:01
He said that the symmetry...
1:33:03
is just an isolated incident of mathematical technique rather than the evidence of a broader mathematical reasoning.
1:33:09
If you just got to divide the circle nicely, of course you're going to get powers of two.
1:33:14
Now, he's an expert in Mesopotamian mathematics, which tells me that he's probably of the Sumerian city-states camp, and he wants the Sumerian city-states to be first.
1:33:23
So, you know, obviously with science there's a lot of, like, human tension.
1:33:28
Yes.
1:33:29
And it's worth something to think about, like, you know,
1:33:33
Is this just symmetry?
1:33:36
Or is consistently dividing a circle into 32 equal parts and then 64, it's...
1:33:42
inherently performing a mathematical operation without written symbols right yeah it's an open question i i think it's the symmetry is clear to me yes and to me we've had a lot of discussions about symmetry on this podcast and so symmetry is inherently a mathematical principle yes so to me i i think this is pretty cool right if you have not listened to it check out our chen ying yang episode i think it's episode 18 or 19
1:34:07
From season one, yeah.
1:34:08
From season one, where we go deep into the concept of symmetry.
1:34:13
And that was a lot of fan favorite episodes.
1:34:18
What's interesting here is I do, I think, get the tension of, like, there's, it looks aesthetically good versus understanding a broader mathematical context of why it looks so good.
1:34:28
Yeah.
1:34:29
That feels fair to me.
1:34:31
Yeah, yeah.
1:34:32
Like, I can sort of see, like, yeah, like, you know, I make things that look good all the time.
1:34:35
I don't know the math behind it.
1:34:36
Yeah.
1:34:37
Yeah.
1:34:38
But given how early it is, even being able to compute aesthetic value.
1:34:44
Yeah.
1:34:44
And have it all over the civilization.
1:34:47
Right.
1:34:47
And clearly it was a part of some aspect of culture, whether it's economics or otherwise.
1:34:53
Yeah.
1:34:53
That it's not just...
1:34:56
for it's clearly a shared visual language yes right yeah yes yes and so i this this is interesting so what what is your what is your conclusion here my conclusion is that i do think that they were doing something mathematical because of the 64 okay that's all yeah the 64 bowl with the four quadrants and then that was the four by four grid in each of them that was like yeah
1:35:19
That was pretty like, oh, okay, that's like, you know, they're doing something.
1:35:22
Yeah.
1:35:23
No.
1:35:24
But again, it could just be symmetry.
1:35:25
I like to think that there's mathematics because I just like to find the beauty in things.
1:35:30
But, you know, who knows?
1:35:34
For those who have stayed this long...
1:35:38
And worked through mathematics, which you've always said we're never going to cover mathematics.
1:35:42
And you broke your rule because it's so good.
1:35:45
And it was kind of light mathematics.
1:35:47
Light, you know, mathematics light.
1:35:48
What do you guys think?
1:35:50
Do you think this Halafi work on the pottery and stuff, do you think it's math?
1:35:56
Do you think it's art?
1:35:57
Do you think it's...
1:35:58
Art math, math art.
1:36:00
Let us know in the comments.
1:36:02
As always, we like to know how folks get to the end of the episode.
1:36:06
And it just helps us with that little comment.
1:36:09
We are going to...
1:36:12
wrap up here today with a follow-up from the cloud nine story.
1:36:18
Yeah.
1:36:19
Potential for dark matter gas cloud from the last episode.
1:36:24
I'll let you take it away here.
1:36:25
Yeah.
1:36:26
I mean, we, at the end of the last episode, we asked our audience, what do you guys think?
1:36:32
caused the name Cloud9.
1:36:34
And, you know, some of the authors of the episode that we sent the link to, they actually made it that far as well, which is great.
1:36:41
Yes, yes.
1:36:42
And they were asking...
1:36:45
They actually let us know.
1:36:46
Yes.
1:36:46
What is the story behind the name Cloud9?
1:36:49
So...
1:36:50
This particular response comes from Alejandro Lambe, but actually most of the authors of that study responded with very similar takes.
1:36:57
I'll just sort of paraphrase how he put it.
1:37:00
So back in February 2023, he read a paper about a potential dark galaxy candidate that was identified by the FAST telescope, which is the big Arecibo-like telescope in China.
1:37:10
And the authors were initially unaware of Alejandro's earlier paper where he had described this idea of relics.
1:37:18
Yes.
1:37:18
He was one of the original authors of that relics paper being like, hey, there should be something like this out there.
1:37:23
Yes.
1:37:24
So he emails them.
1:37:25
The gas cloud that he emailed them about turned out to actually be a galaxy.
1:37:29
So when you point at optical telescope, you see stars and it's just a galaxy.
1:37:33
Then the same fast guys.
1:37:34
Yes.
1:37:35
Yes.
1:37:36
in china they look at this object cloud nine and cloud nine was named cloud nine because it was the ninth object that they had seen in that direction okay so it was just the cloud nine and in china there's no sort of cultural meaning for the phrase cloud nine the fast guys then emailed alejandro back being like hey this could be a new candidate and this was the ninth cloud in their radio survey
1:38:00
The authors were aware of Alejandro, and then they sort of teamed together.
1:38:04
They went, looked at the analysis from the VLA.
1:38:07
The VLA actually could pinpoint where Cloud 9 was because the fast telescope is this giant thing that's just sort of like stuck to the Earth, right?
1:38:16
So it doesn't have very good angular resolution on where this thing is coming from.
1:38:20
VLA could then pinpoint and be like, it's actually right there.
1:38:23
Once you have those coordinates, then you can go to the Hubble guys at NASA and be like, hey, look—
1:38:28
I actually know where it is, and I'd really like to look at it.
1:38:32
The Hubble guys gave the authors some time, and that's where we get that paper from Cloud9.
1:38:37
They kept the name because, you know, if it turns out that Cloud9 is a relic, then the authors would be on Cloud9.
1:38:44
So they kept that name for that reason.
1:38:46
But that's what's behind it.
1:38:49
Which is very cool.
1:38:50
We always love naming conventions on this show.
1:38:53
Yeah.
1:38:54
And it was great to get outreach from everyone on the team.
1:38:57
Yeah, that was awesome.
1:38:58
If we happen to cover, if you're listening now and you happen to be in a lab on a research team, having had a story by us covered at some point and there is a correction, you can always make sure you reach out to us to facilitate that correction in a future episode.
1:39:16
We do our best to cover it as accurately as possible.
1:39:19
But there's no one better than the experts.
1:39:21
And so we are really grateful that the team for the relics paper were just so enthusiastic in providing feedback and loving the coverage.
1:39:31
This was a fantastic episode 23.
1:39:35
Yeah.
1:39:36
The Time Vault story is still so nuts to me.
1:39:39
wrapping notes for housekeeping.
1:39:41
Again, if you want us to cover a paper, you can go to FFPPod.com backslash submissions.
1:39:48
We have a Reddit style upvote, downvote ranking board on there that can help you in the community, provide us stories you want us to cover in the future.
1:39:58
As we mentioned earlier in the pod, we do have an Artemis special coming up next week.
1:40:06
So that will be a one story deep dive.
1:40:09
we will probably still do a rundown as well.
1:40:12
But just if you're listening this far, you're a fan, just be aware of that so you're not disappointed when you tune in next week.
1:40:19
We're going to get into a bunch of weeds.
1:40:21
I am your host, Lester Nare, joined as always by my co-host and our resident PhD, Krishna Chaudhary.
1:40:30
This is From First Principles.