Latest / From First Principles / Ant Scans, Lunar Chickpeas, Hidden Galaxies & Superconductivity (EP 40)

Transcript

Ant Scans, Lunar Chickpeas, Hidden Galaxies & Superconductivity (EP 40) · From First Principles

• 0:00
  They identified 1,300 objects that have an odd appearance, right?
• 0:04
  And more than 800 of these objects have never been identified in scientific literature before.
• 0:10
  The secret to farming on the moon is worm poop and fungus.
• 0:14
  Yes.
• 0:14
  Not quite exactly what NASA put in the brochure.
• 0:17
  However, it could be the future of a permanent mission.
• 0:20
  lunar presence.
• 0:21
  Now that could transform our society more than I think any other technology that I can really think of.
• 0:28
  Hello internet this is your captain speaking Lester Nare joined as always by my co-host and our resident PhD Krishna Chaudhary.
• 0:38
  This is a rundown episode.
• 0:40
  We'll be covering four stories at a high level.
• 0:44
  If you're interested in our deep dives, check out the previous episodes from this week.
• 0:48
  We may even, in fact, have another round of Are You Smarter Than a Scientist?
• 0:55
  So in this week's rundown, first of the four stories, 3D scanning of ants, thousands of ants with a particle accelerator published in Nature Methods,
• 1:06
  Fascinating story.
• 1:07
  We're following that up with growing chickpeas on the moon with fungus and worm poop.
• 1:14
  The good stuff.
• 1:15
  Scientific Reports brings us that one.
• 1:17
  We will follow that up with some discoveries from the Hubble Archive.
• 1:21
  And that Hubble Archive discovery of cosmic hidden anomalies.
• 1:26
  but we used AI to discover them.
• 1:28
  That was an astronomy and astrophysics.
• 1:31
  And we will end with an LK99 throwback about the state of room temperature superconductivity, where we may have a roadmap that was published in PNAS.
• 1:40
  We are going to learn about the science from the ground up today because this is From First Principles.
• 1:46
  From First Principles.
• 2:02
  So for our first story, we are going to be talking about ants, like the movie, the Pixar movie.
• 2:09
  3D scanning of thousands of ants with a particle accelerator, a team led by researchers at the Okinawa Institute of Science and Technology, used a psychotron particle accelerator to create detailed 3D micro CT scans.
• 2:25
  I've gotten a couple CT scans, but not a micro one, of a 2100 CT scan.
• 2:30
  Ant farm colony with nearly 800 species.
• 2:33
  And this was published in Nature Methods because the method is what is interesting here.
• 2:38
  Yes.
• 2:39
  So for the longest time, if we want to find, if we want to make like a 3D version of these
• 2:45
  really tiny insects, and we want to do it to super, super high resolution, you use something called a micro CT scanner.
• 2:52
  This thing is going to send out x-rays, and then those x-rays are going to scan and allow the scientists to examine the physical structure, the stuff inside without actually destroying the specimen.
• 3:03
  Each of these scans takes about 10 hours for a single specimen.
• 3:09
  So it's not very high throughput, right?
• 3:11
  And we want to be able to do a lot in parallel.
• 3:14
  And so that's what these researchers did.
• 3:16
  They employed the Karlsruhe Institute of Technology, which has a giant synchrotron particle accelerator, X-ray imaging.
• 3:27
  It's also got robotics.
• 3:28
  It's also got...
• 3:29
  artificial intelligence and what they're able to do is generate interactive digital reconstructions of 800 ant species it's pretty incredible if we if we look at the the headline right and if we look at this animation look at the detail that we're getting this and this is this is crazy yeah this is that micron resolution of an ant right and
• 3:55
  This is rendered using AI because a lot of times we get little slices.
• 4:00
  When we do a CT scan, we get like little slices or like different sort of configurations of what the ant is in, depending on the specimen.
• 4:08
  But you can use AI now to reconstruct that 3D version of the ant.
• 4:13
  It's really quite amazing.
• 4:14
  With that setup, they've scanned 2,000 specimens in a single week, 800 different ant species.
• 4:22
  The effort is called AntScan.
• 4:24
  They've got a free public website called AntScan.info, and it's kind of like a Google Maps for ant anatomy.
• 4:31
  The data is also synchronized with genome sequencing projects.
• 4:35
  That's really cool.
• 4:35
  So you've got 585 of those scans linking to the genomic data of about 186 different species.
• 4:44
  I think it's a really cool application of particle physics and, like, particle accelerators that we don't really think about.
• 4:51
  It reminds me of a very old story that we did on this podcast about the T-Rex bone.
• 4:56
  Yeah.
• 4:56
  Remember that?
• 4:57
  Yeah.
• 4:58
  Where it was like they found...
• 5:00
  blood clotting in a T-Rex bone and showed that the T-Rex was capable of rapidly healing its bones after some kind of, I think it was a fight with another T-Rex probably.
• 5:14
  But it's a really cool application of particle accelerators that doesn't have anything to do with fundamental physics.
• 5:21
  It's just leveraging it as a tool to do very high-resolution 3D imaging of biological specimens.
• 5:28
  Cool.
• 5:29
  Which, if you asked me before, we did the original T-Rex story, which my reaction was like, this is unbelievable.
• 5:37
  Because we were sort of repurposing this very expensive, large-scale device to become multipurpose beyond its sort of initial mission or context itself.
• 5:49
  Yeah, and the scans are like so detailed that they can be used for further training of machine learning systems to recognize ants in the field.
• 5:58
  For example, you know, I'm out in the field, I get some specimen.
• 6:01
  Now I have a pre-trained network, a machine learning model that is trained using this really high fidelity, really high resolution data.
• 6:09
  And it can immediately tell me what kind of species it is, or at least what species it is most related to if it's a new species, right?
• 6:19
  Like if Disney or somebody wants to make, what is it called?
• 6:25
  The visual effects.
• 6:26
  Yeah.
• 6:26
  Like would like a giant ant, you know, remember that, that one movie where like, it's like, honey, I shrunk the kids.
• 6:32
  Yes.
• 6:32
  You want to make like a, you know, a giant visual effect for that.
• 6:35
  Now you can be really, really biologically accurate.
• 6:38
  Yeah.
• 6:39
  So if you want to be as precise as Interstellar was in their replication of the physics of a black hole.
• 6:47
  Yeah.
• 6:47
  And the amount of science that went into creating so.
• 6:51
  Now particle accelerators can give us animated 3D ants that are giant at very high fidelity.
• 6:58
  Yeah.
• 6:59
  It kind of is a little bit the example you just brought up sounds a little bit like a Pokedex.
• 7:04
  You can tell if it's a Caterpie.
• 7:05
  Yeah.
• 7:06
  Or a Beedrill or a Weedle.
• 7:08
  Yeah.
• 7:08
  Or any of these.
• 7:09
  Butterfree.
• 7:10
  I'm trying to flex my Pokemon knowledge.
• 7:12
  I don't know any.
• 7:13
  I just know Pikachu.
• 7:15
  And Charizard.
• 7:16
  Charizard.
• 7:17
  Classic.
• 7:18
  Classic.
• 7:19
  Always interesting to see the intersection of different disciplines.
• 7:22
  Yeah.
• 7:22
  Yeah.
• 7:23
  And how, again, past work can inform new work.
• 7:26
  And so this was, again, a method story.
• 7:28
  Yeah.
• 7:29
  Fascinating and published in Nature Methods.
• 7:32
  We have this coming out of the Okinawa Institute of Science and Technology.
• 7:35
  And I'll let you... You had the good pronunciation on the second institute.
• 7:39
  Oh, yeah.
• 7:39
  That's the Karlsruhe Institute of Technology.
• 7:43
  And Okinawa Institute of Technology, of Science and Technology, OIST, is what we used to call it.
• 7:48
  I actually did a summer...
• 7:50
  workshop there um for like computational neuroscience that's so it's an amazing like place that's it's on it's on a hill in okinawa in the middle of the pacific ocean beautiful beautiful spot a lot of bugs i have to say a lot of bugs a lot of bugs but you know it's a tropical island and you know what does not have bugs this podcast
• 8:12
  And if you've been listening or watching, or if you're a new listener or a new watcher, we're grateful for you joining us today to learn more about the universe around us.
• 8:20
  A like, a comment, a share, a follow on any of the platforms helps us reach more curious people and helps us continue to do this podcast and deliver the best science podcast on the planet, in our opinion, every week, day in and day out.
• 8:35
  And we're going to now transition into our second story,
• 8:39
  which is about growing chickpeas on the moon.
• 8:43
  Now, we are from California, but this story comes out of Texas.
• 8:48
  Not only just out of Texas, but from researchers at UT at Austin and Texas A&M, published in Scientific Reports, where they demonstrated that they can successfully grow and produce seeds in soil
• 9:03
  that is very similar to the soil that we know exists on the moon.
• 9:09
  And if we ever want to expand the human experience beyond our local Gaia on planet Earth, being able to do agriculture in space is valuable.
• 9:20
  And it seems like we may have an interesting finding in this recent paper.
• 9:25
  Yes.
• 9:27
  Trying to establish a permanent presence on the moon will require probably, you know, doing some form of agriculture there.
• 9:34
  Light is pretty easy to come by, but soil is pretty hard to come by because moon regolith, which is the name for lunar soil, is very, very different from...
• 9:44
  from earth soil earth soil has an entire ecosystem inside right there's bacteria there's fungi there's all sorts of creatures living in there and so soil is not just like a bunch of like you know small particles of rock which effectively is what the in you know the a biological part of soil is you need a lot of these
• 10:07
  helper agricultural necessities to actually create an environment to grow crops in the Martian with Matt Damon, right?
• 10:17
  He grew potatoes, I think, but he supplemented it with his own poop, right?
• 10:22
  This is trying to get that dream.
• 10:26
  a reality without the poop or with without the poop without this case it's not human poop they're actually going for um worm poop ah i see okay so what they're saying is we can take lunar regolith which is this super you know kind of dry dead substance we can supplement that with both fungi which are going to help the plants
• 10:52
  And also poop from worms where the worms eat like sort of biological matter and trash and other kinds of stuff that comes out of the human presence there.
• 11:06
  Right.
• 11:06
  Because we're going to have a lot of like biological waste.
• 11:09
  The worms can subsist on that, create organic material.
• 11:12
  We supplement that organic material into dead lunar regolith.
• 11:18
  And then we grow chickpeas.
• 11:20
  And it actually worked.
• 11:22
  That's incredible.
• 11:22
  Yeah, and one thing that they actually did test was without the fungus, the plants did not produce any seeds.
• 11:29
  So it's really important that we have the fungus and we have this biodegradable waste that comes from the worms.
• 11:37
  So part of the idea here is we know we can't grow in a completely dead substrate.
• 11:44
  we do need some biological ingredients mixed in to make the recipe work.
• 11:52
  Yeah, and they simulated the lunar regolith because obviously you can't get the little bit of lunar soil that NASA brought back
• 12:00
  from its space missions.
• 12:01
  And now other people have also brought back with their non-human space missions there.
• 12:06
  But, I mean, lunar soil is very hard to come by, okay?
• 12:09
  I don't think you're going to convince anyone, hey, I want to grow chickpeas in your lunar soil, right?
• 12:14
  That's going to completely contaminate lunar soil.
• 12:16
  So they simulated lunar soil,
• 12:19
  And then they made sure that that sort of composition closely reflected the stuff that actually came back.
• 12:26
  And then they did this experiment with creating better soil with worm compost and fungi.
• 12:33
  The secret to farming on the moon is worm poop and fungus.
• 12:37
  Not quite exactly what NASA put in the brochure.
• 12:40
  However, it could be the future of a permanent lunar presence in better understanding how we can sustain worm.
• 12:48
  particularly on the ability to have a food supply when we might not be able to send starship rockets or whatever the defense prime rocket of the day is that can carry humans to the lunar surface.
• 13:04
  Yeah, and one thing I do want to mention is that we don't quite know whether these grown chickpeas are safe to eat.
• 13:10
  We just know that they've been grown and they look like chickpeas, right, in every sense of the word.
• 13:15
  So there's still more testing that needs to be done to make sure that these plants don't absorb, like, harmful metals that are in the soil on the lunar surface.
• 13:26
  And they actually provide the nutrients that the astronauts would need.
• 13:29
  So there's still more to the story, but it's a great first step.
• 13:34
  we will one day traverse the stars.
• 13:37
  And one of the key aspects to us being able to do so is being able to survive.
• 13:42
  And food is a key aspect of that.
• 13:45
  An early first step.
• 13:46
  Are you going to be the first to try to test the lunar chickpeas?
• 13:48
  I won't be the first.
• 13:50
  I'll tell you that.
• 13:51
  You sound like Elon.
• 13:52
  He's not going to be the first to Mars either.
• 13:54
  Now, I'm very excited to transition to one of my favorite parts of the show.
• 14:10
  Ladies and gentlemen, welcome back to another episode of Are You Smarter Than a Scientist?
• 14:15
  And the quizzical look on Krishna's face is how he's going to attack this question on this week's rendition.
• 14:22
  Name the 10 deadliest questions.
• 14:25
  animals to humans as you know are you smarter than a scientist we have a question we have 10 answers we have three strikes you at home you might have some ideas in your head our resident phd's has some ideas in his head and we are going to see how many of the 10 deadliest animals to humans interesting that krishna can name in this round the floor is yours good sir
• 14:50
  Man, I don't think I'm going to do well on this, but let's give it a try.
• 14:54
  You know, the first thing I can think of is mosquitoes because of malaria.
• 15:01
  It is a very good guess.
• 15:04
  Okay.
• 15:05
  And mosquitoes is number one.
• 15:07
  Okay.
• 15:07
  750,000 per year.
• 15:08
  Wow.
• 15:08
  That's a lot.
• 15:10
  It's not a small amount.
• 15:10
  And it's incredible that such a small thing can kill so many of us.
• 15:15
  Yeah.
• 15:15
  Okay.
• 15:17
  Let's do...
• 15:18
  Okay.
• 15:18
  In that vein, let's do like insects that carry disease.
• 15:21
  So like what about like fleas?
• 15:24
  Fleas.
• 15:25
  Or ticks.
• 15:26
  Can you give me that?
• 15:27
  Fleas or ticks.
• 15:28
  Yeah.
• 15:30
  It impacts dogs a lot, but it is not.
• 15:33
  It's not for us.
• 15:33
  Okay.
• 15:34
  It is not on our list.
• 15:35
  Damn, I'm already on a strike.
• 15:36
  All right.
• 15:37
  This was a tough one.
• 15:38
  This is outside of your purview.
• 15:40
  Definitely.
• 15:43
  I hear about bears a lot.
• 15:44
  No, I'm not going to say bears.
• 15:46
  Nobody interacts with bears.
• 15:49
  Sharks are probably also overblown.
• 15:51
  Yeah.
• 15:56
  I will note that some of them are classic.
• 15:59
  Okay, they are.
• 15:59
  Let's go with bears.
• 16:02
  Well, that's not what I expect you to say, unfortunately.
• 16:05
  Oh, no.
• 16:05
  It's not even.
• 16:06
  Oh, my God.
• 16:07
  I'm going to be so bad at this, guys.
• 16:09
  We're on strike two.
• 16:10
  I will note that Christian is a biophysicist.
• 16:12
  So this is.
• 16:14
  Yeah, this is zoology and like society.
• 16:17
  Oh, my gosh.
• 16:18
  I'm doing so bad.
• 16:18
  I'm already on two strikes.
• 16:21
  Some of them are classic.
• 16:22
  Yeah.
• 16:24
  Maybe I should not reference that because I don't want to throw you off.
• 16:29
  Okay, okay.
• 16:29
  Can you give me a hint because I'm only on one left?
• 16:32
  So many of us may be familiar with a gentleman who, rest in peace, was really impactful for us when we used to watch things about animals.
• 16:39
  A stingray.
• 16:40
  Oh.
• 16:41
  Or a manta ray.
• 16:44
  Not what I was expecting you to answer.
• 16:46
  Oh, I think you said Steve Irwin.
• 16:48
  I was talking about Steve Irwin.
• 16:50
  Was it not a stingray?
• 16:51
  But it was not a stingray.
• 16:52
  Oh, it's a crocodile.
• 16:53
  Yeah.
• 16:54
  All right.
• 16:54
  Okay.
• 16:55
  Now let me just see how many I can get.
• 16:56
  Come on.
• 16:57
  I got one, bro.
• 16:59
  So crocodile is number nine on the list.
• 17:02
  We will have one lifeline.
• 17:03
  Again, we're very new to this.
• 17:05
  All right.
• 17:05
  How about I'll give you one extra strike, which is against the rules, but we will have one mulligan in this exception.
• 17:11
  So we have number one mosquitoes, number nine crocodiles.
• 17:16
  This is a tough one.
• 17:19
  Hippo?
• 17:22
  Do we have hippo on the board?
• 17:25
  Number 10, hippo.
• 17:26
  We have one, nine, and ten.
• 17:29
  We have a lifeline.
• 17:31
  Damn.
• 17:33
  Dude, I'm like...
• 17:35
  I guess I'm so sequestered and like...
• 17:38
  You know, normal society.
• 17:39
  I have no idea.
• 17:40
  Sharks.
• 17:42
  Let's go with sharks.
• 17:43
  Sharks.
• 17:44
  Is that on the list?
• 17:45
  I don't have a four strikes, but I do have the sound effect.
• 17:49
  And that is not correct.
• 17:52
  That is not correct.
• 17:53
  Unfortunately, we have mosquitoes at one, crocodiles at nine, and hippos at ten.
• 17:59
  If you're at home, make your list now for what you think is.
• 18:04
  Two through eight is I purposely threw you a curveball.
• 18:07
  I gave you two easy ones.
• 18:08
  Yeah.
• 18:09
  The first two episodes.
• 18:10
  I had to curve.
• 18:11
  Number two.
• 18:13
  You could argue this was not going to be easy to get, but the answer is humans.
• 18:17
  Oh.
• 18:18
  Are the deadliest animals to humans.
• 18:20
  That's fair.
• 18:21
  We know that very well here in the United States.
• 18:24
  Number three.
• 18:26
  With snakes.
• 18:27
  Oh, why didn't I guess?
• 18:28
  Yeah, duh.
• 18:29
  Okay.
• 18:29
  Very poisonous animals.
• 18:30
  If you're a black mamba, R.I.P.
• 18:32
  Kobe or others.
• 18:34
  Number four, it might not have been your corgi.
• 18:38
  Really?
• 18:38
  But some dogs are our fourth.
• 18:41
  My fear of dogs as a child is now...
• 18:44
  Reignited.
• 18:45
  Is cats on here?
• 18:47
  So, unfortunately, cats are busy taking over the world intellectually.
• 18:51
  So they're not killing people.
• 18:51
  They're not killing people.
• 18:52
  They're smarter than that.
• 18:54
  Number five.
• 18:55
  There was no way you were going to get this one.
• 18:58
  Free water snails, which was news to me.
• 19:01
  Fresh water snails.
• 19:02
  Wow.
• 19:03
  Okay.
• 19:03
  So this is our number five.
• 19:04
  Number six.
• 19:06
  In our order, assassin bugs.
• 19:09
  I've never heard of that.
• 19:10
  Assassin bugs.
• 19:12
  I got to look this up later.
• 19:14
  So we may do now an episode on assassin bugs.
• 19:16
  Yeah.
• 19:16
  Number seven, the tsetse fly.
• 19:20
  Oh, I've heard of that.
• 19:21
  In the continent of Africa.
• 19:22
  And it's interesting.
• 19:23
  Some of my dad's research has been around this area.
• 19:25
  Yeah.
• 19:25
  The disease is a big one.
• 19:27
  Another funny reference to a family story, which we'll talk about in a future episode, about how someone ended up in the hospital for 21 days is scorpions.
• 19:37
  So our top 10 deadliest animals to humans, mosquitoes, humans, unironically, snakes, dogs, freshwater snails at five.
• 19:47
  I think I said something else earlier by accident.
• 19:49
  Six assassin bugs, CC flies, scorpions, crocodiles, and hippos.
• 19:57
  This was a bad one, guys.
• 19:59
  This looked like Ferrari at F1 two weeks ago.
• 20:02
  Not a good start to the season.
• 20:04
  No, this looked like Aston Martin.
• 20:07
  What's the other one?
• 20:07
  Haas?
• 20:08
  Is Haas doing anything?
• 20:08
  Haas is doing okay.
• 20:09
  They've got a Ferrari engine.
• 20:11
  They're doing okay?
• 20:12
  Yeah, they're doing okay.
• 20:13
  Okay, so it is simply the Ferrari.
• 20:16
  This is something, again, we're trying to see if you all enjoy.
• 20:20
  Let us know in the comments how many you got.
• 20:23
  Are you smarter than scientists?
• 20:25
  Probably on this one.
• 20:26
  And we'll bring it back for next week's rundown.
• 20:29
  If you want it more, you got to let us know you want it more.
• 20:32
  But in the meantime, before our next episode game show, we are going to go to our third story, which is about...
• 20:42
  Hidden Cosmic Anomalies in Hubble's Archive Discovered by AI.
• 20:49
  Researchers at the European Space Agency, ESA, published in Astronomy and Astrophysics, a new AI system called Anomaly Match, very on the nose, that scanned approximately 99.6 million images, cutouts from the entire legacy Hubble Archive site.
• 21:08
  In two to three days, which is an unbelievably short period of time.
• 21:15
  And what did we find?
• 21:16
  Yeah.
• 21:17
  So for background, machine learning and AI systems are very good at something called anomaly detection, right?
• 21:23
  Which is something that is out of the ordinary.
• 21:26
  Just because of the way that they learn data and they learn the data space, there's ways that they can group data.
• 21:33
  similar items together in their latent space and they can figure out outliers.
• 21:38
  It's kind of a, I mean, if you think about like rudimentary clustering algorithms where, you know, you project your data into some kind of subspace and then in this case, let's say, oh, normal spiral galaxies are over here.
• 21:51
  Normal barred galaxies are over here.
• 21:53
  There's going to be little tiny sectors where you find extremely rare objects and machine learning algorithms can do this very, very efficiently.
• 22:01
  Now, what these scientists did from the ESA was apply that method to the Hubble Legacy Archive, which is over the past 35 years, the Hubble Space Telescope has just been taking image after image, right?
• 22:14
  And one of the great things about the Hubble Space Telescope is all of their images are now public access, okay?
• 22:23
  If you're, you know, taking something today, then perhaps it's not public access.
• 22:27
  You've got like six months or something to like grind out as many papers as you can.
• 22:32
  But pretty soon it's going to go into the public archive.
• 22:35
  So all of these images were cut out into 100 million different little image cutouts that are measuring just a few dozen pixels on a side.
• 22:47
  Right.
• 22:48
  And from those hundreds of millions, from those hundred million image cutouts, they identified 1,300 objects that have an odd appearance, right?
• 22:57
  And more than 800 of these objects have never been identified in scientific literature before.
• 23:03
  Most of these anomalies are just like galaxies going through mergers or interactions, and so they're going to look weird.
• 23:08
  Yes.
• 23:09
  But every once in a while, there was some pretty...
• 23:12
  Cool stuff.
• 23:13
  So if we go to our next photo, you'll actually see some of the very cool stuff.
• 23:17
  So on the upper right-hand side, you're seeing galaxy mergers.
• 23:21
  But the lower two on the right, you can see one galaxy that's kind of curving around another.
• 23:28
  That's a weird gravitational lens.
• 23:29
  Yeah.
• 23:30
  Okay?
• 23:30
  Where because of Einstein's relativity...
• 23:34
  The gravity from the foreground galaxy is bending the light from a galaxy that's behind it.
• 23:41
  And so these are very, very cool gravitational lenses of a background galaxy that's sort of twisting its light around the foreground, which is in the front.
• 23:52
  We've also got really weird ring-shaped galaxies, like the one on the upper right.
• 23:56
  And these are all, like, completely new.
• 23:59
  I think that's so cool.
• 24:00
  And just look at the ring-shaped galaxy on the upper left.
• 24:03
  Oh, sorry.
• 24:03
  Yeah, the upper left.
• 24:04
  Yes, you're right.
• 24:05
  Got it, got it, got it.
• 24:05
  Yeah.
• 24:06
  Yeah.
• 24:06
  And I just think that's so cool that, like, a machine learning algorithm can go through these within two and a half days and figure out all these new galaxies that now we can
• 24:18
  Maybe get our other telescopes to look at, right?
• 24:21
  Take a look.
• 24:22
  Get some spectra.
• 24:23
  Figure out if maybe there's a supernova that's happening in one of these lensed galaxies that we can spot at multiple different times because maybe the light is taking longer on one end than the other end.
• 24:34
  If it's multiply lensed, you can do all sorts of very, very cool science by looking at these new objects, right?
• 24:42
  It's, again, one of these...
• 24:45
  Examples of a news story about AI helping out scientists.
• 24:49
  Yes.
• 24:49
  That, again, makes me very hopeful in all of the drab that we're getting about how AI is going to end the world.
• 24:56
  I mean, I think we are very much in danger of that as well.
• 24:59
  Don't get me wrong.
• 25:00
  But there's also hope.
• 25:01
  And I think if it's in the right hands, tools like artificial intelligence and the anomaly detection that comes with it.
• 25:08
  can be very fruitful for the pursuit of fundamental science.
• 25:12
  That is always the question is who controls the button.
• 25:15
  But the anomaly match, I'm sure, I mean, part of the Verisirubin already has this built into its existing architecture.
• 25:22
  And that's going to generate like unprecedented volumes of data.
• 25:25
  So an anomaly detector is going to be very crucial for...
• 25:29
  analyzing that data just to begin with and look at the fact that we're still discovering new stuff today yeah for so for this archive that's you know over you know had that has been gathering for over 30 years yeah and the Hubble only looked at a small patch of the sky right in in all of its 35 years of history it didn't like tile the entire night sky like the very Rubin will so it's very very exciting time and you know what that means there's a chance I will see the
• 25:58
  Yeah, one day one day we haven't looked this is the analogy I always bring up is we're looking out our front window at a house and we're saying there's no one in the street.
• 26:09
  But we haven't looked out the backyard.
• 26:10
  We haven't hopped over the fence.
• 26:11
  We haven't driven around the neighborhood, but we're just looking at the front windows like oh, we're all alone here.
• 26:17
  And it might be some roadblocks, right?
• 26:20
  Maybe they're doing construction on the Interstellar Highway.
• 26:22
  I kid.
• 26:23
  Maybe.
• 26:24
  Maybe.
• 26:25
  Our last story, one of my favorite almost things that happened that did not happen.
• 26:30
  Yes.
• 26:31
  Is about room temperature superconductivity.
• 26:36
  And this is out of the proceedings of the National Academies of Science from a combination of universities, including MIT, Harvard, Columbia, University of Houston, Carnegie Institution.
• 26:46
  Carnegie Institution, Graz University of Technology and Intellectual Ventures.
• 26:52
  And this was a programmatic paper that lays out a roadmap for achieving one of the pinnacles of, I guess, would you say engineering and science, the intersection of the two, which is the idea of this creating a room temperature superconductor.
• 27:09
  That's right.
• 27:09
  This is not a...
• 27:12
  This is not a proper science article that we usually cover.
• 27:16
  This is, in fact, a perspective article.
• 27:18
  So a lot of times scientists will get together and they will write effectively an opinion piece about where the field should be looking at and how the field should be structured
• 27:28
  for future advancement.
• 27:30
  And this is one of those, okay?
• 27:31
  It's in the proceedings of the National Academy of Sciences.
• 27:34
  It's a strategy paper that assesses the current state of research for room temperature superconductors and then sets out future directions.
• 27:42
  So superconductors are these materials that...
• 27:45
  have zero electrical resistance.
• 27:49
  Not negligible, not next to zero, but because of fundamental quantum mechanics, it is literally zero.
• 27:56
  The resistance of these electrons moving through this material is literally zero because of some very fundamental, very cool quantum mechanics that is going on.
• 28:05
  Now, that could transform our society more than I think any other technology that I can really think of.
• 28:12
  A very...
• 28:14
  Base example would be if we could have room temperature superconductors or high temperature superconductors, we can transfer electricity and power with almost no dissipation.
• 28:25
  Right.
• 28:26
  The problem with modern day superconductors is they either require extremely low temperatures.
• 28:32
  So even colder than liquid nitrogen sometimes.
• 28:35
  Or.
• 28:37
  extremely high pressures.
• 28:39
  If you want to get to high temperature, like room temperature, you got to like stick it inside a diamond anvil cell where like you're squishing stuff inside of a diamond and then finally you like make something that is superconducting, right?
• 28:53
  If we want industrial scale applications, we need something to be superconducting like on the table.
• 28:58
  Right.
• 28:59
  You know?
• 29:00
  And so there's a prediction challenge which comes from our ability to predict new superconductors.
• 29:06
  Now, that has advanced dramatically because we figured out a lot of the material science and the fundamentals of how to, you know, simulate these things in a computer.
• 29:16
  Now, what this paper is proposing is a shifting of focus towards like thermodynamics and synthesis modeling.
• 29:23
  Okay.
• 29:23
  Because a lot of times when we try to predict new superconductors in our computer programs, those can't be synthesized properly.
• 29:32
  Ah.
• 29:32
  Through the normal processes.
• 29:34
  It's like you've given me like the recipe, but I have no way of cooking this thing.
• 29:38
  You've shown me that unobtainium is a thing.
• 29:40
  Yeah.
• 29:41
  I have the root ingredients to create unobtainium.
• 29:43
  It's like, what are we doing?
• 29:44
  Right.
• 29:45
  And then there's an engineering challenge because we've got all these different knobs that we can turn in our lab.
• 29:50
  Things like pressure, things like the nanostructure of light.
• 29:53
  Yeah.
• 29:54
  Let me just say that again.
• 29:56
  So there's also an engineering challenge, right?
• 29:58
  Because we have these various knobs that we can turn in our lab.
• 30:01
  Things like pressure, things like the nanostructure of the material, light, lasers, just like pile a bunch of lasers on it, right?
• 30:09
  To control that superconductivity, but our ability to predict how each of these knobs...
• 30:15
  affects that material is pretty limited.
• 30:18
  Right?
• 30:18
  So what this perspective paper is doing is saying, you know, there's actually no physical law, theoretically, that is preventing room temperature superconductivity.
• 30:27
  No one's come out with a, like, this is impossible paper.
• 30:31
  Right?
• 30:31
  So it's definitely there.
• 30:33
  And superconductivity...
• 30:36
  Superconductivity has been observed in so many different materials, in so many different conditions, and it's almost like a generic property of materials at some point.
• 30:45
  If you lower the temperature down, if you increase the pressure enough, things are going to become superconducting.
• 30:51
  So the idea is we want to be able to now gear our research towards creating something as a community.
• 31:00
  Yeah.
• 31:00
  Right?
• 31:00
  Rather than just going off in all these different tangents trying to do our own thing, let's have a sort of Human Genome Project style moment.
• 31:08
  Yeah, okay.
• 31:09
  Right?
• 31:09
  Where all of these different researchers from around the world sort of get together.
• 31:14
  We plan out how are we going to get there.
• 31:16
  Yes.
• 31:16
  Right?
• 31:16
  Right.
• 31:18
  And the first task is always to improve computer-aided models.
• 31:22
  So not just like predict random stuff that we can't cook, but maybe things that we can.
• 31:27
  Things where the recipe actually makes sense.
• 31:29
  We've got the tools to do it.
• 31:31
  And there's a light at the end of the tunnel.
• 31:34
  In software, there is this phrase or this rallying cry of it's time to build.
• 31:40
  Yeah.
• 31:41
  It sounds like in the superconductor community, the rally cry is it's time to cook.
• 31:47
  Exactly.
• 31:48
  Yeah.
• 31:49
  It's time to cook.
• 31:50
  Exactly.
• 31:51
  And the other thing that they really highlight in the strategy paper is that AI is here.
• 31:56
  Yeah.
• 31:56
  And we need to now start leveraging that in the way that AlphaFold kind of leveraged AI to solve protein folding, which for the longest time it was like, oh, this is an impossible problem, right?
• 32:07
  There's two to the 300 different configurations.
• 32:10
  There's more atoms in the universe than they are.
• 32:13
  There's more atoms in the universe than there are structures that a protein can take.
• 32:18
  And so it's an impossible problem.
• 32:19
  Well, AlphaFold is pretty good at like 90% of proteins now, right?
• 32:24
  So we need to start leveraging these technologies to our advantage.
• 32:29
  Right, right.
• 32:29
  And this dovetails with, I mean, we don't look to always cover AI in our stories on the pod.
• 32:36
  Right.
• 32:37
  It is just a factual reality that so many labs and researchers, especially things that are breaking through, happen to have a companion AI component.
• 32:49
  And it's just kind of the way that it is.
• 32:52
  Yeah, yeah.
• 32:53
  Especially for use cases where it's like protein folding or where you just want something that can go through like almost like there's a brute force kind of pathway where it's like throw compute at it,
• 33:06
  you'll be able to get down to a more narrow band of options than we otherwise would be able to without being able to just throw compute with you at the problem set.
• 33:16
  And that's kind of like a very level one implementation, and there's also level two, three, four, and five.
• 33:21
  But superconductors are going to totally change the face of humanity on this planet if we can get to a way to build them in a way that's...
• 33:32
  scalable uh in terms of format from a manufacturing capacity yeah exactly we had we had this whole hype cycle with lk99 where everyone was doing the think pieces of oh my gosh it's gonna mean for everything that one can i just say i i looked at that archive paper and i was like there's no way like we we should do like a joke episode on it sometime of just like how to recognize nonsense how to sniff the bs uh
• 33:59
  Bunch of four, again, really fun papers today.
• 34:02
  You can kind of see how the difference between our deep dive episodes where we really break down the fundamentals.
• 34:07
  So it's not just sometimes in the rundowns, I can feel like, well, this is just what they're saying.
• 34:14
  This is why we have the deep dives to make sure we can understand the real fundamentals that build up the knowledge base to be able to have confidence in making these conclusions.
• 34:23
  Yeah.
• 34:23
  from these studies so we touched on the 3d scanning for ants incredible use of a particle accelerator for 3d model generation yeah i mean i might want to use that for my 3d gaming so how how how unnecessary it is to use a particle accelerator to do like 3d video game yeah mini particle accelerator in the back it's helping us model this we followed that up with growing chickpeas on the moon if you could go to the moon mm-hmm
• 34:47
  And know you're coming back and it's fine.
• 34:49
  And it's like as safe as air travel.
• 34:52
  As safe as air travel.
• 34:54
  And you don't have to pay.
• 34:56
  Yeah, I'd probably do it.
• 34:57
  Go.
• 34:57
  Yeah.
• 34:57
  If it's as safe as air travel.
• 34:59
  Yes, I think I would do it.
• 35:00
  I would go.
• 35:01
  Dude, the earth is going to be the size of the moon.
• 35:04
  I'm so excited.
• 35:05
  You know, imagine looking at the earth and it's just like, oh, I can like cover it up with my...
• 35:09
  thumb hey jeff if you want to do another launch but instead of another identity based cohort you want to do people of color where we are very yeah yeah i'll i'll be your token you know whatever for the marketing we'll go but like there needs to be like 100 launches before because i need i need to start with orbit we'll just we'll just do a little small orbit or we're being a little facetious although he is a princetonian so that's part of the connection
• 35:34
  We hit the cosmic anomalies.
• 35:36
  I love these telescopes, both ground-based and space-based, and the whole ecosystem around them.
• 35:43
  It's always fascinating.
• 35:44
  The data continues to give us value and room temperature superconductivity.
• 35:50
  Which I was going to ask you a question about, can you explain resistance?
• 35:53
  But that's for a deep dive episode.
• 35:55
  So if you want to know more about superconductors, let us know in the comments because I want to know more.
• 35:59
  Yeah, we could do a deep dive on the history of superconductivity.
• 36:02
  It's a fascinating tale.
• 36:04
  BCS theory is what it's called from the University of Illinois Urbana-Champaign.
• 36:09
  Very proud of their work there.
• 36:12
  They have a little plaque in the physics department.
• 36:15
  Like this is where BCS theory was.
• 36:17
  I thought you were talking about college football for a second, but you were not talking about the old BCS bowl or whatever it is.
• 36:23
  I clearly don't watch college football.
• 36:25
  I am your host.
• 36:27
  Lesser Nari joined as always by my co-host and our resident PhD, Krishna Chowdhury.
• 36:32
  As you can tell, we are having a fun time on this pod.
• 36:34
  We really appreciate you.
• 36:36
  We will see you all next week.