Latest / Paths of Curiosity in an AI World / Q4 - AI in Engineering, Science and Discovery: Mysteries at Every Scale with Michael Zevin, PhD

Transcript

Q4 - AI in Engineering, Science and Discovery: Mysteries at Every Scale with Michael Zevin, PhD · Paths of Curiosity in an AI World

• 0:04
  SPEAKER_00: Welcome to Pass of Curiosity in an AI world where we explore Tamara's careers through the lens of AI.
• 0:12
  SPEAKER_00: Somewhere in the data from the most sensitive instruments ever built, plane noise can look almost exactly like two black holes colliding.
• 0:22
  SPEAKER_00: Dr.
• 0:23
  SPEAKER_00: Michael Zebin helps teach machines and volunteers to tell the difference.
• 0:28
  SPEAKER_00: Let's get curious.
• 0:36
  SPEAKER_00: Dr.
• 0:36
  SPEAKER_00: Michael Zebin is an astronomer at the Adler Planetarium in Chicago, who studies how black holes form and what makes two of them collide.
• 0:47
  SPEAKER_00: He is one of the lead researchers of Gravity Spy, a project where machine learning and volunteers work together to start real signals from noise and LIGO's data.
• 0:59
  SPEAKER_00: Hearing his project in astronomy and physics from Northwestern in 2020, and then called NASA Hubble, Copley Institute for Cosmological Physics for an Enrico Fermi Fellowship at the University of Chicago.
• 1:24
  SPEAKER_00: Dr.
• 1:25
  SPEAKER_00: Devin, thank you so much for joining us.
• 1:28
  SPEAKER_01: Thanks for having me.
• 1:30
  SPEAKER_00: In 2015, LIGO picked up two black holes colliding more than a billion light years away.
• 1:38
  SPEAKER_00: For someone who has never heard of it, what is LIGO?
• 1:43
  SPEAKER_01: So LIGO is a series, a set of observatories that we have here in the US.
• 1:47
  SPEAKER_01: We also have secure observatories that are operating in Italy and in Japan as well now.
• 1:52
  SPEAKER_01: But LIGO stands for the Laser Interferometer Gravitational Wave Observatory.
• 1:57
  SPEAKER_01: Those are a lot of words, but essentially what these devices do is they are very sensitive to perturbations in the fabric of space-time.
• 2:06
  SPEAKER_01: So extreme events in the universe actually cause ripples in the fabric of space-time itself.
• 2:11
  SPEAKER_01: This is something that was predicted by Einstein in the early 1900s, around 1915.
• 2:17
  SPEAKER_01: And about a century after that prediction, we finally had the technology to measure these disturbances in space-time.
• 2:24
  SPEAKER_01: So that's what these detectors do.
• 2:25
  SPEAKER_01: They're called observatories because they're actually observing events that are happening out in the universe and teaching us about them.
• 2:32
  SPEAKER_00: That is so cool.
• 2:34
  SPEAKER_00: And we will definitely get into that.
• 2:36
  SPEAKER_00: But first, how would you describe what you do today to your student self?
• 2:42
  SPEAKER_01: Yeah, so what I do today, I guess I would say that I'm a uh stellar paleontologist.
• 2:48
  SPEAKER_01: So I was very into the idea of paleontology as a kid.
• 2:51
  SPEAKER_01: I love dinosaurs.
• 2:52
  SPEAKER_01: And in some senses, I'm doing a similar thing today because I'm interested in learning about the fossils that stars leave behind, things like black holes and neutron stars, which we can talk about more, to try to get a better understanding of how those stars live their lives, where they evolved in, what kind of processes happened during their lifetime to shape what they became, and then how their fossils evolved over the cosmic eons to give us the collisions that we see today.
• 3:24
  SPEAKER_00: Walk us through your path and how you ended up studying black holes.
• 3:30
  SPEAKER_01: So I've always been interested in astronomy.
• 3:33
  SPEAKER_01: I love the idea of these questions that don't necessarily have a clear answer.
• 3:38
  SPEAKER_01: And one of the great things about the field of astrophysics is every time you answer something new, you're opening up a whole new box of questions.
• 3:46
  SPEAKER_01: So there's really this kind of never-ending search for understanding of how our universe came to be, what our place is in it, and what are the kind of the cool things that are happening out there that helped shape how humans are today.
• 4:00
  SPEAKER_01: So I did my undergraduate studies in physics and astronomy, um, and I pursued a PhD after working as a science educator for a couple of years at Northwestern University in Chicago.
• 4:15
  SPEAKER_01: And the the path to studying black holes in particular was kind of a kind of an odd one.
• 4:20
  SPEAKER_01: Um I, as a beginning graduate student, I started working a little bit in this field of gravitational waves, which um is what LIGO detects when two black holes collide, for example.
• 4:32
  SPEAKER_01: And I didn't know much about it, um, but decided to give it a go because it sounded interesting.
• 4:38
  SPEAKER_01: And within the first year after I started grad school, we found the first ever gravitational wave event.
• 4:45
  SPEAKER_01: So this was a humongous discovery in the field that was awarded the Nobel Prize in 2017.
• 4:52
  SPEAKER_01: So I kind of just jumped in headfirst into this kind of new cutting-edge field in astronomy, and now all I think about is black holes and how to form them.
• 5:01
  SPEAKER_00: In plain terms, what is a gravitational wave?
• 5:06
  SPEAKER_01: Yeah, a great question.
• 5:07
  SPEAKER_01: So Einstein, as I was mentioning before, re-described what our perception of gravity is in the universe, what gravity really is.
• 5:16
  SPEAKER_01: And it differs from what's typically taught in these kind of you know textbooks that we might be used to from middle school and high school.
• 5:23
  SPEAKER_01: And Einstein's description of gravity is that gravity is an aspect of the geometry of space-time itself.
• 5:30
  SPEAKER_01: So breaking that down a little bit, space is this malleable fabric that can bend and distort in the presence of matter and energy.
• 5:39
  SPEAKER_01: And things moving through this kind of curved fabric of space will travel on curved paths according to how space-time itself is bent.
• 5:47
  SPEAKER_01: So one of the byproducts of this theory is that as things are moving through this malleable space-time fabric, they can create ripples that propagate away, kind of like if you were dragging your hand through a pond of water, it'll create ripples that move away from your finger.
• 6:02
  SPEAKER_01: These ripples are what are called gravitational waves, and they've been predicted in theory since 1915.
• 6:09
  SPEAKER_01: But detecting them is a very, very difficult thing to do because what these gravitational waves actually do is they cause space itself to expand and contract.
• 6:19
  SPEAKER_01: So anything sitting in space like you or me or the planet Earth will stretch and squeeze a little bit as these gravitational waves pass by.
• 6:27
  SPEAKER_01: But the amount that they stretch and squeeze is unbelievably minuscule.
• 6:30
  SPEAKER_01: From a typical gravitational wave event like the first one we discovered in 2015, it would cause the Earth to stretch and squeeze by about the width of a proton, which is 1,000 trillionth of a meter.
• 6:43
  SPEAKER_01: So you need to build really, really sensitive devices that can measure these tiny, tiny changes in distance and characterize them so that we can learn about the objects that are creating on the universe.
• 6:55
  SPEAKER_00: LIGO's arms are four kilometers long, and the change it measures is far smaller than a proton, like you said.
• 7:04
  SPEAKER_00: So how can instruments detect something that small?
• 7:08
  SPEAKER_01: These instruments uh are really some of the most sensitive instruments that we've ever built as humans.
• 7:14
  SPEAKER_01: And the technique that they use is something that's called laser interferometry.
• 7:19
  SPEAKER_01: So they are, you know, observatories, kind of like telescopes, because they're looking at things down the universe, but they operate in a very, very different way.
• 7:26
  SPEAKER_01: And essentially, they use very powerful lasers that are shot down these long vacuum chambers.
• 7:32
  SPEAKER_01: Like you said, each of the arms are about two and a half miles long.
• 7:36
  SPEAKER_01: And laser light has a property called interference.
• 7:39
  SPEAKER_01: When two waves of light interact with each other, they can either constructively interfere and amplify the signal, or destructively interfere and diminish the signal.
• 7:49
  SPEAKER_01: The same thing happens with sound waves, with ocean waves.
• 7:53
  SPEAKER_01: So it's using this property of interference of the wave-like nature of light to measure when these two arms of the detector change their length by infinitesimally small amounts.
• 8:05
  SPEAKER_01: So this is a very simplistic overview of how this works in theory.
• 8:10
  SPEAKER_01: In reality, these detectors have all sorts of different mechanisms to help isolate them from external sources of noise, to amplify the power of the lasers, even turning to quantum physics effects that can help improve the sensitivity to the gravitational wave universe.
• 8:28
  SPEAKER_00: When two black holes do collide, what can you learn about them from the signal?
• 8:34
  SPEAKER_01: So, what happens when these two black holes collide is they emanate a lot of these gravitational waves.
• 8:40
  SPEAKER_01: Um, they're extremely powerful in gravitational waves.
• 8:43
  SPEAKER_01: Uh, it's something that you know doesn't emit any what we call electromagnetic radiation, so light.
• 8:49
  SPEAKER_01: So we can't see these things happen with normal telescopes.
• 8:52
  SPEAKER_01: But with these LIGO detectors, as they emit a lot of these gravitational waves, they cause the detectors to do something like this, to wiggle like that.
• 9:01
  SPEAKER_01: So one arm gets a little longer, the other one gets shorter, and then they wiggle faster and faster until we get the waves from right when they merge.
• 9:10
  SPEAKER_01: So the behavior of the signal that we get can actually tell us a lot about the system.
• 9:16
  SPEAKER_01: Like you were mentioning before, we can tell how far away these things are.
• 9:20
  SPEAKER_01: The first one that we detected was merging about a billion light years away.
• 9:25
  SPEAKER_01: And it's just because these events are just so intrinsically rare in the universe that we need to see very far out to catch enough of them.
• 9:33
  SPEAKER_01: But we're also able to tell things like how heavy the black holes were.
• 9:37
  SPEAKER_01: So the two in the first event that we detected were each about 30 times more massive than the sun.
• 9:43
  SPEAKER_01: And we can even tell things about how the black holes are spinning.
• 9:47
  SPEAKER_01: So black holes actually carry something that we call angular momentum, and the spin of the black holes, the angular momentum, imprints itself in the signal that we detect.
• 9:57
  SPEAKER_01: So with every one of these events, we're able to characterize how far away these things were, how heavy they were, how fast they were spinning, and even the orientations of their spins.
• 10:08
  SPEAKER_01: Um, so we're able to tell a lot from the signal itself, and then the difficult part is trying to take all that information to understand how those black holes came to be.
• 10:19
  SPEAKER_00: As space is mostly empty, how do two black holes end up close enough to merge?
• 10:28
  SPEAKER_01: This is a question that's uh one of kind of the the leading unanswered things within my field right now.
• 10:35
  SPEAKER_01: Uh getting black holes close enough so that they merge is a pretty tricky business.
• 10:40
  SPEAKER_01: Um, gravitational waves, once they get close enough, the energy that they lose from these ripples in space-time will eventually drive them to merge together.
• 10:50
  SPEAKER_01: But if they're too far apart, the energy that they lose is so inefficient that it would take much longer than the age of the universe for these things to merge.
• 11:00
  SPEAKER_01: So you need these things to get to very close separations for gravitational waves to kick in and cause them to merge.
• 11:07
  SPEAKER_01: There are a few ideas of how this can happen.
• 11:10
  SPEAKER_01: Um, kind of two of the leading ideas is one is that two big stars are born in isolation.
• 11:16
  SPEAKER_01: So they're a binary star system orbiting around one another.
• 11:20
  SPEAKER_01: And throughout their lives, they go through many processes where one star is gobbling up material from the other, the other star is gobbling up material from the other one.
• 11:28
  SPEAKER_01: Sometimes one star can expand and envelop both of the stars in something that we call a common envelope that can, through friction, drive these cores closer together, and then lead to these black holes that are really close separations.
• 11:42
  SPEAKER_01: Another way that this can happen is in what we call dynamical environments.
• 11:46
  SPEAKER_01: There are things called globular clusters out in the universe, which are kind of dense balls of stars, and black holes that form from those stars will kind of sink towards the cores of these clusters and enter a black hole moshit of sorts where strong encounters between them can cause two of these black holes to come together and possibly merge.
• 12:12
  SPEAKER_00: How do you tell from the signal which path a pair of black holes took?
• 12:18
  SPEAKER_01: So, this is one of the things that I'm I'm most interested in in my particular research is how can we tell from one of these signals whether these black holes formed in kind of the galactic suburbs, like I was talking about, just two stars evolving on their own, or in one of these metropolises, like inside of a globular cluster.
• 12:36
  SPEAKER_01: And there are subtle hints within the signals that can help tell us about those different formation pathways.
• 12:43
  SPEAKER_01: So, tying it back to the stellar paleontology, it's trying to take the information that we get from these black holes to learn about where these things actually grew up.
• 12:55
  SPEAKER_01: And the ways that we can do this are based on some aspects of how heavy the black holes, the spins of how of black holes can actually tell us quite a bit because in one of these clusters, the spins are expected to be totally randomized, whereas if these two stars are evolving together, the spins might be aligned with one another.
• 13:13
  SPEAKER_01: So the information of the spins from the black holes can tell us a lot.
• 13:17
  SPEAKER_01: And even some things like how stretched out the orbit is, how eccentric the orbit is, when the two objects merge, can help tell us whether it formed in a cluster or in the galactic suburbs in isolation.
• 13:32
  SPEAKER_00: What does knowing how a pair formed tell us about the stars and galaxies they come from?
• 13:39
  SPEAKER_01: Yeah, so as I was saying before, it can help tell us the environments within a single galaxy in which they came from.
• 13:45
  SPEAKER_01: Another thing that they can help teach us about is something called the metallicity evolution of our universe.
• 13:52
  SPEAKER_01: So when astronomers talk about metallicity, um, since the universe is mostly made of hydrogen and helium, we are talking about the abundance of everything in the periodic table higher than hydrogen and helium.
• 14:05
  SPEAKER_01: So everything above helium is a metal to us in astronomy.
• 14:09
  SPEAKER_01: And the amount of metals inside of uh stars affects how they live their lives very significantly.
• 14:16
  SPEAKER_01: Um, so it can cause winds in these stars to be much, much stronger when they have lots of metals in them.
• 14:23
  SPEAKER_01: So not just the environments within a particular galaxy, but the masses of all these black holes that we're seeing and how that changes as a function of time in the universe's history can help teach us about how metals were synthesized and populated within our universe as well.
• 14:41
  SPEAKER_00: Beyond black holes, what else has gravitational wave astronomy taught us about the universe?
• 14:48
  SPEAKER_01: So there are lots of things that we can feasibly see with gravitational waves, but the other thing that we've seen so far with these LIGO detectors is not the merging of two black holes, but the merging of two objects called neutron stars.
• 15:02
  SPEAKER_01: Now, these neutron stars are like remnant cores that are left behind when a big star ends its life.
• 15:08
  SPEAKER_01: And if you piled a little more material onto these neutron stars, they might collapse down into a black hole.
• 15:14
  SPEAKER_01: Um, but in 2017 we saw the merger of two neutron stars, and these are really, really extreme objects.
• 15:22
  SPEAKER_01: They're about the size of a city, maybe 10 miles across, which is tiny in astronomical terms, but way more than the sun.
• 15:30
  SPEAKER_01: So they're the densest objects that we know of in the universe.
• 15:34
  SPEAKER_01: And when these things actually collide with one another, they create effectively a thermonuclear explosion called a kilonova.
• 15:41
  SPEAKER_01: And this is something that we can see with telescopes, something that we can see with light.
• 15:45
  SPEAKER_01: So this event in 2017, when we saw the gravitational waves from these two neutron stars merging.
• 15:51
  SPEAKER_01: Telescopes around the world and in space also witnessed the explosion that ensued in this distant galaxy that was associated with this event.
• 16:01
  SPEAKER_01: Um, and this this was actually the first direct evidence we had of heavy elements being synthesized in the universe.
• 16:09
  SPEAKER_01: So it turns out that um elements like gold and silver and platinum and uranium are actually formed in these kinds of cataclysmic collisions in the universe, which was theorized before but never directly observed.
• 16:22
  SPEAKER_01: So this particular event that we saw with the gravitational waves and with light was able to help fill out the periodic table and tell us where some of the heavier elements that we know of are actually synthesized in our universe and how they got here to Earth.
• 16:38
  SPEAKER_00: Which creates stronger gravitational waves?
• 16:40
  SPEAKER_00: The collision of black holes or neutron stars?
• 16:45
  SPEAKER_01: It's a great question.
• 16:46
  SPEAKER_01: So black holes do create stronger gravitational waves.
• 16:49
  SPEAKER_01: The heavier the objects are, the stronger the gravitational waves are.
• 16:53
  SPEAKER_01: So that means we can't see neutron stars colliding quite as far out as we can see black holes colliding.
• 16:59
  SPEAKER_01: But they contain a lot of extra science that we won't be able to get with just black holes.
• 17:04
  SPEAKER_01: And to give you a little perspective, the uh energy emitted from the two black holes that we first saw, like I said, this is completely invisible, we don't see any light from it.
• 17:16
  SPEAKER_01: But the energy emitted in gravitational waves from these black holes spiraling towards each other was more energy than all the stars in the observable universe put out at a given second combined.
• 17:29
  SPEAKER_01: So humongous amounts of energy, but it's all locked into these ripples in the fabric of space-time.
• 17:37
  SPEAKER_00: Wow.
• 17:38
  SPEAKER_00: LIGO is also sensitive enough to pick up things that aren't gravitational waves.
• 17:44
  SPEAKER_00: What is a glitch?
• 17:48
  SPEAKER_01: So, yeah, like you were saying, because of the sensitivity that we need to make these detectors that to see things from space, they're also susceptible to lots of noise.
• 17:58
  SPEAKER_01: And by noise, we mean things that can be aspects of the instrument, so you know, a magnetic field that's coming out from one of the things in the instrument causing some noise, or environmental noise, things like strong gusts of wind, or cars driving nearby, or even earthquakes on the other side of the world.
• 18:15
  SPEAKER_01: LIGO actually sees all of these things too, and the ones that happen that are transient, that happen randomly, rather than underlying noise that we know we have in the detector, those are what we call glitches.
• 18:29
  SPEAKER_01: Um, understanding these glitches is really important because if we don't understand the noise, we can't get rid of the noise, and it makes it much harder to see things in the universe.
• 18:40
  SPEAKER_01: Trying to understand all the different kinds of glitches, these noise artifacts that we have, is pretty essential for us to be able to do the science that we want to do.
• 18:49
  SPEAKER_00: You helped start Gravity Spy as a graduate student.
• 18:53
  SPEAKER_00: Can you explain what it is?
• 18:56
  SPEAKER_01: So, Gravity Spy is a citizen science project, which means we're counting on volunteers from around the world to help us analyze data.
• 19:05
  SPEAKER_01: And it's specifically looking at these glitches that I was talking about.
• 19:08
  SPEAKER_01: We get glitches in our LIGO detectors way more often than we get real astrophysical signals.
• 19:14
  SPEAKER_01: And characterizing and classifying these glitches is something that is a difficult task for an individual research to do just because we have so many of them.
• 19:24
  SPEAKER_01: But we are able to project these glitches into an image that we can look at to see what this looks like in a map in time and frequency.
• 19:32
  SPEAKER_01: And they could look like little teardrops or look like fish in some cases, or look like wiggly lines.
• 19:39
  SPEAKER_01: So we have taken all of these glitches that we get within LIGO and given them to the general public as a means to help better understand which ones are important, which ones are affecting our detectors, and possibly giving us the means to actually remove them from the instruments themselves to better improve our sensitivity for the things that we really want to see.
• 20:02
  SPEAKER_01: So that's what Gravity Spy is.
• 20:03
  SPEAKER_01: It's this project that is taking a lot of the data from LIGO and letting volunteers from the general public help us understand it.
• 20:12
  SPEAKER_00: Volunteers found glitch types the machine had never shown before.
• 20:16
  SPEAKER_00: Like 1,080 lines and 1,400 ripples.
• 20:20
  SPEAKER_00: How does a person catch something a machine wasn't trained to look for?
• 20:26
  SPEAKER_01: So, as we'll talk about a little later, I think uh Gravity Spy does use AI techniques.
• 20:31
  SPEAKER_01: It uses uh machine learning algorithms called neural networks to help it do a very quick classification of all the glitches that are coming through it.
• 20:40
  SPEAKER_01: Um, but machines are still not always the best at identifying new types of objects or new types of glitches within the detectors, whereas the human eye and the human intuition can be really good at seeing these kind of pattern recognition or identifying anomalies that they're not used to seeing.
• 21:01
  SPEAKER_01: So volunteers have been really essential in helping to find new glitch types that have appeared as the detectors have evolved and become more sensitive.
• 21:10
  SPEAKER_01: So it's really kind of a tandem of using research, uh, using citizen science researchers as well as using machine learning and AI techniques together that can help improve the you know the characterization of all of these noise artifacts.
• 21:27
  SPEAKER_00: Do you think algorithms or AI are getting better than people at catching the unexpectedness in this data?
• 21:35
  SPEAKER_01: I don't think that they're getting better than people.
• 21:38
  SPEAKER_01: They're getting better at different tasks.
• 21:41
  SPEAKER_01: So I still think there is um I still think there is a lot of room for both human classification and AI, and they can work together in really magical ways.
• 21:52
  SPEAKER_01: Um but I think that AI is still isn't able to really capture the The human intuition and pattern recognition and anomaly identification that human brains can right now.
• 22:07
  SPEAKER_01: So pairing them together is really the most powerful thing that we can do in these kinds of projects because both have benefits that can help each other.
• 22:16
  SPEAKER_00: 10 to 20 years from now, what, if anything, will someone doing your job be doing differently because of AI?
• 22:26
  SPEAKER_01: So I mean AI is something that is we've so we've been using AI in some capacity for a while in astrophysics, 15, 20 years maybe.
• 22:39
  SPEAKER_01: And AI has really become in the forefront of the general public because of these things that we call large language models, things like ChatGPT and Gemini and Claude, which are helping our day-to-day lives become much more efficient.
• 22:53
  SPEAKER_01: And they also help with making silly pictures and stuff like that.
• 22:58
  SPEAKER_01: So it's impossible to ignore AI as a tool.
• 23:03
  SPEAKER_01: And one of the things that AI can do to help us as researchers, just in our day-to-day lives, is helping with efficiency.
• 23:13
  SPEAKER_01: So one of the things that we do quite a bit in astrophysics is coding.
• 23:17
  SPEAKER_01: And because so much code is open source, available to the public, it can be adjusted by these large language model algorithms.
• 23:26
  SPEAKER_01: So the efficiency at which you can code either simulations or analyses has just skyrocketed because of the use of these AI tools.
• 23:37
  SPEAKER_01: But AI is also much broader than what we typically think about it as is these large language models.
• 23:43
  SPEAKER_01: There are lots of different AI techniques that are essential for doing identification, classification, characterization.
• 23:51
  SPEAKER_01: And with upcoming observatories and astronomy, we're going to be getting so much data that it would be impossible for humans themselves to look through all of it.
• 24:00
  SPEAKER_01: So using AI to help understand these large expansive data sets of tens of billions of galaxies and helping to identify the anomalies or the interesting things within these data sets is really going to be an essential tool for us to be using to do astrophysics research.
• 24:21
  SPEAKER_00: What is one concrete thing a student could do in the next year to move towards this kind of work?
• 24:29
  SPEAKER_01: So I think I think one thing that many people, including myself, didn't realize before getting into being an astrophysicist is that a lot of what you do is based in computers now.
• 24:41
  SPEAKER_01: So whether it's you know scheduling what a telescope is going to be look at, to taking the data that a telescope has received, to doing the analysis of that data, to running simulations of galaxies merging, this is all done through different coding mechanisms.
• 25:00
  SPEAKER_01: So learning, you know, how to code and how to use these AI tools to help you code in a more efficient way, I think is really going to be essential for people that are interested in going into the field moving forward.
• 25:15
  SPEAKER_00: What is the biggest question in your field that no one can answer yet?
• 25:20
  SPEAKER_00: And is AI getting any closer to it?
• 25:24
  SPEAKER_01: That's a good question.
• 25:25
  SPEAKER_01: Um there are, I'd say the the biggest questions, so there are a few.
• 25:35
  SPEAKER_01: I mean, there are a lot of big questions in astronomy.
• 25:38
  SPEAKER_01: Some of the big ones that come to my mind are um is there life out there in the universe?
• 25:43
  SPEAKER_01: Is there intelligent life out there in the universe?
• 25:46
  SPEAKER_01: Um, what happened before the Big Bang?
• 25:49
  SPEAKER_01: And what are all these kind of missing pieces of our universe that we don't understand yet?
• 25:53
  SPEAKER_01: Like dark matter and dark energy.
• 25:55
  SPEAKER_01: These are all kind of essential questions about, you know, answering some of the fundamental things that we think about as humans, like why are we here?
• 26:02
  SPEAKER_01: Are we alone in the universe?
• 26:04
  SPEAKER_01: And as of right now, AI, like I was saying, can be used as a tool to help outer investigate these questions, but I don't think they're things that AI can solve on its own.
• 26:15
  SPEAKER_01: We really need the creativity of the human brain to be able to start to probe these questions in more and more detail.
• 26:21
  SPEAKER_01: But what AI can do is it can take all this data that we're gaining from all of our observatories around the world and in space and help us process that data in a more uh in a way that might be more apparent what the answers to these questions are.
• 26:37
  SPEAKER_00: What a nice partnership.
• 26:40
  SPEAKER_00: What is something your work has required that most people would never guess?
• 26:46
  SPEAKER_01: Ah, so being an astronomer at a planetarium, um, one thing that I do quite a bit outside of my research is engaging with the public.
• 26:56
  SPEAKER_01: So um explaining hard science concepts uh to people from all walks of life, um, and doing a number of kind of partnerships and news interviews and things like that.
• 27:08
  SPEAKER_01: So some of the you know odd things I've done is work with, for example, um a chef at a Michelin star restaurant to design a space themed menu.
• 27:18
  SPEAKER_01: So I didn't think I would be doing these kinds of things.
• 27:20
  SPEAKER_01: Um but it's it's a lot of fun.
• 27:22
  SPEAKER_01: And uh you know, upcoming I'm gonna be working with uh Botanic Garden on helping to design a space-themed orchid show.
• 27:33
  SPEAKER_01: So and collaborating with them then.
• 27:35
  SPEAKER_01: So there are a lot of fun things that I get dragged into that way.
• 27:37
  SPEAKER_01: As far as my research life goes, I think many people don't realize how much coding and computing goes into what you do now in most scientific fields.
• 27:50
  SPEAKER_01: So um I personally love doing that stuff.
• 27:52
  SPEAKER_01: It's like uh fun problem solving, um, but it's something that you should be you know aware of that you would need to have some experience in outside of just kind of the math and the physics that goes into astrophysics.
• 28:05
  SPEAKER_00: Before you leave, are there any books, podcasts, or other media you would recommend for a student who wants to learn more about this field?
• 28:14
  SPEAKER_01: Yeah, so there's a popular astronomy blog called Astrobytes.
• 28:19
  SPEAKER_01: Um I used to write for it when I was a graduate student, but it um it uh every day takes uh a scientific research paper that was put out in astronomy and distills it into a kind of more digestible level.
• 28:32
  SPEAKER_01: So it's a great way of, you know, outside of what's in the news, hearing about some of the new discoveries that are happening in the field.
• 28:39
  SPEAKER_01: Um a book I recently read, which I I really enjoyed, um, is uh a short history of nearly everything by Bill Bryson.
• 28:48
  SPEAKER_01: It talks about some things that are astrophysics, but also things that are related to evolution and um discoveries that we make about our own planet.
• 28:57
  SPEAKER_01: Uh but I think uh one of the nice things about it is it kind of tells the the personal story of how these discoveries are made.
• 29:04
  SPEAKER_01: These things aren't done in a vacuum, scientists are people.
• 29:07
  SPEAKER_01: And learning about the kind of human side of how these scientific discoveries are made, I think, is uh really interesting.
• 29:15
  SPEAKER_00: Well, thank you so much for joining me today.
• 29:18
  SPEAKER_01: Yeah, thanks for having me on.
• 29:20
  SPEAKER_01: This was fun.
• 29:25
  SPEAKER_00: This has been Paths of Curiosity in an AI World, exploring careers through the lens of AI.
• 29:32
  SPEAKER_00: Until then, stay curious.