Podcasts Season 3

S03E03: How AI is transforming astronomy (and other stories), with Dr Sara Webb

The idea that artificial intelligence is taking our jobs can be scary – but in actuality, there are cases where this is a good thing. Dr Sara Webb (Swinburne University of Technology) shares one of these stories in today’s episode, which begins with a TedX talk in Melbourne and ends with a discussion of some of the many ways techniques developed for astrophysics are transforming seemingly unrelated fields. Sara is an astrophysicist based at Swinburne University and is also a published author with a talent for communicating complex ideas about our universe (and AI) for broad audiences. Listen in to hear more about the role AI is increasingly playing in astronomy, how she got into astrophysics in the first place, and more in this wide ranging episode that paints a picture of what a career in STEM can look like.

Listen and subscribe on Apple Podcasts, iHeartRadio, PocketCasts, Spotify and more. Five star ratings and positive reviews on Apple Podcasts help us get the word out, so please, if you enjoy this episode, please share with others and consider leaving us a rating!

Episode credits:

Guest: Sara Webb

Co-hosts: Zena Assaad and Liz Williams

Producers: Zena Assaad, Robbie Slape, Liz Williams, Martin Franklin (East Coast Studio)

Transcript:

Liz: Hello, listeners. Thank you so much for joining us today. We are in the Australian National Center for Public awareness of science. And we have a very exciting guest on today.

Zena  

Yes we do. We have Dr. Sarah Webb today. So for anybody who doesn’t know Sarah, she is an astrophysicist at the Swinburne University of Technology. And she’s also a Mission Director for the Space Youth Innovation Challenge. Now, I met Sarah last year at the Women in AI awards. And I started following her after that, because she is so so active on social media specifically for science communication. She does really incredible things like Tiktok or instagram videos all the way through to articles in the Conversation and Cosmos magazine. She’s so incredible. And I just absolutely loved having her on the podcast today.

Liz 

Yeah, no, it was wonderful. One of the my favorite moments was when she shared the experience she had of actually creating one of those early videos. What about you, Zena? Was there a moment that you particularly caught you in this episode? 

Zena  

For me, it was definitely when she was talking about AI and how it’s changing her job as an astrophysicist. And I think the reason it was my favorite part is because she kind of spoke about how opportunistic it was and how great it was. But she also was really honest and transparent with some of the challenges of having AI kind of embedded as a primary tool within astrophysics and how that’s kind of changing what it means to be an astrophysicist. And I really love that. 

Liz  

Well, it was definitely a fascinating conversation and you are gonna get to listen to it right now.

***

Liz  

So Sara, thank you so much for joining us today. We really appreciate you coming here on the Algorithmic Futures Podcast.

Sara  

I’m excited to be here.

Zena  

I’m excited to talk to you about all of the amazing things that you do. And one of the amazing things on that list was that last year you did a TED talk. And the TED talk was called “Why I Want AI to take my job”. Is that right?. So can you talk to me a little bit about how that TED talk kind of came about and why you do want AI to take your job?

Sara  

So I had gotten approached by the organizers of TEDx Uni Melbourne. One of them had been following me online for a couple of years and had followed that, you know, not only was I an astronomer, but I started talking about machine learning a lot. And they realized, oh, oh, machine learning is huge in astronomy, they didn’t realize and then they they’d followed me along, as I talked about, like transitioning from pure astronomy into some other different contexts. And so they were really, really kind. And they were on the Organising Committee and emailed me and went, Hey, I don’t know whether you’d want to do a TED talk. But if you did, what would you want to talk about? And, you know, we’re trying to have it be about the edge and what we’re thinking, you know, edge computing, other things like that. You know, would you want to be talking about astronomy or AI? both, or neither? And so, I replied back went “Oh, my gosh, I would love to do one. And I think I’d actually like to do on on just AI in general. So why I think AI is important to take our jobs. And I would put it in the astronomy context and how in astronomy we, if artificial intelligence and machine learning algorithms weren’t as powerful as they are, we couldn’t get through our data.” It’s impossible. And the fact that astronomy is the perfect kind of sandbox for testing complex and almost sometimes risky algorithms like generative algorithms, in a risk free environment, and so that’s how that kind of came about, and it was really cool. I was really glad that I got to do that, and kind of shape it into something I’m very proud of.

Zena  

You know, I think the conversations around AI, in my opinion always fall on like, a hard “I love it” or a hard, “I hate it.” And I think it’s the end of the world. And so I think your TED talk really kind of fell into the it’s actually okay. And it’s not bad kind of bucket. What was the reception that it got? 

Sara  

Yeah, it was really good. So the people who were in the audience, had quite a few people come up to me afterwards, saying that they hadn’t thought about the fact that AI might take something away from them. So part of the talk was that it’s taking parts of our job, yes, but it means it’s giving back time and energy to do other things in our jobs. And a lot of people kind of appreciated that context. And then a lot of other people saw how potentially AI could replace or help them do what they’re doing. And it wasn’t just astronomy, it was people in medical science and, and different types of areas, who, who kind of got a little bit of a new understanding of the different types of machine learning. So we talked about supervised and unsupervised. And it’s kind of like a mini Crash Course. And I think it made, at least from the feedback, people were like, oh, it’s not so scary. Like when people say machine learning or artificial intelligence, I don’t know, I think it’s like a brain in a computer that’s doing things but you know, it’s okay, a human designs it, and I think that that was probably the biggest highlight was that people had a little bit more understanding about something that — you’re so right — traditionally, people have kind of a black and white, either it’s good, or it’s bad. And I think the media can also either have, it’s good, or it’s bad, depending on what day it is, what policies have come out. And I think if you’re just a normal consumer of being alive and existing on this planet, it’s really hard to determine ‘is AI good or bad?’ And sometimes, you need to be able to talk to experts to kind of nut out the nuances of “it’s neither, it’s just a tool.” 

Liz

So I do want to talk a little bit more about what it actually looks like to make use of AI in your job. And what it allows you to do that maybe you didn’t have time to do before, or maybe you didn’t have enough time to do before. What does that look like?

Sara

So it was specifically in my astrophysics career. A lot of the artificial intelligence algorithms that we’ve designed, are all around reducing the amount of time it takes us to look at data. So astronomy is very data intensive, lots of different types of data that are coming in. And it’s impossible. If we were to look at every single bit of data, we’d be here for the rest of our lives — beyond our lives, it’s too much data. And so we really harnessed — the research groups and our work with — both supervised and unsupervised learning. So supervised learning to look for things we already know, what they should look like in the data, whether it’s images, or whether it’s time series, to help us classify very quickly. And then my expertise was in using unsupervised learning to be able to find anomalies and things that were unexpected or unknown previously within data. And so that was a massive help. Because if we hadn’t have used that, we would have missed so many different objects that we hadn’t seen before, within our datasets, just because we didn’t know that we should be looking for them.

Zena 

I want to focus in on that point for a second. You know, the AI picked up something that you missed yourself. And one of the criticisms of any kind of automated tool is the loss of nuance. People talk a lot about how you know, when human beings are analyzing data, yes, we can’t analyze as much in as quick timeframe as a generative tool. The benefit of having a human to it is that we have the element of nuance, like we can find these really finer details that might be outliers, or might not be outliers. And that’s something that an AI tool can’t really do. So can you talk to me a little bit about how you kind of manage that when you’re using AI tools in astrophysics? 

Sara  

What you say about that nuance is really important. And I think it’s important to highlight that pretty much every major astronomical or astrophysical discovery that has ever been made has been made by accident. You know, we work very, very hard, and we build amazing instruments and tools. But most of the time, it’s things that we find that surprise us. And throughout history, you can look at every every type of detection that we’ve had through from supernovae, exploding stars, to burping black holes to flaring stars, even neutron stars, these stars we didn’t know could exist. All this has found by accident in data by somebody looking at it, and understanding the nuances of how that data was collected and what the physics might be. And so you need that. You need to kind of like the human inquisitiveness to be able to look at data and really dive into it. And you do if you’re using a supervised algorithm where we’re training just on things that we already know, you’re losing that. And so that’s something that was really important to me. And when we started to look at different unsupervised algorithms, so you’re not telling it what it’s looking for, you’re just asking it to find patterns in the data. And there’s some unsupervised algorithms where you can have it explicitly outline and highlight noise in your data. So things that it thinks is not like anything else, it’s kind of like that playing what is the odd one out, the algorithm does that for you. And so a few of our different research papers that we’ve done, have been looking into that noise. So you have things that are grouped and clustered together that are known, pretty much, and then you have a big group of stuff that was unclusterable. No one knew what it is. So then how do you use that? And how do you kind of make an anomaly score to be able to rank what a human should look at. And by doing that, we were able to take things that still the machine didn’t know that it was important, but a human when they looked at it, they did. So that’s kind of how we handled that was understanding that the machine was not going to find everything, and that we still had to look at it as a human. But how do we intelligently figure out the best way and the best process for humans to look at it?

Zena  

I think I like this element of serendipity that comes with astrophysics. So most things were found by accident. Do you feel like there’s a loss of that serendipity with AI generated tools? 

Sara  

Sometimes I do. Yeah, I think that’s probably one of the biggest concerns for astronomy in the future is we’re building mega facilities, massive, massive telescopes in all sorts of wavelengths from optical to radio. The optical telescopes will generate about 10 million alerts per night of things that are in theory a human should look at, and we can’t. So we have to build in these these algorithms. And I think it’s a concern of many people who are working on this on these different things is well, how do we make sure that we’re not losing anything that a human might deem as important? And it’s really hard, because eventually, we’re going to get to a point where astronomers won’t look at data how they used to, and actually, I co-supervise a PhD student who’s working, trying to understand this problem in astronomy is that an astronomers job today is very different than it was 10, 20, 100 years ago. And in 10, 20, 100 years from now, their job is going to be completely different. And what does that mean? How do we need to prepare? And how do we need to prepare to be able to not only train algorithms, but prepare humans to do that job. And of course, astronomy is just a small example. But this is true for almost every career that there is now: it’s different now than it was in the past. And it almost certainly will be completely different in the future. And so how do we how do we prepare for that? Which is tricky, and I don’t have the answer.

Zena  

I think that’s something I always say — I always say AI isn’t going to take your job, it’s going to augment your job. And that’s the case for any technology. 

Sara

Yeah, absolutely. And it’s the case for, like anything that we’ve seen in human history, the industrial revolution, it changed the way jobs worked, or the invention of the automobile that eliminated a heap of jobs, but made a heap of other ones. And I think it’s– people can be scared or worried. I think when you’re in the when you’re in the here and now and you’re thinking that something you do or your your profession might not exist how it is now then that’s a scary thought. But it’s happened for millennia. And it’s going to keep happening. And this time, it’s just technology. It’s, it’s algorithms rather than machinery or whatever else it was.

Liz  

What got you into astrophysics? What drew you to the field?

Sara  

I love astrophysics so much. So I think I got into physics kind of out of spite, because in high school, I had — I changed classes in grade 11. So in Australia, you kind of pick what your what your senior years are going to be. And, like maybe a quarter of the way through my grade 11. So my, my second to last year at school, I had changed from law to physics, because I realized law was just not for me. I love it. I think it’s interesting, but I really liked physics. And my physics teacher at the time had said I don’t think that’s a good decision, had like sat down with my mother at a parent-teacher interview while I’m there and said, I just don’t think she’s cut out for it. I don’t think she can — I don’t think this is a good decision.

Zena  

I love a spiteful story. They’re my favourite.

Sara  

Spite can make you do a lot of things but I loved it. I wouldn’t have done it if I didn’t love it. But that put kind of like a fire in my belly because I’m like, Who are they to tell me that I can’t give something a go? And I loved it. I loved physics, ended up doing very well in it. And then knew that I wanted to study at a university and knew that I really liked astrophysics just because of all of the unknowns. So many questions that we still don’t have answers to just understanding the universe and I loved that and so it was really easy. From there. I just kind of went straight into it. 

Liz

I love it. I have a similar story.

Zena  

The amount of women that we have interviewed on this podcast, and we’ve asked how did you get into your field and the amount of women that have started with a story like, well, there was this guy who told me…

Liz  

There’s always this question –when you’re introducing new technologies into a field, and you are changing the nature of work, right? There’s, there’s always this question as to like, Are you changing who gets attracted to the field? Are you changing who might might engage with it? Right? Are you changing the kinds of skills that you might be selecting for when, you know, inadvertently set up this like, chain of data processing that ends up providing, like, helping you find what to look for? Right? And so the reason I asked that question is, I’m actually wondering if, if you’ve thought about that at all, in terms of like, who is being drawn into the field? Is AI or are machine learning tools changing who might actually get drawn into doing this kind of work?

Sara  

Yes, yes, absolutely. We are. So part of that my PhD students work is trying to understand, like your, your occupational identity is kind of the words that we’re using for it because an astronomer say 50 years ago, knew that they were going to be looking at telescopes, looking through telescopes, or working with glass plates, or working physically with data that was taken by a telescope. And doing a lot of hands on work, lots of hands on calculations, like the original Harvard computers, the female computers that did all of those painstaking calculations by hand for spectra. And that’s very different now. We would never do anything by hand, pretty much never going to measure a spectrum by hand ever again. And sort of what does it mean now to be an astronomer and and likewise, when I decided that I wanted to do physics, and that I loved astronomy and astrophysics, I had no idea I needed to know how to program a computer. None at all until there was maybe my second year and I did a summer research project. And my supervisor, who was very kind was like, Okay, do you know, what languages do you know? And I laughed and went English, and I’m not very good at Spanish. What do you mean languages? And he was like computer languages. And like, we had a crash course. And he’s like, okay, it’s okay. You don’t need to know right now. But in the next six weeks, you’re going to learn one. Pick one. I picked Python, because we use it a lot in astronomy. And I’ve since had to learn a little bit of things like IDL and C, which I don’t love. But you know, it’s like learning different languages. I had no idea and would I have picked to study astrophysics if I knew that that was a requirement? I don’t know. Because it took me years to build up the confidence to not only feel comfortable programming, but feel like I can do it competently. It’s kind of a very steep learning curve. And I think that’s, yeah, it’s something that’s interesting, in that we’re trying to understand with some of my student Hugo’s work is how did actually doing astrophysics and astronomy, so how did people doing a PhD — what did that do for their career? Did they realize that astronomy is not what they wanted, or not what they thought it was? Do they realize they loved programming, they could do it in any industry? It’s still ongoing research. But something interesting that we’ve we’ve found is that in astronomy, a lot of what we do is visual inspection, lots of looking at data, maybe not physically working with it with your hands, but you’ve got to look at it at some point to evaluate it. And it turns out that we’re not really taught how to, nobody sits down and does a course — not like if, say, You’re a stenographer or a radiographer, where you get taught how to look at an image, where do you look and why. That’s not how astronomers are taught. And so part of kind of that research is us trying to understand what we’re moving to automated machines and computers. And we don’t even know how we decide what’s important in an image. That’s kind of scary, because now in 20 years, we might — my PhD student in 20 years or so might never physically look at an image, they might only ever get data from a data broker with a machine learning algorithm. And who knows, whoever designs that, hopefully it was me and hopefully it was, well, there’s still biases in there. And there’s still, Yeah, things that you can’t, you can’t know fully, if that makes sense. I think it is something that I’ve thought about a lot. And something that is exciting, but also a little scary, because you don’t want to lose like what Zena had said kind of that, that inquisitiveness or the human nuance in things and I think this is true for everything. It’s like you hear people talking about self driving cars and the nuances of what is a good decision and a bad decision. And could a computer make that better than a human? And I mean, I think theoretically, yeah, our computer could make that decision. But you know, are we losing human nuance in it.

Zena  

I think so, I want to talk to you a little bit about language. So you do a lot of science communication across different things, you do it across social media, videos, interviews. And there really is an art to the kind of language that you use across different platforms, right, like the kind of academic language we use versus something that you would use on a tik tok video, vastly different. So can you chat to me about some of the challenges, but also some of the skills that you’ve developed with communicating your research across such a vast difference of mediums?

Sara  

Communicating my research and just astronomy and AI in general is like my one of my favorite things to do, because I think it’s challenging, but in the best way, because it’s challenging to do research and to communicate that to our peers, we’re all on the same level, we’re all using the same terminology– that’s already challenging. But then to communicate it to say, a five year old and a 99 year old, across a spectrum of mediums is also really fun. And as you said, really challenging. And so, yeah, I do lots of different types of communication. So one of my favorite is written communication. And for that I do– I figure out who the target audience is, it depends what what outlet you’re writing for, whether it’s something say, for the Conversation, where a lot of the readers are people who want to be informed by experts, so they have a bit more, sometimes a bit more base knowledge than than other people. But still, they’re not experts. And then I also write things that are for children for for people between five and like 15 years old. And that’s a whole breadth in itself. And so figuring out exactly what your audience is that you’re writing for, and then trying to make sure that the language that you use and the way that you explain things, that we often use metaphors, or you’re often giving kind of like visual description, like descriptive examples, making sure that it’s something that they can identify with. So when you’re trying to explain what gravity is, for a child, you might explain it using a trampoline, because they know what trampolines are. It’s something that’s fun and bouncy, and they can picture that very easily. With an adult, you might be able to use it a little bit differently. But yeah, it is it’s still a challenge. And I still find it challenging. Figuring out what is the right communication and some of the things that I do online like making videos and, and through different mediums. So for Tik Tok, or Instagram or for for YouTube, they also have wildly, like wildly different audiences. And I sometimes will have a video where I’m explaining something that –one example is explaining the in billions of years and like four or 5 billion years, the Andromeda Galaxy is going to crash into the Milky Way. Not everybody knows that. And it’s kind of like a wow fact. And I was just explaining it in fairly simple terms, but for a general audience, mainly adult audience. And — it was very lucky, it went viral a couple of times. But with that became an onslaught of comments saying or everybody knows this. And then on sort of comments saying, Oh, my gosh, I had no idea. It was interesting, because some comments were rude to the other side. So someone might be like, I had no idea, someone else might comment and be like, how? And I’m like, well–

Zena  

You’re not allowed to have a difference of opinion on social media.

Sara  

Yeah. And if it’s something that’s not very controversial, just go. But it’s almost a similar thing happened where I made a video that had gone viral about how on a mirror, if you put something quick you can it the way the way the reflection works, you can you can see it in a weird way. And people commenting Oh, my gosh, I had no idea. That’s awesome. And then other people being like, how did you not know? Did you not take physics. And it’s that very niche way of trying to combat that because you want to make it accessible for everybody. And you want to make it a safe space as well. It’s okay if you don’t know things, none of us can know everything. That’s impossible. But you also want to foster an environment–because a lot of the time on social media, the video is one thing, but then the comment section is another thing–where people aren’t being shamed for wanting to learn or for not knowing. And I think that’s probably the most challenging part for me is making sure that I communicate things in a very kind and open way. But then also moderate what people say online because I think, yeah, the internet is a weird place. 

Liz  

I mean, it’s designed to amplify things that we engage with. And unfortunately, you know, we are drawn to things that feel dangerous and controversial. And, you know, like, it’s, it’s amplifying some of our worst tendencies. It also makes it easier for us to not, it makes it easier–I think it can make it easier for us to forget that there are other people on the other side, you know, reading our comments, 

Sara  

Yeah, I think it loses humanity. 

Liz  

Right. That is another topic.

Sara  

But communication is really hard. And it’s really fun.

Zena  

It is really hard. 

Sara  

I think the best thing I’ve learned from it is to just to kind of just do what makes you happy. So for a long time, when I first started making videos, I was trying to make videos that I thought people were going to, were going to be the more watchable ones. So more, not click baity, but the topics that people are like, Oh my gosh, this is breaking news or whatever. And I didn’t–It wasn’t it wasn’t making me happy. And then I started making videos about things that I really enjoy, which are often like nuanced physics concepts, or astrophysics and it’s not breaking here’s a headlining news story. But it’s just like, here’s something interesting. And I found that that brought me a lot more joy. And I wasn’t kind of chasing the algorithm, you might hear that a bit, is people trying to figure out what does the algorithm like? Does it like when you start this way? Yeah, exactly. And people will chase it. And it’s mentally exhausting. And I think I’ve been the happiest when I haven’t. And I just make things that make me happy that then I get good reception that make other people happy or interested or feel like they’ve learned something. And that’s that’s made a huge difference over the last couple of years.

Liz  

Is this the kind of thought that brought out this deck of cards that that you put out? Can you tell us a little bit about it? I mean, it talks about the science behind each constellation? 

Sara  

Sure. 

Liz  

Tell us where that idea came from?

Sara  

Yeah, it actually was born out of, yeah, social media. So a publisher had seen me explain–I don’t even know what video was anymore. But it was explaining something fun to do with with stellar astronomy and the fact that the I think it might have been that the North Star is not always the North Star. And the South Pole isn’t always the South Pole, it changes throughout time. And we have, we have like amazing history from from all over the world, but especially in Australia with some of the song lines from our Indigenous Australians about how the sky has changed over 10s of 1000s of years. And that’s fascinating. And they had maybe seen a video of me explaining something like that and had reached out and said that they had they wanted to do — so the publication house wanted to do this like deck of constellations or deck of stars, and they wanted someone to write it and that they had kind of gone down a rabbit hole — they’d seen the video. And then they went and read all of the different articles I’ve written online. And they really liked the way I explained things, which was very kind and asked, you know, would you like to be commissioned to do it? And I said, Absolutely, I would love to. And so that was really fun, because it was about the constellations, which I just made up patterns in the sky, that we’ve we’ve had a consensus that we said that there’s 88 of them. But the history and the law and what’s in them. So the history of a lot of constellations is pretty fascinating. A lot of them are very, very Eurocentric, very Northern. But most of them that end up being along the equator have very similar storylines between what, like Northern European thought and southern countries and different territories had seen as well. Some constellations that say, look like a dog or look like a warrior have been those things throughout history in different languages and different different places without communication, which I think is interesting. So it was talking about a little bit about the lore behind them. So a lot of the Norse mythology, which is pretty cool. And just how that translates down, but then also what you can see in the constellation, so if you get a telescope or binoculars, what type of stars and like globular clusters? What type of things can you actually see and the astronomy side of it, which was really fun. And that was a really cool experience, because I had a really great relationship with the publisher. And I had already started writing a book book that I knew that I wanted to write for years, and I’d already outlined it and wrote a few chapters. And so I pitched to them. I said, Hey, I’m working on this, I have — I’m starting to pitch it to other publishers. But do you want it? Because I loved working with them so much. And then almost immediately, within a day they said, yes, yes, we would like it, we’ll draft up a contract. And so now I have, later in the year, The Little Book of Cosmic Catastrophes is coming out, which I’m so excited for. And that wouldn’t have happened if it hadn’t been for social media. And it hadn’t been for me trying to communicate something about science, and it was like, a kaleidoscope. But yeah.

Zena  

Can you share more? Are you able to share a little bit about what the book is about who the audience is? And then if you have, and I know, I’m gonna say this very tentatively, because I know, publishing dates can move and change, if you have like a vague idea when it’s going to be published.

Sara  

It will be out towards the later half of this year. And the concept is all of the ways that the universe could theoretically end the earth. So all of the different ways that the Earth couldn’t, like might not have formed, could not have survived in the very beginning. So it’s broken up into three sections. That’s what could have happened, what could still happen, and what absolutely will happen. So you’ve got what could have happened is Jupiter didn’t move in the early Solar System and the Earth never formed in the habitable zone, or our sun formed with a solar twin rather than alone, or the universe never formed, because it has no obligation. It didn’t have to form and yet, here we are. All the things that could still happen is like massive asteroid impacts. I mean, the dinosaurs, they can’t — because they’re gone — but they would tell us that that’s not a good time. Gamma ray bursts or black hole like a rogue black hole would theoretically just end it in any second. And then the things that will happen, so our merging — our merger with Andromeda Galaxy, our Sun is going to die. You know, all of the different things and all of the different things in between. We’ve even got a chapter on simulation theory and the fact that we could just be big simulation that someone could turn off. 

Zena  

That’s it — you sold it to me immediately.

Liz

You’ve got two customers here.

Zena

 A really funny thing about me is that I don’t actually believe any conspiracy theories, but I’m obsessed with reading them. But I will go down every rabbit hole. Have you read the ones about when Britney Spears shaved her head? And apparently it was like a distraction? Because the American government was like doing some underground stuff? Yeah. I don’t actually believe that’s true. Have I read everything about it and watched all the videos about it? Yes. Yeah. So literally all you have to do conspiracy theory. You’ve got me as a customer. I love it.

Liz  

Where’s your interest in in in writing and communication come from?

Sara  

I think I’ve always loved it. I’ve always loved communicating. So one of my jobs when I was in undergraduate was teaching high school students, they would come to the university I was at, we’d teach them — we’d do like a three hour workshop and teach them different things. And I loved doing that. And I realized that that was science communication, didn’t know what that was beforehand. I loved it. And I always loved the idea of like writing and producing videos, and I was always so nervous. I think I wrote and made my first video maybe when I was 19 or 20. And it was a five minute explainer about how the sun works. It took me like eight hours to film and to edit because I was so nervous. I hated everything I was saying. And I just was so self conscious.

Zena  

How jarring is it to see yourself and hear yourself on video. Yeah. Oh my god, is that what I look like? Is that what I sound like? Is that what my face looks like when I talk?

Sara  

Exactly, so that was– I did that and then just stopped for years because like I was just It’s too nervous to be self conscious, not confident enough. And then when I had started my PhD, so maybe, maybe seven years ago now, I still loved doing communication. And a friend had said, do what you want to do communication, like on TV and radio. And I said, yeah, maybe — like, I don’t know. And then it built up very quickly, from there of like, after the jar of seeing and hearing yourself and then just realizing that you can’t change it and it’s fine. Then the confidence came. And then I did all the things that I always wanted to do like writing, videoing, producing, but I didn’t have the confidence to do. So I think it was always there. I always loved communicating, but I was always too nervous. And it just took a few years to get out of that. 

And I’m very, very fortunate that I am at a university that finds the value in that as well and understands that even though traditional academia is all about our research, our peer reviewed publications, which of course, are very important, and I still do all of that and do all the conference presenting. But there’s also kind of the duty to the general public, because a lot of what we do is funded by the public. A lot of what we do is to try and educate the next generation who are going to kind of come up and do degrees or do do any type of vocational education with us. And I think I’m very, very, very fortunate that I have been supported in that way.

Zena  

That’s amazing. You know, I read an article the other day, Cathy Foley, our chief scientist was talking about the irony of misinformation. And she was saying how like, it’s so easy and free to access misinformation. But the academic content that you know, we work so hard and rigorously on is behind a paywall. So to actually be able to access accurate or what we assume is accurate and valuable information is behind this paywall, and then everything that is available, isn’t always great. So I feel like science communication for general audiences, particularly coming out of or not even science, communication, just communication of academic research, to general audiences coming out of universities, is so critical until we can move past that traditional publish or perish and paywall kind of approach that we’ve had for decades. 

Sara  

Yeah, and I think I think Australia from what I see is most of the Australian universities understand the value in that, and especially now that things are changing, and people are chronically online, and will receive, you know, a two minute or even a two second snippet of this as news. They clock it as factual and then that’s gone. And I think, I think that at least in an Australian perspective, it seems that we we tend to value the fact that okay, people are going to take a two minute snippet or a two second snippet, we might as well have an expert do that snippet, or try and make sure that we’re communicating things the best that we can, which I think is a really good thing for for our universities, and also for just the general public is that our media does value expert opinions. And we’re very lucky that we have media outlets, especially our ABC really values making sure that they get experts who are factual, correct, well spoken, so they’re not going to, you know, scare the audience with too much terminology. They’re going to speak to them like they’re a colleague or a peer. And I think that’s really, really important because it could so easily go the other way, it could easily be that having an academic or having an expert is too hard or it’s too difficult. It’s too hard to figure out who is doing what, and we’ll just have a general commentator for everything. And I think we see that in other countries and in other news places where then you just get misinformation spewed continuously. And I’m glad that that it seems to be very important here.

Liz  

I mean, one thing that is maybe worth mentioning or discussing here is like, it’s not actually easy to translate something that starts in a research paper. And, you know, for any kind of audience, you really have to think very carefully how to do that, and how to do that well and accurately. There’s like, there’s always that balance between like, Okay, well, I want to explain this thing, but I don’t want explain the ginormous amount of background information that you need. So what do I leave out, while still maintaining, you know, the integrity of what I am sharing. And that is a skill that takes a very long time to build up. It’s not, it’s not something that like, you know, you wake up with one morning, and you can just do well, right? And I mean, you can so maybe you can talk a bit about your experience, like I’m sure you’ve, you’ve got stories where you’d thought like, where you had to go through a process, figuring out how to explain, you know, something that you wanted to explain to a five year old that was not easy to explain.

Sara  

Yeah, it takes years and years. So I think I’ve been doing outreach and communication. Yeah, I guess almost seven years now. Which is a very, very long time. But it probably wasn’t until three years in that I started to feel confident that I could explain things, how I how I wanted to get across how I was expecting for different audiences. And I still struggle. Like I don’t think I don’t think you ever wake up not like you said wake up magically being able to eloquently explain everything in the universe ever. I still really struggle sometimes to pick my words wisely. And carefully, I guess, depending on what I’m doing. Yeah, I think it’s something that you’re always trying to build up. But I think it’s a skill that is super vital, just in life, but also in academia, because we’re communicating our research, yes to our peers, but also to funding bodies, to other stakeholders, to industry, and being able to communicate to those different levels is really vital in trying to get your research recognized as important. Because yes, we could, we could sell to our, you know, our colleague, our officemate, why our research is important in our specific niche, and they will, hopefully most likely be like, yes, absolutely. We understand. But trying to sell why that’s important to the general public and to other–to other places, is hard because I think ultimately, a lot of what drives research is being able to communicate effectively why it’s important and also communicating that back to the public. Because again, a lot of our research that is funded through say the ARC, it’s publicly funded — it’s research for Australia, this is research for our people. And we have a duty to be able to communicate why we’re doing that research, why it’s important. And I get this a lot. When we talk about space, when we’re talking about space industry, well, why why are we spending so much money going back to the moon? It’s like, Well, we’re not. But other countries are, and we’re just helping. But there’s good reasons why. And you can talk about all of the different things that spun off from an Apollo– and the medical, just the medical innovation from the Apollo program, like paid us back in dividends, we still use that stuff in our ICUs today. And I think trying to explain that to the general public is really important and something that I’m still, and I will for the rest of my life, learn how to do. That’s something that I feel really passionate about, because yeah, I think a lot of the time, you probably see it a lot in your own research and then just in general in the public as well. A lot of the time we’ll be like, Well, why are we spending money on this particular thing when we have a cost of living crisis, we have a housing crisis we have, you know, things aren’t solved in the world. And which is true and accurate. But a lot of the ways that we work towards figuring out solutions or technology that will help us combat bigger issues along the line is with research. 

Liz  

Definitely. Just for our listeners who don’t know what ARC is– Just for our listeners who don’t know what ARC is, ARC is the Australian Research Council. So it’s one of the funding bodies that supports fundamental research in Australia. There are others, but not many. A lot of the kind of fundamental research–so research that doesn’t have a direct and immediate application or, you know, direct benefit for society will most likely be put through the Australian Research Council as a proposal. Some of them get up, some of them don’t. That is one of the ways that Australia kind of supports this, what we call blue sky research, it’s sort of a longer term vision that we are creating. And there are all these knock on benefits for this kind of research. But it does mean that you actually have to tell the story about why this research is important; what kinds of benefits might come out of it. 

Sara  

And that’s part of the reason why I took the original position I did after my PhD. The professor I still work with today, Professor Christopher Fluke, is an astronomer by training like me. But his whole career has been built off really innovative ways of using techniques that we do in astronomy in other areas — so many different things that he works on because he can see the value and the difference in the crossover, which is what really drew me to that original project that I had taken the job with, and I’m lucky enough to still be working with him on different projects. But the cool thing is my day job now is not as an astrophysicist. It’s an interdisciplinary researcher. So working on –at the moment, I’m working on a project that’s through the digital health CRC, that my my supervisor is one of the lead investigators for and it’s AI in medical settings. So for clinicians, radiologists for a few different settings, and how does that help, but how do we know–how can we use what we’ve learned from astronomy with our massive test beds of AI to kind of help us along and get it right, which I think is really interesting and really lucky that I get to work on that.

Liz  

That’s really exciting. So where can our listeners find you.

Sara  

You can find me anywhere online pretty much at @sarawebbscience. You can Google just Dr. Sara Webb and my website will pop up with all the different links. And I’m always open to requests, so if you have a question about universal machine learning. You can always message me or put it in comments and I will nine times out of 10 end up making a video about your questions.

Liz  

Awesome. Thank you so much for joining us today, Sara. It’s been an amazing conversation and learn so much from you.

Sara 

Thank you so much for having me. It was my pleasure.

Leave a Reply

Your email address will not be published. Required fields are marked *