“Look, there was probably not a great deal in formal education. I’m sure I’m not the first person to have told you that. I wish there were more that related to navigating complexity and understanding ⁓ the kind of wicked problems that we face nowadays.”
“you shouldn’t make an assumption that what you see before you is exactly how it might be representing itself or how it might be represented to you. And that you should call on the expertise of epidemiologists and bioinformaticians and biostatisticians and other experts in the field to help explain that in a way that is understandable and digestible, but it’s not always straightforward. that…you do need to give it the space to be able to say, well, what is the story behind this ⁓ and how best can I understand some of the complexity behind the data.”
TRANSCRIPT
Linda McIver (00:00)
Welcome back to another episode of Make Me Data Literate. I’m excited this time to be able to bring you a guest that I pounced on at the ⁓ science misinformation symposium a few weeks ago. Couldn’t resist the opportunity to to ask him to be a guest. ⁓ so this is gonna be really fun conversation. ⁓ welcome Professor Sutton.
Brett Sutton (00:22)
Thanks, Linda. Please call me Brett.
Linda McIver (00:25)
Okay, I will. Thanks, Brett. I I think this is gonna be a super interesting conversation. So let’s start with who are you and what do you do?
Brett Sutton (00:35)
Sure, Brett Sutton. I’m currently Director of Health and Biosecurity, which is a research unit in CSIRO, Australia’s National Science Agency. So I lead a team of 350 or thereabouts, and they’re involved in research across health and biosecurity domains.
Linda McIver (00:59)
What does that look like day to day? So that’s a that’s a heck of a large term.
Brett Sutton (01:05)
It is a big team. yeah, look, there’s the usual administrative management. ⁓
responsibilities. So a lot of the day is taken up in that routine work, if you like. But obviously ⁓ we carve out appropriately ⁓ the bits where we’re either reflecting on our operational needs and refining elements there or reflecting on the strategic direction that we’re taking and thinking about where we make investments, where we dial things up or down accordingly. So, ⁓ you know,
It’s got the usual spectrum of ⁓ the routine and then the stewardship towards a strategic area of focus.
Linda McIver (01:52)
So what kind of things does health and biosecurity cover?
Brett Sutton (01:56)
Well, biosecurity is really the full spectrum, plant, environmental, agricultural, ⁓ animal and human biosecurity. So it’s all things in relation to keeping Australia secure from invasive species or invasive pests and what that means in terms of consequences for the environment, consequences for animal health, consequences for human health. So that’s, you know, that’s everything from ⁓
invasive viruses, invasive species that can threaten those sectors, but also the zoonotic diseases that are a risk to humans as well. there’s also a focus on those high consequence diseases that can impact human health.
Linda McIver (02:43)
So you folks should be on high alert with the bird flu situation.
Brett Sutton (02:49)
Yeah,
there’s a ⁓ separate research unit called AAHL, Australian Animal Health Laboratory is the name of the research unit. down at Geelong in the…
now named Australian Centre for Disease Preparedness and they’re obviously focused on doing all of the in-depth testing that’s required for bird flu to confirm the cases that are initially suspected or reported as we know from Western Australia and South Australia but anything else that will come in that requires that kind of high level expert confirmatory testing.
Linda McIver (03:27)
That’s ⁓ that’s I I don’t wanna get into that one because I’ve been looking at the news and being alarmed by it. But ⁓ good to know you folks are on the case.
Brett Sutton (03:38)
Yeah, and look, ⁓ it’s a big challenge for Australia and obviously threatens both wild bird and other bird industry populations. ⁓ But we’ve kind of seen it coming for some time and we knew that we were the last continent to be unaffected. And so there’s been a long lead time in terms of preparation and we’re now in ⁓ response mode.
Linda McIver (04:06)
That gives me some comfort. do you so you deal with ⁓ the threats even once they’re here? It’s not like you’re trying to keep them out of our borders but once they’re inside it’s out of your hands, like, you know, fire ants or things like that.
Brett Sutton (04:23)
No, no, ⁓ we certainly respond to things that ⁓ are post border incursions and problematic as ⁓ new invasive ⁓ species or pests. And so some of our technologies absolutely applied to the mitigation or control or, ⁓ you know, hopefully eradication in some instances of those invasive species. But it’s always pretty challenging once it’s gotten to ⁓ an established state and ⁓
⁓ We continue to try and apply the innovations and technologies that can keep it under wraps. these are complex things and obviously ⁓ movement across borders and the vast continent that Australia is means that sometimes controlling and mitigating the impacts is the best that we can do rather than elimination in full.
Linda McIver (05:20)
Yeah. That makes sense. so you’ve had a a really impressive career. ⁓ how much of what you learnt in your formal education has ⁓ well how much of what you need in your day to day job did you learn in your formal education? How much was on the job training?
Brett Sutton (05:44)
Look, there was probably not a great deal in formal education. I’m sure I’m not the first person to have told you that. I wish there were more that related to navigating complexity and understanding ⁓ the kind of wicked problems that we face nowadays. The world was simpler when I was in high school ⁓ in the 80s and ⁓ the problems that we face now.
Linda McIver (05:51)
No, not in at all.
Brett Sutton (06:10)
⁓ have multiple levels. They are system problems. They are ⁓ wicked in the true sense of the word in as much as ⁓ they don’t have single solutions and they can’t be overcome with ⁓ a single innovation or a single intervention. they require
all things together to be working and that goes from ⁓ the potential technological innovations that might apply to them but also to the policy regulatory legislative settings and then of course to the human behavioural elements as well which is as complex as all the rest put together. you know don’t think I even really scratched the surface of that in formal education that was very much on the job training and obviously my time as ⁓
Chief Health Officer and Chief Human Biosecurity Officer in Victoria was on the job training for crisis management, crisis leadership, crisis communication. And a lot of that gets applied in normal times as well in terms of managing that complexity.
Linda McIver (07:19)
Yeah, I I imagine that ⁓ the level of on the job training you got during twenty twenty and the subsequent years was unexpectedly intense.
Brett Sutton (07:32)
Yeah, it was. It was obviously in the crucible of crisis ⁓ and very challenging for everyone who was involved. ⁓ But there wouldn’t be anyone who didn’t say that they learned a great deal and got some.
refinement of skills and experience by virtue of going through it, you know, that does include ⁓ how to manage all of those things happening at once, ⁓ how to navigate the nuanced communication of those challenges and how to manage big teams, you know, the team that I was involved in grew to over 3000 and, you know, there’s a lot to… ⁓
There’s a lot to make sure that that’s focused in terms of its strategic priorities and its operational focus.
Linda McIver (08:26)
Yeah, next level challenge. ⁓ we saw during that time that ⁓ people’s understanding of for example, exponential growth is not what we might wish it to be or it wasn’t then. It I think it improved sharply, at the time. Is there any one thing that would make your life a lot easier that you think the world would be better if people understood just this one thing about data?
Brett Sutton (08:53)
I don’t know that there’s any one thing because ⁓ it isn’t one thing. I think maybe the understanding is that it can be all things. It can be your best friend. It can be your worst enemy. ⁓
Yeah, but maybe it’s that there’s a specific and nuanced and complex set of skills to navigate it, understand it, ⁓ transmit and communicate it that is required, that you can look at it.
and think that it’s straightforward, but there’ll always be a more nuanced story behind it. Sometimes it’s straightforward. Sometimes it’s just a case of very simple transmission of numbers that represents something very straightforward and very linear. ⁓ But at other times, ⁓ there are all of these other layers that you need to understand. I think, yeah, so maybe if there’s one thing, it’s that… ⁓you shouldn’t make an assumption that what you see before you is exactly how it might be representing itself or how it might be represented to you. And that you should call on the expertise of epidemiologists and bioinformaticians and biostatisticians and other experts in the field to help explain that in a way that is understandable and digestible, but it’s not always straightforward. that…you do need to give it the space to be able to say, well, what is the story behind this ⁓ and how best can I understand some of the complexity behind the data.
Linda McIver (10:38)
I love that. I did a workshop on Friday with some teachers on ⁓ measurement and I gave them a graph which was actually our the solar output from our solar system over three years and I said, Tell me the stories that this graph tells you but then also give me three other reasons why.
the graph might be s showing what it’s showing. Like the the idea that what we think we see in the graph isn’t necessarily what’s there. You know, you can say that one peak is higher than another, but you can’t say why from just from the the data itself, which is something we tend to forget. It’s easy to get caught up in a particular narrative, especially if it’s pushed really hard.
Brett Sutton (11:12)
Yeah.
Yeah, that’s right. Things get presented for particular reasons and everybody needs to take a step back and make sure that they’ve got either themselves the appropriate skills to interrogate and ask sensible questions. we should all expect that.
We’ve got the ability to call on either the expertise to be able to help navigate that or that we can have it ourselves, but we don’t get it like that overnight. And if we’re going to ⁓ have an appropriate understanding, we know that really complex data or data that is representative of really complex phenomena ⁓ will need that special expertise in order to navigate it.
Linda McIver (11:52)
Mm.
You must have some pretty strong feelings at this point about ⁓ armchair experts.
Brett Sutton (12:21)
I, you know, of course there are frustrating elements, of course there are ways in which ⁓ armchair experts may not have helped with understanding data at different times. At the same time, I’m deeply aware that people have a right to understand things that are affecting them personally and affecting us at scale and ⁓ have deep implications for how we live our lives and ⁓ the settings that we… ⁓
have to move through ⁓ in a policy sense. And they’ve got an entitlement to understand either. ⁓
what it all means or why it can’t be explained in a 30 second radio sound bite or an eight second TV ⁓ grab. Because, you know, oftentimes it’s the presentation of that data through those compressed and restricted opportunities that make it really difficult. think the, you know, the deepest frustrations I’ve had with data are probably around cherry picking. there was this one thing that was said about it without the broader context and it’s the
Linda McIver (13:28)
Yeah.
Brett Sutton (13:29)
broader context that means absolutely everything. And it’s not to say that ⁓ people are wrong in responding to what they get, it’s to say that ⁓ we should all do our best to create the time and space to be able to look at all of that complexity and nuance where it’s present.
Linda McIver (13:33)
Yeah.
And and to communicate it effectively.
Brett Sutton (13:52)
Absolutely, absolutely. And look, there are mistakes I’ve made there. There are mistakes that everyone makes at whatever points. ⁓
you know, it’s with the best of intentions that people try and navigate this stuff. ⁓ But it’s worth all of us reflecting on the fact that it’s not straightforward sometimes, that it has, you know, just layers and layers of, you know, all the biases that it can be subject to, all of the confounding, all of the errors in measurement and interpretation that can occur as well.
Linda McIver (14:07)
Hm.
Yeah. I I love that message that, you know, there’s there’s I always say there’s no such thing as a perfect data set, but there’s also no such thing as a perfect analysis or or one true analysis. There are, you know, so many different ways to to approach a data set and it’s really important that we don’t get too invested in ⁓ believing one track before we’ve thoroughly tested it.
Brett Sutton (14:54)
Indeed, although, you know, I’ve quoted this commentator, philosopher before H.L. Menchen, who said, you know, for every complex problem, there is a solution which is clear, simple and wrong. And, you know, he’s getting to the point that people can get quite attached to that simple narrative because they don’t have to deal with the complex reality that’s in front of them.
Linda McIver (15:16)
Yeah.
Brett Sutton (15:20)
But it’s wrong. And so, you know, take the time and leave the space for that complex answer with all of the caveats and all of the nuances and all of the differing interpretations that others might come to it with because that’s closer to the truth, quote unquote, than a straightforward, simple and wrong answer.
Linda McIver (15:22)
Yeah.
Yeah. I mean complex is is difficult and messy and a lot of work and i i the the simple is very attractive.
Brett Sutton (15:56)
Indeed, Wouldn’t we all wish the world thus?
Linda McIver (16:01)
Yes. Yeah, hundred percent. ⁓ are there things that that ⁓ that data shows or that you can see in data that you wish more people knew about?
Brett Sutton (16:14)
⁓ you know, I think there are are times when ⁓
data collection, data analysis, and all of the curation and collation that happens behind the scenes comes together in analysis, is deeply, ⁓ deeply… ⁓
illustrative, that it tells a really important new story. And I think it doesn’t happen frequently, but I think there are occasions when data can bring you to absolutely new and powerful insights, either by virtue of where data has been drawn from and brought together in a way that hasn’t happened before, ⁓ or it’s gone through the time and care and complexity to allow for an analysis that brings you to really
important new insights. Now that can happen in all kinds of areas across you know the full spectrum of social and health and medical and other sciences but it can be incredibly informative in those regards and you know I think I ⁓ recall a
insight into the harms of gambling ⁓ in Australia. There was a new analysis that had been done with data that was at a bigger scale with greater ⁓ data linkage than had happened previously across jurisdictions, across age groups and with socio-demographic data that hadn’t been linked ⁓ at that level before.
And it was illustrative of the deep harms that happened for a very substantial proportion of people who gamble. ⁓
Linda McIver (18:05)
Mm.
Brett Sutton (18:06)
And it allows you to set aside the simple narratives that you might otherwise hear of that it’s only the X percent of problem gamblers who are really impacted or doesn’t have these harms at a population level. It was very clear that it was deeply harmful to the problem gamblers, but it was pretty damn harmful for a substantial proportion of the population. And you can’t just set it aside with, again, the simple narrative that might come from
⁓ from those who would want to present it differently. The gaming sector in the US was well known for saying, we’re building a gaming center or a casino in your town, but guess what? We built three playgrounds and a kindergarten. Now that’s a simple and wrong narrative because that was built on the back of ⁓ a whole bunch of bankruptcies that were not visible.
Linda McIver (18:40)
Mm.
Yeah.
Brett Sutton (19:05)
these ⁓ real phenomena and consequences to life ⁓ is very powerful and very important.
Linda McIver (19:14)
Yeah, I saw a a d ⁓ snippet of Shaun Micallef’s gambling documentary recently where ⁓ he talked to someone who runs I can’t remember the town it was in, but someone who runs one of those ⁓
a bowling club it was, where they had ⁓ got rid of the pokies and they’d been you know, everyone told them they couldn’t do it and you have to have pokies and they subsidise all the rest of it and they like, I’m not sure that subsidizing people’s food on on the off the backs of you know, people gambling and losing all their money is actually the right thing to do. And they found that it turned out, you know, to have massive social benefits, not just for the club but for the the community around them. ⁓
It’s it’s it’s always nice to see those corporate ⁓ narratives overturned when they’re manipulative like that.
Brett Sutton (20:11)
Yeah, and
data shines a light on that stuff, doesn’t it? It gives the granularity and the depth of understanding of those ⁓ phenomena that you might otherwise not see and therefore dismiss.
Linda McIver (20:16)
Hm.
Yeah, that’s right. I interviewed David Spriggs from Info Exchange late last year and he was they have a platform where people can ⁓ search for services nearby, you know, whether they need food or housing or medical support or whatever it is they need, they can search for it on this platform. And the searches ⁓ plummeted during the pandemic, ⁓ during the lockdowns
when they raised the rate basically when they lifted ⁓ welfare payments above the poverty line. ⁓ and then ⁓ the week before they put it back, the ⁓ demand for services massively increased again. And it was just such a a neat and and impossible to argue with illustration of
what happens when you give people enough to live on. I it’s weird that you need evidence of that, but you know, the sometimes the evidence is really clear.
Brett Sutton (21:22)
Yeah, but it can
quantify the benefit. ⁓ It ⁓ can make real those positive impacts for sure.
Linda McIver (21:31)
Yeah, that’s right. Which is which can be really powerful. ⁓ not always powerful enough, but can be really powerful. ⁓ what are the worst data mistakes that you’ve seen?
Brett Sutton (21:44)
Look, I think it’s probably in the interpretation and it’s not, ⁓ again, it’s not a critique of ⁓ people who are trying to make sense of things that are ⁓ presented in an overly…
decontextualized way, but it’s where something is presented and the questions haven’t been asked and the presentation hasn’t been given of all of the weaknesses, caveats, methodological biases that are at play. And yeah, and it can literally provide an opposite outcome, an opposite ⁓ analysis than might have been presented. ⁓
Linda McIver (22:16)
The assumptions. Yeah.
Brett Sutton (22:30)
when that’s deliberate in order to support a narrative, that’s particularly ⁓ frustrating. ⁓ But I wouldn’t put that in the category of mistake. think mistakes are made when someone hasn’t gone to the, with the care and attention that is required to understand the complexity of certain. ⁓
Linda McIver (22:38)
Mm.
Brett Sutton (22:54)
data interpretations and analyses, you know, so whether it’s survivorship bias, whether it’s ⁓ the healthcare seeking bias or the healthy vaccinee bias or the, you know, the seeking testing bias, all of those things. ⁓
can remain hidden unless someone knows to look for them or knows to call them out as a potential bias at play. And I think they’re sometimes hugely problematic for ⁓ the results as they’re presented because they literally change what you should make of those results. And ⁓ that happens all too often. I mean, obviously, ⁓ we can be focused on ⁓
Linda McIver (23:31)
Fine.
Brett Sutton (23:44)
analyses that have gone through a robust peer review process and have had that you know methodological ⁓ robustness interrogated in full. Even then it can it can get through the cracks or it can still be open to all of those ⁓ all of those challenges but those really good studies they
they’re upfront about the limitations, they’re upfront about the fact that this methodology was applied because it was a practical and feasible consideration or that it was most appropriate in these circumstances. But there are other occasions where ⁓ it’s simply presented and… ⁓
and no one has asked the tough questions of that methodology or of the data. And that can literally be then grabbed by everyone who has that pre-existing desire to see that particular outcome presented and then amplified in ways that can get away from us.
Linda McIver (24:25)
Yeah. And
Yeah, and even where the methodology is ⁓ properly and and realistically presented, that doesn’t always translate into the headlines that that ⁓ result and the the stories that are written based on that, ⁓ you know.
Brett Sutton (24:59)
Yeah,
and science journalism is more important now than at any other time. ⁓ you know, there was probably a simpler world in which ⁓ journalism presented the science and it ⁓ was relatively more simple and more straightforward. We’re now in an era where, ⁓ you know, if it’s just a little bit… ⁓
tempting to distort for dramatic effect. And let’s face it, everyone’s a journalist now. And so it’ll come through in someone’s blog or be presented in someone’s podcast ⁓ and bless you for ⁓ talking to the nuance and complexity and all of these issues. there are those who would just present it straight out with the distortions because they’ve got great dramatic effect. ⁓
Linda McIver (25:33)
Yeah.
Mm.
Brett Sutton (25:59)
That’s a huge problem nowadays.
Linda McIver (26:01)
Yeah.
I saw ⁓ I think the Australia Institute did a takedown of ⁓
Pauline Hansen’s recent quote about the number number of the Australian population who don’t speak English, ⁓ or don’t speak it well. And it turned out she’d included babies ⁓ among among many other issues, she’d included babies. ⁓ but I liked the way you talked about you mentioned ⁓ seeking testing bias and it reminds me of the classic ⁓ genetics thing, genetics results that I learnt about when I was doing genetics, ⁓
back as my son says, when dinosaurs roamed the earth. ⁓ they they told us that ⁓ ten b when when genetic testing is done, ten percent of people turn out not to have the fathers they thought they had. ⁓ and that was before, you know, any anyone could just do a genetic test by s by getting a pack sent to them o online. But ⁓ it it completely
Brett Sutton (27:00)
Yeah, and
these, it’s the suspicious person bias, right?
Linda McIver (27:06)
Exactly, right. It’s like the the people who have doubts are the ones who get it tested and so obviously there’s a much higher you know, it’s it’s surprising in in in actual fact that, you know, of the people who have doubts, only ten percent of those are are ⁓ confirmed in those suspicions. ⁓ which is a you know, it’s the same data but a completely different presentation.
Brett Sutton (27:23)
Exactly.
Exactly.
Yeah, and these are the people who have blue eyes and both of their parents have brown eyes or they can roll their tongue like this and their parents can’t or some other inherited characteristic, know, I’m O negative, but my parents are both ⁓ rhesus positive. You know, these are the things that prompt individuals. so, yeah, it’s in that interpretation that you have to take care, absolutely.
Linda McIver (27:38)
Mm, mm. Yeah.
Yeah.
Yeah. And comes back to context, doesn’t it? You know, you you when you look at the raw data it doesn’t tell you the whole picture.
Brett Sutton (28:05)
Indeed.
Linda McIver (28:08)
we talked about data being deliberately misused. ⁓ what do you look for when when you’re looking at things in the media or, you know, they come across your desk at work? How do you spot the ones where you go, Mm, I’m not sure this is quite what we’re looking at?
Brett Sutton (28:26)
Yeah, look, I guess this ⁓ legacy media or mainstream media that is.
different from social media presentations. In mainstream media, would, you know, I’m working on the assumption that they are trying to present it ⁓ in a way that is factual and is appropriately representative, sometimes with a tendency towards dramatic effect, you know, not trying to make very deliberate distortions. And so in those circumstances, I think the appropriate questions to ask and the appropriate things to look for are what is missing here?
Linda McIver (28:50)
Ha ha.
Brett Sutton (29:03)
and sometimes it’s on an axis in terms of ⁓ you know giving a full representation of what those figures are sometimes it’s ⁓ literally looking for
Linda McIver (29:10)
Hm.
Brett Sutton (29:15)
where the source is, what period of time it relates to, what population it relates to, what the methodology might be behind this. And oftentimes that’s available, or you can go to a link that allows you to see what the source data is and what the fuller picture is. ⁓
But yeah, it’s really asking the question around what is not being presented here and what can I not see in this because the graph is one thing, it’s telling a story, but it is one little chunk of a much bigger, broader and more nuanced context. ⁓ In the world of social media, I would say question everything. I mean, there is obviously… ⁓
a need for a civic uplift in our understanding about how to interrogate all of that, but it does relate to what is the source and ⁓ how might I trust or not trust that source. Don’t just take it.
⁓ as valuable because it’s been shared by a member of your own family or your circle of friends or circle of Facebook mates or whatever. ⁓ But there are also all of those, all of those danger signs, you know, it’s come to me urgent, caps, all caps ⁓ and ⁓ someone speaking in a video and they’re all very emotional and agitated. Now they’re, they’re ⁓
Linda McIver (30:26)
Yeah.
Brett Sutton (30:45)
not inappropriate at times, but you have to ask the question, why is this being presented in some massively dramatic way? ⁓ Ask yourself the question, is that the emotional hook that someone’s trying to reel me in on in order to engage with this? ⁓
The story of social media nowadays is that it’s all about the attention economy and in fact it’s about the attraction economy to hook you in and keep you there. ⁓ Short video feeds are the primary mechanism of engagement with our youth now and I think it’s a pretty depressing ⁓
Linda McIver (31:19)
Yeah.
Brett Sutton (31:36)
landscape in terms of information sharing because ⁓ you know telling a story which is a little bit more boring because it’s longer and needs detail and has no great dramatic you know drum roll conclusion
Linda McIver (31:51)
Ha ha ha.
Yeah.
Brett Sutton (31:56)
isn’t going to get a 15 year old on TikTok. And so that’s something to challenge now. But I think it’s an enormous challenge to overcome because it’s literally built into algorithms and the monetization of that data and the monetization of that engagement on those platforms. And so they’re not really interested in the long, nuanced and somewhat duller story. That’s the reality.
Linda McIver (32:25)
We don’t quite know, isn’t great clickbait.
Brett Sutton (32:29)
Correct. Correct.
Linda McIver (32:31)
The big secret no one’s telling you about this. The secret is we don’t understand it.
Brett Sutton (32:34)
you
That’s the dangerous sign to look for. They’re trying to hide it from you.
Linda McIver (32:39)
Ha ha
I I realized when I was preparing for the last episode of this podcast where I interviewed Ketan Joshi about ⁓ AI data centers and stuff, I realized that th those red flags you’re talking about, the urgency and the all caps and that, you know, you’ve got to get on board, you don’t want to miss out. That’s that those are scam red flags and it’s absolutely the way the AI industry and the data centers are being marketed to us. I was interested to hear you say that.
you assumed when you were dealing with legacy media that they were trying to tell a a ⁓ factual story. That’s not where I thought you were going. Maybe that’s because I’m overly cynical a little little bit paranoid these days. But do you not feel that ⁓ try not to put you terribly on the spot, but do you not feel that mainstream media suffers from the same kind of pressures the need to get
clicks and and and also, you know, often a an agenda to push.
Brett Sutton (33:46)
Yeah, there’s no question they’re feeling that pressure and it’s a commercial pressure. ⁓ And so they, of course, they need engagement ⁓ and they’re competing with platforms that ⁓ are driving that engagement, you know, kind of without moral or ethical constraints in many respects. That said, there are voluntary codes of practice that apply and there are ways in which they ⁓
Linda McIver (33:50)
Mm.
Brett Sutton (34:15)
constrain themselves ⁓ as with industry standards that put them in a different category, I think. That is not to say that they’re not being challenged and it’s not to say that bad practices don’t come to the fore. But I think there is a little bit of, ⁓ you know, there is a little bit of hope in that they are.
Linda McIver (34:28)
Mm.
Brett Sutton (34:36)
also wanting to differentiate themselves from that social media pig slop as I like to call it ⁓ because they want to be able to say ⁓
this information is coming to you in a different way for a different reason. It’s coming to you because we are ⁓ wanting to differentiate ourselves in terms of the way that we ⁓ apply some rigor into the storytelling that we do, the information and fact checking that we do. ⁓ And so, you know, it’s not universal. It doesn’t apply to everything that you ever see. ⁓
Linda McIver (35:14)
Mm.
Brett Sutton (35:15)
There’s no question that there are commercial and financial pressures that apply to ⁓ mainstream media. But I do think there will be and there is also pushback on what people are exposed to through social media. And they are saying, you know, it might be my main form of engagement. It might be ⁓ three hours of my screen time a day compared to 10 minutes of watching ABC News. ⁓
Linda McIver (35:42)
Mm.
Brett Sutton (35:44)
but they are also ⁓ appropriately sceptical of that feed. so, ⁓ you know, hopefully there’s a little bit of a swing back of the pendulum to say, I just need some assurance about what is real in the world and what I can trust in terms of information sources. And so people do differentiate. People say, you know, this might be appealing. Yes, I spent some time looking at these feeds, but I also don’t trust them as much as I would trust our national broadcaster or
those who would apply a voluntary code of conduct.
Linda McIver (36:14)
Mm.
Yeah. I have I have concerns about voluntary codes of conduct, but perhaps there are ⁓ more consequences, more accountability in the mainstream media. A a little more, not as much perhaps as we might like, but
Brett Sutton (36:36)
Yeah, and look, obviously the ABC applies the charter and takes it very seriously and has independent review processes. you know, they’re not all the same in the mainstream media space either. ⁓ people need to think about that in terms of where they’re getting their information from.
Linda McIver (36:40)
Mm.
Hm, that’s true.
I feel like we could sort the not fix the social media problem, but it could be a lot less problematic if we simply legislated the algorithm away. ⁓ you know, just make it chronological, it’s still an algorithm, but you know, legislate the engagement metrics away and just go you can’t do that anymore. ⁓ it
Brett Sutton (37:22)
It is a tricky space and yeah,
we’re dealing with, ⁓ you know, companies that have ⁓ revenue that exceeds Australia’s GDP. ⁓ it’s not an easy ⁓ or symmetrical power play.
Linda McIver (37:32)
Mm. No.
No. I I think that’s a really important point that we tend to forget that that ⁓ asymmetrical. The the ⁓ the power is ⁓ largely on one side and that’s really problematic. So we wind up legislating people instead of legislating platforms. And say, you know
Brett Sutton (37:54)
Yeah, indeed. And
it’s tricky. it applies to us at an individual level as well. We would all be better off if we had a deeper understanding of what we are giving over when we get on platforms in terms of the data that we are sharing, willingly or unwillingly or knowingly or unknowingly. And I think ⁓ that’s also problematic, that all of that data is used and it’s used for commercial gain against us, if you like.
Linda McIver (38:08)
Hmm. Yep.
Yeah. Yeah. Definitely. Do you have th I didn’t warn you about this question, so you can ignore it if you want to, but do you do you have any positive uses of social media yourself? Are there things you like it for? I like it for cat videos and otters and things like that.
Brett Sutton (38:39)
Yeah, look,
for me nowadays it’s mostly entertainment or humour. don’t tend to go to it for trusted sources of information or factual interrogation because it’s largely, it might be there, but you’ve got to sort through… ⁓
Linda McIver (38:46)
Mm.
Mm.
Brett Sutton (39:02)
You’ve got to sort through tons of chaff to get to the wheat and that’s really tricky. So ⁓ by and large, I won’t go there for facts, ⁓ but ⁓ yeah, it’s good for the jokes.
Linda McIver (39:06)
Yeah.
That was that ⁓ your point about having to sort through tons of chaff reminds me of ⁓ something that we learnt at the at the misinformation symposium, which is that repetition of a bit of misinformation makes it more plausible even to people who are disinclined to believe it. So if you r repeat climate misinformation even to someone who is on board with climate action and, you know, believes in
climate change and understands the issues and stuff, just the repetit sheer repetition can have a powerful impact on your ⁓ willingness to believe something. And that that’s a little terrifying.
Brett Sutton (39:55)
It is, but that’s how it’s working, that volume wins ⁓ over quality and integrity. maybe that goes to some deep evolutionary element of what it means to be human, that you want that sense of belonging in tribe or in social group. And if you see something repeated again and again, ⁓
it becomes more more challenging to be the person on the outer. And so either unconsciously or consciously, you’ll have a sense that you should go with the flow. But that comes with its own dangers, obviously.
Linda McIver (40:22)
Yeah.
Yeah. Yeah. I find that with AI I push I I a lot of my work is pushing back against the the AI hype and trying to highlight what it can and can’t do. There was a an article I saw today, a study on ⁓ on the use of AI in health and the the ⁓ number of errors number of ⁓ potentially catastrophic errors that that were made which
lined up with with what I’ve been saying, but there’s so much hype and so many people going, It’s the best thing, it’s so great, it’s amazing and you start to go, Am I nuts? Like I have I just missed it? Like do I not do I not get it? And then you see some other error and you’re like, no, no, I’m But the repetition really it is weirdly powerful.
Brett Sutton (41:19)
Yeah, but…
Yeah, AI talks with the confidence of a middle-aged man. And so sounds very convincing because there’s no doubt. It’s like, yes, this is how it is. And just last week, I’m trying to dig down to where my grandmother’s middle name came from. And I couldn’t work out whether it was her.
Linda McIver (41:31)
Yes.
Mm.
Mm.
Brett Sutton (41:50)
mother’s maiden name, which it isn’t her grandmother’s maiden name or her great-grandmother’s maiden name. And I was trying to work out whether it might have been an association with a well-known family in Bendigo when my great-grandfather came over with the gold rush from Cornwall, or whether it was actually on the maternal grandparent side.
And I asked AI ⁓ one way and got, yes, it’s related to association with this famous family in Bendigo. And then I asked it a different way and it said, no, no, it absolutely isn’t. It relates to the maternal lineage. And this is why. And so both confidently presented and sounds very convincing. it’s grain of salt stuff. You’ve got to take it with a grain of salt.
Linda McIver (42:26)
Ha ha ha ha.
Yeah.
Yeah, yeah. And that’s we’re not really well set up to do that psychologically. You know, it’s not we’re not it’s not the default. The default ⁓ we’re we’re quite accepting and and believing people, sort of at bottom, I think. You have to really work to build that scepticism.
Brett Sutton (42:48)
No.
Indeed, it’s AI getting ⁓ to that level where it’s very hard to differentiate from someone very familiar to us. Sounds like normal human conversation. And of course, we’re going to anthropomorphize to the extent that we think there’s ⁓ sentience and emotion and connection behind all of that.
Linda McIver (43:21)
Yeah.
Yep. Yep, definitely. ⁓ this this has been a a wonderful conversation. Thank you so much for your willingness to be involved. We come to our last question, which is my favorite. What excites you about data?
Brett Sutton (43:42)
pleasure.
I think I’ve said it already probably, but it’s the ability to reveal new things and through that telling of a story, bring us to new insights that hopefully mean something to us in the world that is…
is a revelation about us as human beings, is a revelation about society, about the fascinating world we live in. There is so much ⁓ out there that is yet to be discovered, ⁓ but there is also a whole bunch of stuff that is yet to be fully understood. And it’s ⁓ in data that gets us to new understandings that can be fascinating and exciting and revelatory. And I think… ⁓
you know, I think that’s the magic of it all.
Linda McIver (44:44)
I love that. That’s a beautiful note to end on. Thank you so much for your time today. This has been wonderful.
Brett Sutton (44:50)
Pleasure. Thanks, Linda.
