Marnie Hughes-Warrington, "Artificial Historians" (Routledge, 2025)

0:00This episode is brought to you by Facebook. So you were scrolling on Marketplace. And there it was. The bike you've been searching for. You sent a message, and it turned out the seller was super chatty, kind of funny, and an avid cyclist. The next thing you know, you're in a cycling crew. Well, a community cycling group. The thing about Facebook, you might find more than what you're looking for. From a browse to a bike ride, this summer find more on Facebook. This podcast is supported by Anthropic, the public benefit corporation behind Claude. Everyone's got a hard question about AI. Your job, your kids, where it's going. Anthropic was built to surface those questions and share what it finds along the way, because there's hope in hard questions. Ask yours at Claude.ai slash Spotify and keep thinking. That's Claude.ai slash Spotify.

1:01Collast it up with Crocs. You know back to school is coming in fast. So why wait to find your new faith footwear? Step into a local Crocs store and step into your new look. Try it. Style it. Make it yours. Because the right pair doesn't just show up. It shows off. First day fits. Handled. Walk out ready for whatever's next. Visit your nearest Crocs store today. Welcome to the new books network. My guest is Professor Marnie Hughes-Warrington. She's Bradley Distinguished Professor of History and Philosophy at Adelaide University in Australia. Professor Hughes-Warrington studied at the University of Plasmania and Oxford. She's a former Rhodes scholar. She published widely on theory and philosophy of history, exploring the multiple and often surprising connections between these disciplines and their place in the overarching search for truth across different cultures and societies. And today we're talking about her most recent book with contributions from Ann Martin and Louis Yarla Purgat-Brien.

2:07Title artificial historians published this year in 2026 by Rutledge. The book is available in digital open access format. Marnie, welcome to practical history and thank you so much for joining us today. Oh Patrick, it's actually a real honor and a pleasure to be here. Thank you. Because you're a historian and philosopher. I thought maybe one way of starting this conversation was to ask, where do you think? If you like one end, the other one starts and vice versa. It's a really interesting relationship. You know, you're working with one discipline philosophy that's searching for kind of perennial truths or really groundings or truths for people. And the other one is all about change and contingency. So it's a really beautiful and uncomfortable relationship between the two. And various theorists have written about how they might or might not relate to one another. But for me, it's been decades of fun and it's no sign of an end. It's a process it sounds like and I saw that all of your works actually go from one to the other and sort of intersect and play back and forth with each other, which is fascinating and great exercise for the mind.

3:12I'd like to start with an important sentence from your book, Artificial Historians. It's from page 167 and goes like this. The logic of history is the thread which ties AI, Artificial Historians, and Artificial Histories back to where we stand and to a present and the future where we can make histories for good visibly. I thought about this for a long time and it's kind of like a thread that goes through your book. And in line with this thought, your books makes the case that AI and history are intimately related, even condependent, I would say. And it makes the case that it's important for us to understand that complex relationship so that we can build better AI. And furthermore, it's important for us to understand that logic and that it's potential for AI so that we don't hand over too much power to AI or bad AI. I'm using shortcuts here in a lot of ways. Would you say this is the first, like a fair first take on your book and maybe we can use this as a starting point.

4:14So maybe you can correct me, but at least we can start. I think it's actually spot on because I think my primary argument is we perhaps have a wider and a slightly different view of AI that's available for all of us. If we think about the way that contemporary AI works, it processes and makes sense of a lot of our data from around the world. So in a way, it's taking data from the past and that past could be very distant or it could be very, very recent. And it's processing that data and making sense of it. Well, that's a process of history making. The interesting thing is that there's no human involved in that, typically. And professional historians hadn't been involved in that. But if you wanted to do that, you know, better to draw on the expertise of historians who've been doing this for a long time and have great expertise around how to work with data that's, you know, incomplete. Creditual, problematic in lots of ways. Or you want to get more efficient, you know, processing of data so that you don't need as much of it and you don't need to build as many data centers, for instance.

5:15Then you might, you know, it might not hurt you to have a conversation with a discipline that has what I call a historical intersection with you. And that future and that present is actually the focus of the book that I'm just beginning to work on now. So you put your finger right on there on that factory. And you anticipated my last question too with your future projects, but I think maybe we'll come back to this. Yeah, we can come back to that. You mentioned something interesting working with incomplete data. This is what historians do all the time. And of course, data is always incomplete and I imagine will always be incomplete for AI too, correct? Yeah, which is why there's so much at stake. Right. There's just some great thinking out there that says if you've got complete data, what's the point of things? And, you know, is that a world that we want to have? You know, I've just I've just reread the circle. It's a really interesting book, right? From 2013, this notion of a continuous recording of your life recording every aspect of it. Now it's not true that everything's been captured. But it's a really interesting space for people to think about not just technically, but ethically.

6:18Is this what we want? And, you know, we're all familiar with students in a history class saying, oh, the data's biased, the evidence bias. And they go, yeah, okay. So what are we going to do about it? Saying that it's biased is just kind of question one of the class. Right. Doing something about it is the second and the whole point of studying history and philosophy is to think, well, what can we do about that? Because we will never have a complete data record. Right. And it's not what can I do with it, but what should I do with it? How should I think about this? And I think historians are really well placed to take us in the direction of thinking about those questions. So it makes sense that you call your book artificial historians, but it's also at the same time in an intriguing title and a little provocative title. I think you mean a couple of things in by saying this by using this name. You talk about actual histories written with some difficulty as you point out as you make the case that's very compelling by generative AI. And you also talk about the big data that companies collect about us.

7:19And then they use to predict behaviors and suggest all kinds of actions for us to take. Right. Could you talk more about how you chose the title? Yeah, it's a really interesting story, Patrick. So I started working on the book really late 2018, 2019. And I said to everyone, I really want to look at the relationship between history and AI. Everyone said, oh, you're writing a history of AI. And I actually had a bit of difficulty explaining to people that I was interested in the idea that machines might make histories. And that humans might find themselves actually not the only historians on the planet. And in 2019, a lot of people just said to me, I don't know what you're talking about. And they were struggling a bit with what I was, you know, some people really got it. 2020 hits and generative AI, I mean, there have been papers on generative AI. As soon as generative AI hit people's consciousness, they went, oh my goodness. So I really did want to be slightly provocative in the title and say, and I wrote a paper around 2022 saying, you know, is this time that we start to recognize that machines actually might be writing histories and producing them.

8:29And the ethical message of that for me is every one of us that's connected to digital platforms is in a way agreeing to give a perpetual license to our data or store our data. And I think we all have the right to know what our histories are and to have some role in playing making those histories. And so I thought I'm going to be a bit more provocative in the title because I want to drive home to everyone that that kind of bargain that we've all driven between ourselves and various companies might come home to roost when we don't know what histories are written about us. What happens with our information and that's not all AI and the books not only about generative AI. There are of course little language models that people build in research teams. So AI is a big, bulky, interesting collection of technologies and I've tried to look at different ones in the book. So it's not all gen AI and it's not all big corporate AI that's out there. But what I really want to drive home to people is who do you want making your history?

9:29What histories are made about you and what are the consequences of those histories? Do they mean that you get a financial line that you get seen in the emergency department of a hospital? Really driving home to people the centrality of history in the making of AI and the decisions about all of us every single day. Like what's the best route to drive my car or to walk? That seems pretty innocuous. But when it comes to people to say will I get a spot in university or a school? Am I actually going to get interviewed for a job? Will I be seen in an emergency department? promptly will I get health insurance? Then it really comes down to well what's your data history? How in a way has that shaped that conversation and do you know the ways in which those decisions are being made and are they fair for you and for everybody else? So what I'm really asking people to think about is this is the time to get engaged with history and in a way the history curriculum. In particular a historical research can respond to these questions and make people aware you want to make a fairer world.

10:31You're going to have to think about history more broadly and you're going to have to invite historians to the table. One thing that fascinates me as I'm going through your thought process is is the role of intention because we're so used to thinking about history as an intentional process. It should be that way, arguably. And yet we have those histories where we become kind of an object more of a subject. And so yeah it seems that we take some of the agency back and start to think about how these things are being used. Right. I quite like Aristotle and you'll notice that in a few and in fact my goal is to write four books where I'm riffing off Aristotle. Some people will notice that the wonder book can begin little histories and artificial historians are Aristotle's in there. And I've said you know I originally set out to write a kind of revisitation of Aristotle in the book on wonder. And as I went on I thought oh okay I'm going to revisit his metaphysics I'm going to revisit his ethics I've revisited his logic.

11:32Well you can guess what I've missed is his epistemology. So if you're predicting what book I was going to write next that's the one that's coming. And Aristotle says you can't get other people to do your ethics for you. The whole point of ethics is it's hard work. And you're going to have to put the effort in right you are going to have to put the effort in and that's not just about your agency. So I was very honored to give the Couseric Collection Bill Felt University this year and actually talked about not just individual relationships each of us has with AI but actually the beginnings of what we could see to be cyber susceptibility and even cyber history wars and cyber cyber warfare around history. So you know the more data that you've got about what people are interested in in history the more you've got some leverage to produce really good histories because you've got market data and you can say you know for instance everybody loves World War 2 let's write another book about World War 2 and we'll get a best seller. You can do it that way or you can be really malicious about that and you don't need to intend to do that in fact algorithms could coordinate in the generation of histories.

12:37Humans might not actually be directing the show there and you could get what we would recognize as malicious or collusion or outcomes and no human being has sat down there and instructed those in those algorithms what to do they've just taken market data and they've decided we're just going to do this or this is going to get a lot of attention or this is going to do they've optimized the information that's out there. And so where I'm working on it the moment is not just that future between history and AI but you know what does cyber what a cyber history wars look like and what does a history market look like at this point and these are really interested questions to explore. And I feel like there's a larger context beyond AI for this which is why I kind of come back to this question all the time about what history so I was wondering if we could pause for a second and talk about what history is and is not. And I think it's interesting to think about this for so many reasons one like I regularly run into a computer science person that I started talking to more and more to because of AI and when they find out I'm a historian say well you see in a couple years all histories will be written by AI because AI can take the best folks by the best minds from all over the world and it's just going to give you the optimal history and I go out and struggle with and I think about it and I think about responses and I okay.

13:57Second 22 in one survey by American Historical Association reported the two thirds of Americans think that history mostly about name and dates and only 17% people who view history as an explanation of experiences in the past which is exactly the opposite of how historians think of themselves as mainly explanations of of of experiences in the past right. And then we're seeing these clashes over the correct version of history you talk about history wars I mean in America for sure other places in the world so it seems that it all shows that history is so important but it's also so vulnerable and so misunderstood which which I think you know just showcases how important your book is. Well it showcases I think that public history data that you've referred to and the opportunity we have as professional historians to really work our way through what the implications of that information is you know historians have over a long time build up some really sophisticated ways of demonstrating their logic so using some really subtle language like the if and the button the maybe of the historian you know technically that's called trivalent logic well the general public's probably not got much appetite.

15:12For a lot of that really technical language you know there might be more appetite for certainty out there but if you look at that public survey data I don't think anybody out there saying there's only going to be one history in the world and if you ask people do you think we'll ever come down to a world where there's just one history most people go oh no no that's not going to happen okay so why isn't that going to happen well because my family story might not be you know relating that history it might not be enjoyable to me or. Actually my nation has a different view of that event to this other nation so then you begin to take people on the journey of saying okay we can agree probably there's not going to be one history there's going to be more than one and those histories are not going to have comfortable relationships with one another we may disagree with one another and some people might read write histories that you don't necessarily agree with that are reasonable and rational histories and then we haven't even started talking about the ones that are malicious that are actually just promulgating hatred or harm. And I'd love to say that that's a new problem but it's not people have been making histories for a long time in order to basically eradicate or to express hatred or to foment violence towards other people and always say to people of histories part of the problem it's part of the solution.

16:27So one of the opportunities we have is to think about the clever ways in which historians have expressed that trivalent logic the point that not everything is certain because we might discover more evidence or you can read the evidence really responsibly in more than one way all the evidence is incomplete you can do that but how do you communicate that to your audience in a way that they say okay get that I really get that and you know obviously filmmakers have been doing this interesting ways for while I've written on film I find it interesting how filmmakers communicate you know I don't know this bit I've just watched the Odyssey over the weekend and you can see Emily Wilson's translation. Emily Wilson's translation in there it's fantastic but there's a bit in there around the treatment of the women for instance where you go okay Chris Nolan's clearly you know not entirely agreed with Emily Wilson there I can see some modern sensibility about that and a lot of film audiences go oh my gosh that's really modern or I saw the Roman soldier wearing a wristwatch and you go yeah that's point of the point of a film is that it really engages you in that debate and people get really passionate about the inaccuracies of films and I love that but they get less passionate about that.

17:35They get less passionate about that about professional histories and so the invitation is for all of us to say are there new and different ways of communicating that subtlety that responsibility and also push back on people say there's only going to be one version of this because one version of this is not only problematic for most people you end up in some really susceptible positions there whereby you know algorithms could take advantage of those susceptibilities and actually really manipulate the message in some ways that you don't like. They're quite dangerous so yeah so do people get ready and and you might not like it those surveys also show half the people half the population want the history to be really positive and the other half want it to be really negative so I say people good luck getting that one. We're going to get 100% agreement but if you ever sit by and do yourself in a situation where there isn't a debate about history that's a pretty unhealthy place to be that's an unhealthy place. Let's talk some more about the logic of history and maybe unpack some of the motifs from from from your book while we're doing this.

18:40You talk about history made walking versus history made sitting and I love this distinction and I'm going to read a sentence from from your book here. Could the corner and the UN people's these are names of two Australian Aboriginal nations never sit down or more precisely sit sit down all the time. They have told story history live in the moment for you and for me for tens of thousands of years they've spoken it painted it drawn it woven it and sung it they've also written it down since the mid 19th century. The right history but they also recognize history and everything they do and they recognize its power and potential for the present and for the future. They shift from sitting to standing and then walking conscious that all these stances provide opportunities for listening and then acting. I think it's a beautiful way of thinking about history and throughout the book you make that case that what makes history is this this logic characterized by open-endedness and effort to uncover this world of possibilities a world in which there are few absolutes and little finished business which is actually an expression you also use and I think it's great.

19:45And then you conclude on 164 that history is not perfectly or fully logically true or a story again it sounds cruel to put it this way but it is technical right there's no statement about the past can be accepted as true without referring to other things like country footnotes archives participants objects and so on potentially add in fun item. And I think this is very interesting and I think that all of us historians would agree with you but the novelty here is that you actually analyze formally the language and the logic of historical thinking writing talk about tribal logic here you showed just how much the language of truthful historical writing depends on these uncertainties not just the words and wise but the ifs they talk about the maybe. Could you maybe develop this this idea a little bit more. Yeah I got really excited about this I do want to acknowledge I am on Ghana country you know and I'm on in the money money so I'm saying hello to everybody in beautiful Ghana language I'm really privileged to live in Australia and to be surrounded by very wise elders and some people have always said to me what's interesting you've got the endings of that book and the previous book are co-written with an Aboriginal co author what are you doing there and I go if you ask the question what are you doing there then you haven't listened closely enough to Aboriginal elders.

21:02So that to people they will often put these things out for you to get you to think about you know so I you know early in the book I found this beautiful piece by Chinese historian who he says at one point you know historians used to wander around and they'd interview people and they read books and they'd go and talk to witnesses and you know that's a lot of some cities to a lesser extent but they would go and wander around and he has this lovely analogy at some point historians sat down and they started reading other people's eyewitness accounts in books. And listening to them and then they wrote their histories there so they sat down literally in monasteries and they started transcribing histories and putting them together and we you know had a different relationship to the information different form of mediation at that point and you know there was this notion that history becomes the book and I'll say to people the idea of history is the book is really interesting because you might not know at all who the author is and do you even know that the author is human. So you might have noticed that Amazon and Spotify for instance and now putting limits on the number of book publishing outputs that are uploaded per day to Amazon because there's a lot of synthetic material is being produced by AI producing exactly those books we were talking about before the people really interesting these particular topics so we don't even know whether some of these books are actually written by humans anymore and that's the thing about books and for the garlic people they said well you don't depend on one way of thinking about it.

22:29So you're telling the story you are going to have to think about the various different powerful ways of communicating about past that includes walking on country and learning from places very powerful painting singing dancing and the truth is that we have all of these forms of expression for history and occasionally for instance I remember Hayden White said we need to have a thing called history of photo talking about film because history of graphy is writing writing history. No no no no no no no no no yeah graph is about writing but for Aboriginal people for instance about seven hours north of here in the beautiful API lands they draw pictures in the sand they call them walker and they call them meaningful marks as a writing of history it's still historiography but it's not a book form that we're familiar with so one of the opportunities we have is to say history has never been about one way of telling there's all these different cousins and family members of the world. So the history film which the public perhaps have stronger affinity to in some ways museums you know and that's something the other public data tells us is that people interested in museums they love watching documentaries and they may trust those things more than they trust a historian strangely enough and there actually aren't enough professional historians out there in the world and unfortunately smaller and smaller as universities restructure they decide the humanities is not as powerful to them which is interesting because some parts of the Asia Pacific is not a big deal.

23:58So we're going to go oh crikey we cut the humanities back too much we now need them for AI because they're exploring exactly the things that we absolutely need. So taking this people on a conversation to say the forms of mediation and the forms of expression for history is it's more than one way of doing. And we may view the book by the professional historian is having more credence but when you start looking in closely at the way that historians write histories they're actually really magnificent complex vehicles. where you see them switching chronologically, flicking back it and forward, using perhaps a maybe to signal that responsibility, but in some cases to prevent a fight with another historian. That was fun thing I found really interesting is how people use that couching language to really polite to their colleagues as well as signal responsibility. Absolutely, where do they ask questions? Mostly at the end of the books rather than start of their books, because they want to kind of bed down the topic, but they want to open up some questions at the end

24:58for the reader to say, OK, you're going to have to shoulder some of the responsibility here. You're going to have to pick this up. And for me, that indicates that perhaps the revival of an interest in the logic of history is well over too. Good old Avisia Tucker, he's been writing on logic and history for a while, but I would love to see more writing on logic and films exploring the ways in which people convince us. Show us with credence, I say, it's not absolutely perfect. But nonetheless has authority and credence because of the ways in which people build up the story and use logic to convince us that actually this is really plausible. Not only plausible, this has a lot of authority. And I actually believe this is really, really amazing work. Versus, that's just pretty dodgy. Oh, don't write, write, write. I don't like that. Yeah, I don't think that's thorough. And if we're going to be able to think about that conversation around how does AI make history well, that's not going to come down to is Gemini human

26:00or is flawed a human, they're not. And I say to people, I don't view them as humans. They're not agents with intentions. Some people like to think of them that way, I don't. But I can see the differences in the way that they use logic to the way that humans use logic and that gap between what professional historians do and what the public actually thinks the good history looks like. This is the frontier of where the disputes and where the exciting stuff is going to be and the next idea is in my view. Super exciting. This episode is brought to you by Facebook. So you were scrolling on marketplace. And there it was. The bike you've been searching for, you send a message and it turned out the seller was super chatty, kind of funny and an avid cyclist. The next thing you know, you're in a cycling crew well, a community cycling group. The thing about Facebook, you might find more than what you're looking for. From a browse to a bike ride, this summer find more on Facebook. This podcast is supported by Anthropic,

27:03the public benefit corporation behind ClawD. Everyone's got a hard question about AI. Your job, your kids, where it's going. Anthropic was built to surface those questions and share what it finds along the way because there's hope in hard questions. Ask yours at clawd.ai slash Spotify and keep thinking. That's clawd.ai slash Spotify. Collast it up with Crocs. You know back to school is coming in fast. So why wait to find your new faith footwear? Step into a local Croc store and step into your new look. Try it. Style it. Make it yours because the right pair doesn't just show up. It shows off. First day fits handled. Walk out ready for whatever's next. Visit your nearest Crocs store today. And it also makes me think, you know, after everything we've been talking about the last five minutes in terms of those models

28:04and the ifs and the maybes and the, even like the history, professional history is being complex artifices in some way. Is there sort of couching the uncertainty in a certain language? One would, you know, it seems that that history is basically relational, maybe relative. I'm not sure which word is better, but it's, but it's important to say that it's not completely subjective in this extreme sense of it's all text and you mentioned hide and white. So ring me think about this too. It's all text. It's all narrative because it's not. It's not all opinion, for instance, either, correct. Right. And hide and always said that. He actually said, you know, the content of the form is really important. Why the history is written. But he too thought that these decisions about narrative came from ethics. He always believed that it was grounded and, you know, people who study hide and white will emphasize that point. And I'm not sure it's always been heard as loudly as needed to be. He always believed that. So I always say to people, it's not anything goes. You don't get to just write whatever you like. You know, a good history will have a relationship

29:07to body's previous histories. That's the other thing is how does a history relate to what's come before, you know, I would say to people, it's a bit like an elastic band. A great new history will stretch the elastic band at the topic, but it won't break it so far. That everybody else working in there, he goes, I don't recognize that. Like that conversation I had with people in 2018 and I'm 10 and they said, I don't know what you're talking about. I actually don't know what you're talking about. And then the world changes a little bit and they go, I do know what you're talking about. It's really interesting what you're talking about. And now I'm really worried about that. So partly I'm kind of interested in how the focus of history shifts over time. And that comes back to people like Collingwood and people, you know, I'm really interested. I did my doctorate on it. I've never really escaped. Yes, it's not it. I don't think the same things as him. But he was interested in the metaphysics of history. How do they make sense for people? How do they make logical sense for people? How do they have authority? And it's not just that there's one history in the world doing that. There's histories relating to one another,

30:08sparring with one another. Some history is much more accepted and seen as having much more authority than others. Some being very influential. So the point where everybody quotes them, and you go, OK, I'm going to have to quote X in order to write a book on Y basically. But then 30 years later, nobody quotes X and nobody quotes Y. And again, it's not about fashion. I always say to people, again, it's not about anything goes. You can't just generate a generative AI that's really eloquent. So I say to my computer science colleagues, many of whom really love history, by the way. Alliquence is not the secret here. Creatability and authority are really, really important in this space. And a gen AI that's really eloquent will never take the place of somebody who's done the hard yards around an archive and found an archive. And gen AI doesn't discover archives, at least at this point. So it's going to tell us stories or new combinations of things that we already know. But it is not going to bring to light many of the archives that have not

31:09been digitized around the world. Those stories have not been taught. There are many at this point. And again, big question is, does AI exacerbate the historical graphical divide that's already out there? It just accelerates the production of histories around topics that everybody knows. Do we have a popularity bias? Will we miss a trick? So some people from different nation states have said to me, I'm really worried about our particular nation's history being swamped by these other popular stories. And I go, well, it's not as if it hasn't already happened. But this is where these debates about sovereign AI are coming from. Do we have to build our own AI? I said that our own local story is told and that our own forms of credence, because if I go to an Aboriginal community and tell a history, they'll go, yeah, that's really nice. But it's not as authoritative as this particular way of telling history. So the age of historiography is a harmless Patrick, and I couldn't be more excited. I was too, yeah. You know, I made a bet on this discipline 30 years ago. And everyone said, you're out of your mind. I'm going to, I think it's a really good time

32:09to be a historiographer. I think it is a very incredible moment. Couple of thoughts in response to what you just said. First of all, you were talking about this unfulf potentially growing divide, a historiographical divide. It makes me think of what you said in your book about Spotify and its promises to actually, even out the playing field for less known artists. So actually, it turned out to be a popularity, context, a large extent, right? And the second thought, again, you mentioned the case of the Australian Indigenous communities and how they might potentially be implicated in that process of historiographical divide. And so that brought me to this beautiful tradition of dreamings that you describe in your book, Can Guru Dreaming? Could you talk about what this means and how that fits into your argument? Yeah, now there's different words for it around Australia. And I said to people, a lot of people around the world are familiar with songlines or dreamings, but it will differ according to the language group.

33:09So I just want to acknowledge that. And I always say to people, I'm not an Aboriginal person, so I can't speak for Aboriginal people in Australia, the best thing you can. It was read Aboriginal people writing about, now for Aboriginal people, the ways in which we tell history are absolutely grounded in the wear of the history. The places that people have a deeply meaningful to them. And so often the telling of history is about where it happened and the power of that particular location. And so the location can take primacy over the chronology. So it's possible in some histories where you might have people from different historical periods appearing in a history and to a Western sensibility or to a particular sensibility that can't happen. You've got to respect chronology, but actually the moral force and the power of that history is that it's turning on the place that it happens. It's a really powerful way of thinking about the world. It says, never forget where you are and where you've come from. And what I love about that is it's reminding us, AI can help us to think we're kind of dematerialized history.

34:13We can just produce histories really quickly by punching in a few words. But that goes back to some data centers that are located all over the world, right? And more being built all the time. And some of them being built in really, you know, insecure places right now. And some of them being built in communities where people really objecting to the environmental toll. And I've just published a paper in environmental history on, you know, the use ironically of synthetic insect algorithms, you know, connected with AI and the building of data centers where some insects may be actually made extinct by the movement, the global movement, which is material, absolutely material. So Aboriginal colleagues and other First Nations colleagues often remind us, where are you speaking from? Where are you speaking from? Is history gonna end up like an airport where you land and the plane you get out and every shop looks the same? Is that what you want? Or are you gonna acknowledge that the place you come from has got some really special things about it?

35:14And deep history here, over 65,000 years of history and the passing on of histories, teaching us how to live with this place, this very fragile country, you know, and there's wildfires in Ontario at the moment. And notion here of living with, understanding the fragility of the places that we are. So Aboriginal people in these dreamings are Jirkaban, different languages talking about, you know, what have we been taught over time from these places? How do we go back to and revisit what really matters? So it is history made every moment for all of us to think about. And again, that ethical conversation to say, I'm gonna tell you a story and that's the beautiful thing about Uncle Lewis, Yolabuka and Arnihan, they tell a story and you had just sit there and think about it. What am I being told here? And you'll go, I think you're saying this and they'll go, good, you got the message for you. Now that's again not and everything goes. The message is very strong and very solid.

36:15But there's no doubt, it's powerful, right? History is meant to be for the moment and for the future. It's almost as if we're going into the space of religion art when you, you know, the way you describe these experiences. It can be, I mean, it's, you know, when you're here and you're walking on country, capital C country, the power of it is just extraordinary. And I would say to people, every history we read is powerful. But if we've forgotten that's power, have we read so many of them or we've lost the sensibility of how awful and terrifying some histories are, you know, it's like you show some people some histories and it's like the fact compassion fatigue. I've seen, I've already read 20 books on that, you know, I already know what the story is and you go, yeah, but people were killed in that. The story there is really horrific. Have we remembered we don't want to be overwhelmed by the call that we're getting? And it's not necessarily religious, but it is ethical. It could be religious.

37:17It could be about wonder. It could be about or it could be all about those things. And I say to people whether you're religious or not, there is an ethical calling in the telling of these stories. But that doesn't mean that history is all about ethics. It's also about great stories. It's also about capturing people and really inviting them to think about politics and think about all kinds of things. But it is a profoundly powerful discipline and it's all forms and it is made for the moment. History is are written for the times that they're produced in and they are all calling us in different ways to think about what's going on right now and what's gonna happen. You know, people talk about the lessons of the past and they say, well, AI has been a really good sharpener of that notion that we should be learning some lessons from what we're doing here and think about what each of us needs to do right now. Not to overwhelm everyone with the ethics of the burden. I don't want your poor listeners to think, oh, I've got everything weighing on me terribly. Part of the conversation is figuring out

38:17what are the most important things that we need to think about right now. Yeah, as I was grappling through with your argument in the book, you know, other opening that I found into this line of thinking was what I remember and what I've read up on quantum physics. And particularly this book by John Gattis came to mind. He's called the landscape of history where he was essentially defending history against these claims that it's not a social scientific enough. And he argued that while the social sciences were trying to live up to the standards of Newtonian science, meanwhile, this fuzzy and dual and pluralistic world of history that you're engaging with became somewhat ironically more scientific with the development of quantum physics and relativity theory. So that's just kind of like, I'm sure this is on your mind too as you're actually going to see. Well, actually, I came to my mind as I was writing big and little histories. And there's actually a chapter where I look at the really tiny,

39:20I was interested in micro history and then got really interested in, you know, barad, Karen Barad's writing on quantum, whether things get different as you go really, really tiny, you know, and has micro history got smaller, you know, Mike Mugneson and his colleagues, he's done the magnificent job of really making a aware power of micro history. And I said, is it big? Is it too big? Should we go smaller? Does quantum get us to think differently about these things? I was really interested in that question. In that book, I kind of argued and said, well, OK, histories are ethical. They're generative. You know, metaphysically, they are, you know, strong together and made together. They make sense because of the ethics that's driving them. And the decisions that people make about how big their history is going to be, is it going to be about one second? Is it going to be about the quantum? Is it going to be about the whole world? Is it going to be about 14 billion years? Those are ethical decisions. And I argued in that book and said, you know, histories

40:21are, and don't get offended anybody, they're really artistic, which means no two of them are the same. They're crafted with care. And they're beautiful. They're absolutely beautiful. The ones that have done well. And in the wonder book, I had really good fun. You'll be aware looking at histories that perhaps not as beautifully crafted. So really interested in books that were released, that flopped when they were released, or just a bit strange or a bit weird, really interested, that obviously resonated with their audience, but you read them now and you go, that is really, really weird. Like world histories of ghosts, you know, some fantastic strange histories from China. Really, really interesting stuff going, OK, these are all our relatives, they're all histories. But there's some stuff going on here that we need to reckon with. And I've said, it's artisanal, and that's not an insult. That's like a crafts, who are all thinking, a beautiful thinking about what's going on, the labour of that work, for ethical purposes. It's not to diminish it to say it's effects. It's not saying that history is collapsed into philosophy,

41:21but it's to acknowledge just the level of virtuosity that's in that space. And so I always say to people, history doesn't have a replication crisis. Because if two histories came out the same, you'd be deeply suspicious. They both said the same. You know, argue to the same end, you'd think that some, there was some deficiency in the research. So I always say this to my science colleagues, we don't have the same issues that you have around replication. We don't expect people to replicate the same methodology. Even that, we don't have a methodology. The history, A methodology. We don't have a scale for histories. We don't have a, they must be this long, which my colleagues in computer science, they say, well, if you look at the data around what's best selling in history, oh, it'll be about 80,000 words long. It'll be on World War II. You can get all the data, and you can get a machine to write on. And I go, yeah, sure, you can absolutely do that. Lots of people probably will find those histories interesting, but I will always get fascinated in the person that, you know, says, I'm going to take this data that everybody's read this particular way, and I'm going to read it against the grain.

42:21I'm going to give you a new history, which if you look at the recent winners of the Pulitzer Prize, of the Cundle Prize, which I did in this prize book, they are amazing books. Absolutely amazing books that are, you know, even in some cases, topics that we think we know really well. And you read the book and you go, oh, I had no idea. That that could have been the case. And I find myself educated and I find myself corrected. So that was part of the fun of the book, was reading all these prize winning histories by humans, including children's books as well, which was great fun. Really fun with this going through these introductions. I really had fun. I was looking at, you know, the way they're using counterfactuals and models in children's books and saying, this is not an adult-only thing in historyography. And in fact, there's not enough historyography written about children's books to look at the kind of complexities. If you've only got 150 words, how do you communicate complexity to an audience of 10-year-olds or nine-year-olds? And people do that really spectacularly well.

43:21And to my view, it's a bit of a shame that we haven't looked at the kinds of forms of logic that people have used, communicating to children and saying, actually, children do get these complexities. Yes. Pretty well. We don't just magically figure out how to read histories. We guide them by people along the way. Absolutely. Yeah. I think the children's culture is getting the short end of the stick all the time because maybe there's not so much money in it, but that's unfortunate. Well, just people just have this to be the research is more prestigious than education outside that belong together, people that have always belong. This episode is brought to you by Facebook. So you were scrolling on Marketplace. And there it was. The bike you've been searching for, you sent a message, and it turned out the seller was super chatty, kind of funny, and an avid cyclist. The next thing you know, you're in a cycling crew, well, a community cycling group. The thing about Facebook, you might find more than what you're looking for. From a browse to a bike ride, this summer find more on Facebook.

44:24This podcast is supported by Anthropic, the public benefit corporation behind Clawd. Everyone's got a hard question about AI. Your job, your kids, where it's going. Anthropic was built to surface those questions and share what it finds along the way because there's hope in hard questions. Ask yours at clawd.ai slash Spotify and keep thinking. That's clawd.ai slash Spotify. Collas it up with Crocs. You know back to school is coming in fast. So why wait to find your new Facebookware? Step into a local Crocs store and step into your new look. Try it. Style it. Make it yours. Because the right pair doesn't just show up. It shows off. First day fits, handled. Walk out ready for whatever's next. Visit your nearest Crocs store today. It also makes me think that this kind of ambiguity

45:25that we should be grasping for more as you make the case and that history relies on it. Technically speaking, within the world of the enlightenment of which we are airst arguably, this is like only there are only a few pockets of this vague, open-ended space because otherwise institutions operate on the precision mode and on a very strict organization mode who wouldn't you agree. So we're kind of fighting against going against the grain in that sense. Yeah, look amazing. I was, I'm doing a new edition of 50K and I'm riding a new entry on Walter Benjamin and talks about a thief jumping out on the road and stealing the certainty that you think you have. And you just go, he was so onto it. He was so onto it saying, the angel of entry is just looking back and seeing the wreckage. Can't look forward, but what it can do is see the wreckage and you just look at the fragments of his writing and he is constantly tripping us up

46:27and we're falling over and kind of trying to make sense of what he's talking about. He's really hard work and fantastic. And so there are, absolutely. This is what the humanities are all about. It's the saying people, there is a gap and that gap is an invitation for you to do the work here. And so you do, you know, my colleague's saying, my students are cognitive offloading. People are just doing this and I go, well, the invitation is to figure out how do you make the gap visible. So the student understands that closing that gap could be a problem for them. If they don't spend the time thinking about the importance of these things themselves, that they themselves could be at the wrong end of the closing of the gap in those ways as well. So yeah, but I think that's what the humanities is for. Not to be a nuisance, I said to people, we are not a nuisance. We are a reminder that thank goodness we have these gaps to think through and think what the right thing to do is what we want to do. So I've said to people, I think what's happening is that the economy is shifting towards an ethics base

47:27or a values base away from a knowledge base because, you know, although it hasn't got the archival piece sorted out, knowledge is pretty cheap, but ethics is not. Values are not. And so we are going to see continually reworking around our economy to the value of those things that a person who looks at two pieces of information and says, I don't like what I see here. Or I'm just remembering Mary Beard. She said, the biggest problem we ever have is when there's only one source and you've got no other source to compare it against, it can tie on more authority than you want it to have. You've got to reckon, you've absolutely got to reckon with that. Yeah. Speaking of dangers, you point out that the histories made through online data collection can be used to facilitate, I'm quoting, facilitate the surveillance of political dissidents. They can fuel conspiracy theories, enable cyberbullying, some pretty awful stuff. And deny people access to education, healthcare, justice.

48:27And I think everyone who's listening will be familiar with some of these. So it's only likely that AI will enable these process on a greater scale, isn't it? Yeah, look, it accentuates what people are already able to do, what criminal folks are already able to do, it just makes it easier in some cases or faster to do those things. And it's interesting, all the topics that I've ever worked on, AI excites the strongest polarization I've ever seen. Some people are really excited about it, and I think that's great. And some people are really angry about it and really anxious about it. And there's a lot of people in the middle going, I'm just not sure what's going on here. And I would say to people, my responsibility is to make you aware of the really good stuff it does, but also to be really responsible and say, like any object, it can be weaponized, and its susceptibilities can be used by people who have malintent, or in the case the example that I gave you before, you've been algorithms going about their business, and they could generate malicious outcomes

49:28without any human, really telling them what to do. And that can enable not just, you know, attacks on individuals, it can actually end and stay on state actions as well. No different to what we had before in history, let's be honest, it's just easier to do it and faster to do it now. So I don't want to terrify people, but saying, you know, do you know where your data is? You know, none of us want to be the victim of a cyber attack to have our data put on the dark web and used in different ways. None of us want to be open to that kind of risk. So in the discussions about AI, are we paying enough attention to safety and cybersecurity? So the space I'm working in at the moment is kind of pulling in a bit more of that market orientation around history, but also talking more about cybersecurity. And again, that's space where people go to me, why are you working on cyber security and history? And I go, it's fundamentally, you know, if it's taking my data and processing my data, my data is a susceptibility, my history is a susceptibility that people could take advantage of.

50:30And some people will, you know, not be thrilled that I'm telling you this on the podcast and I go, well, I'm sorry, the criminal is already know this. So it's not as if I'm listening to your podcast Patrick that people will suddenly wake up and have a malicious idea. They're already doing these things and there's already great people out there that are countering it in different ways, in really interesting ways, most of which we just can't talk about because we don't know what's going on in there. But it does take us to a discussion about international relations to cybersecurity and to the dark web and people's data being misused and stolen. And, you know, again, it's not to say that historians have to shoulder all of this responsibility, but in the teaching of history and research of history, we do actually have to talk about archival, you know, kind of taking advantage of mobilizing archives and misusing them as well as using them for good. It seems to me that, you know, we're looking back with, so used to apologizing for this logic of history that you're talking about for the openness. It might be even difficult for historians alone

51:31to turn around and somehow take pride. And also it seems somehow look for sources of authority in the public eye that suddenly historians come out and talk about these things as real phenomena that can actually endager real human beings. Where do you think the opportunity will come from other than words? What kinds of partnerships might you envision in this process in the future of historical discipline, even that would make it possible and somehow feasible for historians to talk about it and be listened to other than, you know, just kind of a whole bunch of nerds who are sitting in the office and writing and writing beautiful stories about the past. This is kind of a devil's eye, I'll get the question. It's a really interesting question. And some people say to me, I don't want to go out and speak publicly because I'll be attacked by trolls or, you know, I'll have death threats. And I do know colleagues that have actually received death threats for just, you know, pointing out with responsibility that the archival record is perhaps a little bit more complicated than people thought.

52:33And so that's not a necessarily invitation for all of us to get out into the public media or social media. And some people do it really, really well, I have to say. But one of the most popular trends in AI is the customization of big AI models for particular corporations and institutions. So, you know, people say, oh, they'll say, oh, we're going to use this tool, we've named it Fred. And I'll go, okay, Fred is just Claude. That's been customized for my business or my institution. Yeah, Fred is actually a Claude, okay. This is, I think, the biggest opportunity for historians to say institutions and organizations everywhere are taking some of these bigger corporate models that they've developed to their own models. And they customize them for particular workplaces. The responsibility of the historian is to point out and say, we don't just need a technical authority in organizations to implement AI, we actually need a trust authority within these organizations. And a lot of people are saying that, you know, if you need to tell me where you're using AI and I go, we've just done some work with beautiful work

53:34with the Australian Historical Association saying every spell checker and every piece of grammar checking now has got AI sitting behind it. That's about being visible in that space. But historians putting themselves in those spaces to say, sure, you could be using, and you're going to Fred in your organization. But what is Fred going to do with my data and everybody else's data? And is Fred focused more on some kinds of data than other kinds of data? And is therefore Fred, could Fred be adjusted to be fairer to people? Could Fred, particular, you know, could we erase the data after a certain period of time? And could Fred be more respectful for first nations people? How are we going to treat that data with respect because of indigenous sovereignty? Claims. So I think one of the biggest opportunities we all have is within our own organizations. That's not us just talking to one another. That's actually being at the space of saying, okay, testing out and asking in these questions and then every school on the planet is going to have its captured AI, every hospital on the planet. There's got its captured AI. Everyone is doing this at the moment.

54:35So historians are probably interesting and valuable everywhere. So everybody from cemeteries right through hospitals has been interested in what I do. And I say to historians, how you do what you do with great integrity has got a pretty broad application as people think about, should I make an avatar of my grandma? The little book I've just published for Cambridge University Press, I've got a little section on the ethics of avatars. You know, is it okay to make an avatar of your granny? Is that a problem? You know, does she have any rights at all? You know, if I make that avatar of granny, is she just going to be abusive to me if she was not a nice person? Or am I going to be abusive to her? And you might have noticed that China, for instance, has just actually made a ruling to regulate more heavily AI companions. And that includes avatars of people that have died. They just announced that two days ago. And they've said they want people to form real relationships. Or relationships, right? But, you know, I know people have been generating avatars of loved ones that died because they've got elderly relatives

55:36with Alzheimer's, keep saying where's X? Oh my God, oh, here's X. Here's X, right? It's a really complicated and interesting space. So I'll say to historians, you've probably got more going for you than you've ever assumed. But you can feel a little beleaguered when you look at all the cuts across the universe that is and you can feel like you're really on the down and out, actually not. I don't view it at all. I say this is our time. People stand up. We hardly, this is our moment when we get to stand up inside of people, we can be really helpful for you. So on the one hand, it seems like one way to think about this anchoring of historians position would be to think of of historians as kind of authorities and people who try to read the world and make it more just. But this might not appeal to everyone. And so another way of approaching the problem was to think about this authority as an authority on trust and reading sources, right?

56:39Is what I'm hearing because this is potentially relatable to everyone in those organizations that you were talking about. If there's no trust, and there's very little trust right now. Right, right? If everything's polished, if everything's perfect virtually, then how can I trust that this is not actually doing something weird with the data that I'm putting in? I absolutely agree with you, Patrick. And I have been a senior executive, I did that for over a decade. And trust is the missing middle in a lot of organizations right now, the reasons why people don't have an optimistic view of governance in some spaces or an optimistic view of where things are going is that they don't feel part of it. That AI might not have been explained to them in a way that makes sense to them, that they don't feel part of the conversation, that they can't bring their expertise. You know, as a big debate in Australia at the moment about the safeguards and rail lines that have to be put around a rail, particularly for creative creators. You know, is it okay for big models to just take the creative ideas of people and then just replicate the stuff that people who are really, really hard on? So saying to people, yeah, this is a really important moment.

57:43And it doesn't mean you all have to be out on social media. I'm not asking for that. But it does mean asking the right questions in your institutions and rebuilding trust at the centre of institutions in a way that people feel like they're not being left out of it. And starting to use language like corporatisation, alienisation, the things that lead to bigger gaps around wage fairness that mean that people don't get living wages, things like that. This is where people need to be focusing and saying, doesn't mean you need to be an activist. I'm not asking people to be that. I'm asking them to be true to the authority and the trust that they have at the centre of their discipline and to express that in the way they teach students in the meetings they go to. And to remind us of what really, really matters, even if that's inconvenient for a lot of us. And a bit of a new sense, and I go, a bit of a new sense is a great conversation. It's a really important thing. I mean, it's good to have that friction. It's good to have a challenge. And this is the challenge for the moment, it seems like. On the other hand, I am also thinking, again,

58:43devil's advocate, I went to a conference a couple years ago and the big question was, how do we make just AI? And my first thought was, it's a great question. And I think they were also asking, why doesn't the left have their own AI in the way that the right does? And I'm thinking, well, because that's where the money is. It seems like, that's where the money is. That's where the capital is. This is what brings profitability. So surveillance and all this and control and power. So how do you imagine flipping this around, particularly at a time where all these other indices of democracy of trust are sort of going in the completely wrong direction from a lot of people's point of view? Yeah, I chuckled there because a lot of people say that to me. And they just say to me, AI is just the latest expression of global capitalism. That's what it is, right? That's what it is. And I go, well, hang on, no, it's a bit more complicated than that. Depending on which part of the world that you're in, there is definitely a market facing capital model.

59:45But there's also nation-state AI's that have grown up that you wouldn't describe as coming according to a capital model. And then there's places in the world that have a hefty hand on regulation for really good reasons. So, for instance, the EU has been very strong on regulation for historical reasons. They know what can happen when people's data is mobilized in ways that are just genocidal, hateful, and harmful for people. And so I say to people, it's not a one trick pony out there at the moment. You do have some of these market models that have had to change and had to really accommodate the regulatory push of the EU. The EU has had to think about, well, is it going to make models or does it make regulation? Meanwhile, other nation states are just happily generating different AI, so China has just released yet another low-cost AI model is cheaper than the US capital-based models. And so the US ones, there's not lack of competition out there.

1:00:45So I say to people, it's not a one trick pony. It's not done undustard. And I don't agree that there's going to be one AI. People who used to talk about superintelligence in the one AI and I go, no, don't think so, because I know that there's been a thing called into national relations that's been around for a very, very long time, had a lot of AI replicates or takes advantage of. I love that. I love that. I love that analogy, right? As the extension of nation states and international relations between national avatars or what, not maybe, I don't know. And people, it will reshape some of international relations and know down about it, but a lot of it will replicate patterns through pre-existing patterns of relations between nations. And corporations are in there, but people say to me, oh, corporations have taken over nation states. And I say, well, hang on a minute. There was a thing called the Dutch Empire, which was a corporate thing. But as a corporate thing, if you know your history of organizations, you'll say there's always been multinational corporations. Whenever people have wandered around the world, they have done business in ways that cross over borders. So it's not a new problem to people. So again, historians of organizations, historians

1:01:48of international relations are extraordinarily helpful. At this time, to help us kind of calm down and say, it's a little bit more complicated than that, but we can probably make sense of what's going on. But again, that notion of it's all going to collapse into an infinite point where we have one history in one AI and I go, oh, I don't need it. No, I'm not going to. It's not going to happen. I love that we're finishing on an optimistic note like this. And may I ask one last quick question of you. Maybe could you share a sentence or two about your upcoming project? Yeah, so thank you very much. So I'm working on three things at the moment. The first is papers around this notion of cybersecurity and algorithmic history walls. So that's a little bit of stuff going on. But I'm doing the fourth edition of 50 key thinkers on history at the moment. And one of the things I said to the publisher is, could an AI write this? And I wrote, I mean, no, it can't. I can't write this. It really can't write this. I need to do this. And then I have proposed a new book, which I've called Artificial Intelligence and the Future of History, which is looking at the relationship between contemporary AI and history and saying to colleagues,

1:02:48what are the different scenarios? Some of the things we've just been talking about, what are the different scenarios? How could we get engaged? What's the value of history going to be now and in future? Yeah. Super. I look forward to following your work. And meanwhile, thank you so much. It's been a wonderful conversation. Thank you, Patrick, for really enriching, fantastic hour of talking. And I have to just cut it. Thank you so much for engaging with the book. And particularly for Arni Ann and Uncle Lewis's ideas. And as always, you and your listeners are very welcome to get in touch, very passionate about spreading the word about this topic, getting more people engaged. And your podcast is really helping more people to get engaged with history. So thank you. With summer coming to an end, you're going to be spending more time indoors. Get your home ready and save up to 50% off custom window treatments during the Labor Day mega sale at Blinds.com, from room darkening to light filtering shades, finding the perfect fit is easy. Get free samples, expert design help, and professional measure and install services.

1:03:50And if you're interested in learning about the new product, you can find it on the website on the site. Get free samples, expert design help, and professional measure and install services. Or DIY with confidence and support every step of the way. Shop up to 50% off site-wide plus huge savings on door busters right now during the Labor Day mega sale at Blinds.com.