Rewriting History

Speaker A: 00:00:00

Hello and welcome to the Cognitive Engineering Podcast produced by me, Fraser McGruer for Aleph Insights. In this series of podcasts, we take a look at interesting topics and discuss what we think they tell us about analysis and decision making. I am here with Peter Coghill, Chris Wragg and Nick Hare of Aleph Insights. And this week we're discussing, is it ever a good idea to rewrite history? And due to the lockdown, unusually, we're recording this remotely.

Speaker B: 00:00:29

Nick, what do you reckon about this? Well, we're recording this in the wake of a controversy about Boris Johnson's advisor, Dominic Cummings, who has been controversial for the particular reason that he went, during the lockdown, went up to Durham when he should have been staying at home, went to Barnard Castle, which is a tourist attraction when he should have been staying at home, has some very superficially very flimsy excuses for having done all this. And of course, that's particularly controversial because he was a part of an administration who were telling everyone else that they must stay at home and can't see their relatives, etc. We don't care about any of that. The thing we're talking about today is the fact that during this particular time, he apparently edited a blog post he'd made last year to insert a mention of coronavirus. Really, you must assume to make it look like he had some special foresight. We know that he's very into forecasting and that kind of thing. So what it looks like, and I'm certainly not saying this is the case, but what it looks like is that he's tried to make it look like he was particularly farsighted about coronavirus by editing his past blog post to make it look like he was writing about it when he wasn't. That was picked up for two reasons. Firstly, that the past copies of his blog that are kept by the Wayback Machine, which is the Internet Archive, don't have that in it. And secondly, that his website's sitemap information records that that post was edited this year, so on the 14th of April. So he's been rumbled in doing this. And the question I think we want to talk about is this whole question of editing history. Is it harder now? Is it easier now? Is it harder now? Is it something that we should just be okay with people editing past documents if they can? And what are their hopes for in the future for potentially having technological solutions to make it harder and to make it easier to rumble people? Or in fact, should we just be happy about it and just say, well, everyone does

Speaker A: 00:02:57

it who cares? I think I forget the name of the Canadian chap we had on talking about. He was really keen on the Byzantines. Kristian Gustafson. Yeah. I seem to remember we talked about this a little bit with him as well. Because I suspect this sort of stuff, you know, in the nature of history, it's always been rewritten, in a sense. But now there's different ways of doing it and finding out that it's being rewritten. Chris or Peter, Nick posed a question there with some

Speaker C: 00:03:30

questions there. What are your thoughts? Yeah, well, I think, you know, obviously, documentation and the historical record has always been used to sort of counteract, you know, historical revisionism or denial. So you know, you look at things like the Holocaust denial itself, or the Nanking massacre by, you know, Japanese forces, and other, you know, the Armenian genocide, you know, what ends up sort of meaning that those kind of exercises in historical denial are very difficult is the sheer weight of the historical record. And most of that exists in documentation, you know, it's rosters of people being checked into places, it's, you know, train activities, or, you know, other documentation that's been written. So I think we have it, you know, it is difficult to completely change things that are mass events that were witnessed by, you know, thousands or millions of people. And where you've, where you've, you've got large amounts of physical documentation, which is difficult in itself to destroy, right, just by pure volume, you know, I mean, obviously, they tried burning it and so on. But, but I think where there's, where we're now moving into territory, where a digital record creates something valuable is for things which were only witnessed by a few people, and for which there wasn't much documentation, which people can quite easily sweep under the carpet and destroy, which I think is much, much more difficult to do now with, you know, people's social media posts from five years ago, or something being dragged up, you know, when they're appointed to a prominent position.

Speaker D: 00:05:31

Yeah. And particularly when people in power make predictions, or assertions or promises, they need to be held to account. So it's, so I think the internet has given us an amazing tool for this. But I think, but I, but I worry about sort of how sort of fragile it actually is, because the fact that Dominic Cummings' blog was, the edit was discovered was only by chance that it was one of the blogs that happens to be archived on the Wayback Machine. And that's not a unit, that's I can't take that as given for everything. It's just by luck that it was one of the things that

Speaker B: 00:06:09

got indexed. Yes, a tiny percentage of things are actually archived on the internet. Yeah. Which is not something that people think about or worry about enough.

Speaker D: 00:06:16

And it wouldn't, and it wouldn't take much more sophistication on Mr. Cummings' part to cover his tracks. So it'd be quite easy to delete from record something that you had previously published or change it. Much harder than it was, say 30 years ago, when such, like opinion pieces like that would have been published in print or in a paper, for example. And then copies would be distributed and held in libraries and other sort of institutions like that. So it's, on one hand, it makes it very easy. So fairly unsophisticated analysis can reveal an amazing amount about things. But only if you're lucky that there's some record of it somewhere. So I think there is a case, and I think there is a market for something like the Wayback Machine that you elect to put your publications in, to be part of an immutable record for forever or for some period of time. As part of your self-declaration, a stamp of what you believe to be true at that time. Because otherwise, it's a bit of a sort of free-for-all. You can't rely on blog posts and articles on the internet, and in the same way that you would the microfilms at a library of

Speaker B: 00:07:44

er year being produced by the: 1800

Speaker A: 00:11:05

go in and edit it. Nice roundup, nice sort of, yeah, I like that sort of overview. Chris?

Speaker C: 00:11:11

Yeah, well, I think there's also something at play, which is the sort of the trust in the written record as well that is being eroded. So the fact, you know, facts may exist, but those facts are called into question just based on, or the record of the facts may exist, called into question just based on the motivation of the organization producing those facts or whatever. So this is the whole kind of fake news defense really, which is yes, you know, you might have a recording of a particular thing, but because we distrust you and your political motivations, we're just going to dismiss evidence. So I think there's also a dangerous trend towards, you know, even if you couldn't change the written record, the written record itself becomes not believed by, you know, a large group of the public.

Speaker D: 00:12:11

Yeah, a thought occurred to me that we've sort of allowed this sort of thing to happen, because more information over, certainly over the last 20, 30, 40 years, just the amount of information has exploded, the availability of information and stuff, how much you're bombarded with on an early basis. And so statements that historically would have been quite important, so statements made like the equivalent of a Dominic Cummings' blog would have been a speech that he would have given at some big public event. That would have been transcribed and recorded in papers, probably. Because the volumes of information were much lower, it would have been more important in terms of just its informational value. But because there's so much of it now, public statements given by even senior people, the President of the United States, are treated much more flippantly, and we don't hold them to account so much. So Trump is constantly in trouble for pretending to have meant something else in the past when he said something, or not flagrantly denying he ever said something, but trying to twist what he had said previously to mean something else. So we let them get away with this. I would like to see somewhere, in order for you to take anyone seriously on the internet, their blog or their speeches or their YouTube channel, for them to elect to have all of their stuff immutable, immortalized in some sort of record. And if they haven't got that badge, then they're obviously not willing to stand by what they have said as a sort of badge of authenticity.

Speaker A: 00:14:02

Sorry, Peter, I slightly missed it there. I think it's a really nice point, but what are you saying that makes us more forgiving of these sort of statements?

Speaker D: 00:14:12

I think the trend to so much more information being out there, each piece of information becomes less important. So it's treated as being less important. So you're more willing to forgive inaccuracies or things that just don't make sense or don't fit with your worldview. You're just more willing to sort of ignore them or let them pass by, rather than getting exorcised by them.

Speaker A: 00:14:39

We don't mind our celebrities not being Frank Sinatra anymore, we just get that they're lower

Speaker B: 00:14:43

quality. I wonder if, I mean, whether or not the level of scrutiny that we used to apply in the past when there was a low volume and the low volume of sort of public record, and it could be scrutinized, whether that is a bit demanding to the idea that we're always going to have been right in the past. And if we said something and changed our minds, whether we should be held to account for that, which reminds me of this debate about which way it's going to go. And I think the jury's still out, but which way it's going to go about social media catching up with us. Because on one hand, we have a situation where I think we're still pretty judgmental about things people have done on social media. Someone who got a photo of them getting drunk or doing something silly at a party is still considered a dangerous thing to have. Whether or not in the future, we might just be much more forgiving and say, well, we all make mistakes, whatever, and forget about it, or whether we go the other way and simply stop producing. So whether we just stop engaging with social media because we're scared of what might happen in the future if we say things. And I don't know, I feel like there are two scenarios there. We either do try and sort of make it much harder to get away with Dominic Cummings style fiddling with the past record, in which case there'll be less of that past record because people will be much more reluctant to, for example, publish blog posts. Or we just get very lax and say, well, yeah, fair enough. Trump's saying one thing one week and another thing next week and that's just

Speaker A: 00:16:25

what he's like. Where do we go from here? About sort of three quarters of the way through?

Speaker C: 00:16:32

Well, I think the fundamental question is whether there is something sort of inherently sinister about trying to change the past record. Because if you look at generally organizations, regimes, systems that are associated with trying to change the record explicitly, you look at things like book burnings, for example, is generally not good people trying to do that. And I'm sort of thinking of, you know, I'm trying to think of examples where there is a positive aspect to being able to change the past record of things. And I really can't think of one. It feels like even if it has, you know, a sort of a negative effect, you know, like it causes lack of trust in an institution that we particularly need to trust at that time or whatever it might be, it still feels like there's never really a good reason to change the historical record.

Speaker B: 00:17:39

Yeah, if you are motivated as a leadership of a country to behave in a way that resembles something that George Orwell wrote about, it's probably a bad idea. It's a kind of good rule

Speaker C: 00:17:51

of thumb. It's not meant to be a how-to guide. I mean, there's one example where that I kind of came across, which was quite, quite interesting, which was John Flamsteed, the royal astronomer and contemporary of Isaac Newton, who had a massive falling out with Isaac Newton, totally got a massive hissy fit with him, and they had a lifelong feud. And he basically, you know, was publishing lots of observations of planetary and sort of stellar activity and so on, and publishing all this data. And he was very, very methodical about publishing it. He wanted to double and triple check it, and he wouldn't hand it over to Newton and Halley, who wanted it to support their theories and so on. So Newton basically forced him, he went to Prince George and forced him to hand over this data before it's properly checked, and published it in some of, you know, had it published effectively. And so Flamsteed went out and bought as many copies as he could find of his own work and burnt them, you know, outside the Royal Observatory. But the point is, his motivation there was, I didn't want to publish something inaccurate, this is wrong information, I want to rescind it and publish the correct version. And there, I suppose, you know, there is an argument. But I would still say what you do there is you publish an addendum, so people can see the original record, and subsequently that you've changed your mind. That's all people are asking for, you know, that you can see somebody has reformed or improved their information or changed their mind. But getting rid of that record of what they previously thought is probably not the right thing to do. Now there are, I know there's

Speaker B: 00:19:46

arguments as well about the results of Nazi experiments, whether or not they should be destroyed, because if they were used, it would give some, you know, because obviously, there's quite a lot of actually potentially useful medical data there. But if it were used, it would sort of give some kind of seal of approval to the Nazis having done it. I mean, that's very edge cases. I mean, I think the key question here, which Chris is talking about, and this really comes it's a similar situation as with Cummings is who owns information about themselves, who owns the information that you've produced, you know, with so with this Flamsteed situation, presumably, you know, he would have taken the view that this is my information, other people do stuff with it. Well, you know, maybe Dominic Cummings, fair enough. He's like, well, it's my blog, I can change, I can change previous entries, it doesn't become owned by other people. Now, we can make a judgment about his deceptiveness. But whether or not you know, he did he have the right to do it, it's hard to say that he didn't. It was well within his right to do it. And yet,

Speaker D: 00:20:56

as he is well within his right to exercise the the deletion from the Wayback Machine, the assumption assumption, though, within the press that's covering this is that he did it for for insidious reasons, that he wanted to for for for gain, personal gain. But yeah, there's no reason

Speaker B: 00:21:16

n illegal pyramid scheme with: 1111

Speaker A: 00:23:35

can you have an immutable website? I'm gonna say that question is rhetorical. No, it's not. Well, well, okay. If it's not rhetorical, we kind of need to sort of draw things to

Speaker B: 00:23:48

think about that. How can we how can we have immutability? Yeah. Is that? Do you want to say?

Speaker D: 00:23:55

Yeah, I mean, there are there are technologies which are designed to provide exactly that. So the technologies that underpin Bitcoin, for example, provide you an immutable record of all transactions that have occurred. And that is crucial for the current that currency system to work. And similar thing. So the similar similar approaches are used for digital signatures on contracts, as well as what they call distributed apps. So things that trace the validity or quality of something through the manufacture chain or the transport chain, they all rely on some kind of immutable record. And we have digital technologies which can provide that to

Speaker B: 00:24:39

ly an advanced form of making: 1000

Speaker D: 00:24:55

It's but it was more than just a distribution. It's the fact that if you try to amend something, you not only have to amend everything, but by doing so it instantly alerts everybody that has happened. So it's more than is better in many ways than having multiple copies of a newspaper in hundreds of libraries across the country. Because you could still go to all of those libraries, and you could still replace the paper and be do so undetected. But these systems will alert everyone that something has gone wrong. So let's finish this off. Let's sort of bring

Speaker A: 00:25:25

it back. Let's make it personal. Is there anything cringe worth we've done that's on record that we would like to erase? Or I guess, I guess we could be a bit softer and say is there anything that we haven't done we wish we wish we had put on public record bit like bit like Dominic Cummings tried to do? Yeah, I don't know anyone. That's really incriminating.

Speaker D: 00:25:48

There was a there was a photo of me when I was at university and I was a bit overweight. That has mentally scarred me ever since I saw it. But I don't control that image because it was on somebody else's feed. But yeah, it was I think we were on the beach and I was I was unaware I was being photographed. I wasn't sucking anything in. And I just thought bloody terrible for

Speaker B: 00:26:10

a website. About it was about: 2002

Speaker C: 00:26:57

I'd have to leave home. Chris? Well, I mean, I'm particularly thankful we didn't have social media when I was younger, because I think so. You know, I think I've alluded previously to the fact that I wasn't a terribly, terribly positive character at school. And I think I could have been one of these people who would have been haunted by that, that record throughout their life. But the thing, the sort of single piece of information, which I think it doesn't actually relate to me relates to a friend of mine who we convinced to he had a sort of he had a kind of liaison with a with when he was a student with with a woman. And he, you know, it was obvious this thing was, you know, a kind of one off event, but he became a bit smitten. And we managed to convince him out of nothing other than sort of a kind of sadism, really, to, to write her a postcard sort of confessing his, his love. And yeah, and it never, it never got replied to. But we still, to this day, rip him about how that must have been received. So yeah, yeah, he would have definitely asked for that to be burnt.

Speaker A: 00:28:20

Yeah. I can't think of anything for me. I mean, my life's an open book.

Speaker B: 00:28:25

Spotless history.

Speaker A: 00:28:27

Unfortunately, I'm, you know, I'm lucky in this. I admit I've never made a mistake in anything. So, you know, that's fine. Yeah, I can't think of anything, but I like your examples. Okay. All right, we'll stop there. Thank you, as always, for listening to the Cognitive Engineering podcast. I'm Fraser McGruer. We've been here with Chris Wragg, Nick Hare and Peter Coghill of Aleph Insights. Until next time, goodbye.