Backchat: Luck, debate, and the quantum internet

Host: Benjamin Thompson

Hello, and welcome to Backchat. If the Nature Podcast is 1% inspiration, then Backchat is 99% perspiration. Yes, Backchat’s back with our first show of 2018. It’s also my Backchat debut. If it’s yours then this show is a bit more freeform than the regular podcast, and is a personal take on the latest stories from our team of reporters and editors. In today’s roundtable discussion, we’ll be trying to quantify the role that luck has to play in research, looking at how to report on the iterative nature of science, and how to avoid hyping the high tech. I’m Benjamin Thompson, and joining me on today’s show are Heidi Ledford.

Reporter: Heidi Ledford

I cover biomedical research primarily, but sometimes I write stories about plants as well.

Host: Benjamin Thompson

Lizzie Gibney.

Reporter: Lizzie Gibney

Hi, I am a reporter here in London and I cover the physical sciences.

Host: Benjamin Thompson

And Davide Castelvecchi.

Reporter: Davide Castelvecchi

I do physical sciences. Also, I guess maybe I would be the black hole correspondent.

Host: Benjamin Thompson

Coming up on the show, we’ll be talking about lucky breaks. Is serendipity an important driving factor in science or is it an unhelpful stereotype in the discovery process? Can we nudge the finger of fortune to point in our direction? We’ll also be looking at a blooming debate about the first flower. In science, new research can be contentious. How do we go about reporting it? Is there ever a time when a debate can truly be put to bed? Finally, we’ll be taking a peek at the quantum internet. Will it help me load cat GIFs faster than ever before, and how do we avoid overhyping it? Firstly then, science is littered with famous examples of fortitude. I think that Alexander Fleming’s discovery of mould juice, which you might know as penicillin, is perhaps one of the most famous. Nature had an editorial a few weeks back looking at whether you can divide up the concept of luck. Lizzie, I know this is something that you have been looking into.

Reporter: Lizzie Gibney

Yeah, I mean I think that there are so many wonderful kind of fables in science about the role of serendipity, and by serendipity we just generally mean a happy, unexpected finding, and it’s often used as a justification for why scientists should be really given free reign. But the thing is, we are in our line of work, we tend to be evidence-based, we are scientists, and there isn’t a huge amount of evidence that actually says what the role of serendipity is. So, we’re in a situation where it’s used as justification but mainly based on anecdote. So, there was a particular study that inspired this editorial and it was attempting to quantify in some way, to try and make this subject a little bit more evidence-based.

Host: Benjamin Thompson

Is it possible to quantify such an abstract concept then? I mean what have these guys come up with?

Reporter: Lizzie Gibney

Well, so serendipity is this like entangled web of cause and effect, so the very first thing that needs to be done is to have some framework that means you can actually study it. So Ohid Yaqub, who is at the University of Sussex, has been developing this framework for a little while. So, he went through the archives of Robert K. Merton, a famous US sociologist, who kept newspaper clippings and articles and built up this massive archive with lots of examples of serendipity. And he looked through those and he built on them and he tried to classify them in a way, which is step one of trying to understand a phenomenon. So, he came up with the four different types of serendipity in science that he found. So, one is where a discovery in one area comes about from targeted research in a completely different area. So, an example of that was in WWII when there was a mustard gas explosion and the investigators went to try and figure out if it was caused by an enemy bomb, and in doing so they noticed that there was an effect on the white blood cell count in the soldiers who were affected and this eventually led them to the idea of chemotherapy, so science was happening in one area and it led to a discovery in a completely different area. You also then have instances where you are just doing completely free blue-skies research. The analogy is you’re just going window shopping and seeing what you might like to pick up and there are loads of examples of that as well. So, Wilhelm Röntgen who discovered X-rays, he was just tinkering with a cathode ray tube and he saw some flashing light across the other side of the room that he was completely not expecting. And then there are others where, say, you are trying to get a particular result, but the way in which you get that result is completely unexpected. So, for instance, Goodyear discovering vulcanised rubber, that was accidentally putting together two different materials on a hotplate. So, he was trying to vulcanise rubber, but he had no idea he could achieve it in that way. And then there are others such as when you have a solution to a problem that hasn’t yet emerged. So, the discovery of shatterproof glass was by dropping a flask with a particular polymer mixture in it that made it shatterproof, and it was only a couple of years later that there was a realisation that having shatterproof glass in a car, for instance, could save a lot of lives.

Host: Benjamin Thompson

Well, we’ve got these four categories then. What do we do with them?

Reporter: Lizzie Gibney

So, having this exact framework was what this guy has been working on for a while. Now, the idea is that we actually also have data that we can apply that framework to. So, he’s going to get together loads of grants and find out about the patents and the publications that come out of them, and in a way that he didn’t really want to make public in case other people copy his exact method, he is going to essentially see how often it is the case that research in one area produced results in another, how significant it is when that happens, the size of the discovery and importance, and ultimately from this, try and just build up a picture of how often these different kinds of serendipity happen and see what we can learn from that.

Host: Benjamin Thompson

Well that’s definitely something that I’d like to talk about, but before we do, I mean, let’s just have a bit of fun with it. I’ve talked about Fleming there and you’ve given us a bunch of examples as well – I wonder do the rest of us have any examples of their kind of favourite examples of luck.

Reporter: Heidi Ledford

It’s terrible but my favourite is always Fleming, but the reason is because I think I give him a bit more credit than maybe other people tend to do because I think it’s entirely likely if I had seen those petri dishes I just would have said oh that’s disgusting and thrown them away. And I have times, you know, when I was a graduate student, I mean I feel like my thesis advisor was so great and he was always looking for these accidents and he would help us find them in our data. And there were times when I was so focused on getting a result under one condition that I wasn’t really paying attention to what was going on under other conditions and he would say woah, but look at that, and that would end up being my thesis project, so I find the question quite interesting.

Host: Benjamin Thompson

I mean what I was thinking as well, certainly with what you’ve said there Heidi, and Lizzie what you’ve covered there, is reminds me of Gary Player, the old golfer who – it’s probably apocryphal – but he said the more I practice the luckier I get, which I think, well it suggests that really, it’s graft and hard work that maybe puts you in the right place, and that that eureka moment isn’t really the thing. And is that perhaps even an unhelpful stereotype, that people just assume that someone sitting there, chewing on the end of their pencil, and suddenly they have that ‘ding’ and then science happens?

Reporter: Lizzie Gibney

Well, that’s it and I think often actually what it is, is the environment, all the different discoveries and research up until that point that make everything ripe for a discovery to happen. And yes there might be one person who has that particular lightbulb moment where they make this discovery, but as we saw with the most famous example is probably Darwin and Alfred Russel Wallace, they both came up with this theory of evolution at the same time even though it seemed to be so remarkable and counter-cultural and we think of them, that’s a great example because they were both fantastic scientists but working in a time that made that discovery possible, and I think that it often the case. It’s very hard to know whether it’s some attribute of a particular person that means they were likely to see something others didn’t or whether it was a big combination of factors.

Host: Benjamin Thompson

In the world we live in now, people want bang for their buck, right?

Reporter: Lizzie Gibney

Which seems fair enough – taxpayer money.

Host: Benjamin Thompson

Yeah, of course. I mean most research is publicly funded I would imagine, and so it seems only fair that you get results from it. But are we getting away from just being able to use luck at all – does everything have to be so mechanical?

Reporter: Lizzie Gibney

Well, I think one of the things that this study may come out with, which was certainly suggested from looking at a lot of these examples from history, is that often targeted research doesn’t have to be a bad thing, as long as you’re able to make some discovery in another field at the same time because in an awful lot of cases, that’s what was happening. They were looking for something and that was fair enough because they needed the next whatever bit of technology, but the situation around it was such that when this serendipitous thing dropped into their laps, they were able to exploit it and I think that’s the really important thing. So it’s not necessarily bad that we use funding for applied research and that we try and go in a particular direction, as long as it’s then possible for us to exploit the unexpected findings that happen while we’re doing it so that, you know, Heidi, when you were looking through that data and your supervisor said no, wait, look at this, you didn’t have to go oh god, but I’ve got to publish three papers by the end of my PhD or I won’t get my postdoc after that, you know, you don’t ignore those things when you find them because that would be the real problem.

Reporter: Heidi Ledford

I mean I like this idea of challenging the blue-skies notion and saying oh, you know, we need to be free to pursue our curiosity because we may turn up things that we didn’t expect and that’s where lots of great discoveries come from because, exactly as you said, as long as you’re still doing the research, you still have the chance to make these observations. But then I do wonder maybe you do still have to be willing to support something that seems like it’s quite basic in those initial discoveries because otherwise you’ve got to carry it a bit further.

Reporter: Lizzie Gibney

I think that’s it. I think there’s always going to be a balance, and I think this study is hoping to come up with maybe even just a little bit of evidence, however weak because at the moment there’s none, that says, for instance, where that balance should sit or if we find that the most important instances of serendipity happen in one type of completely open-ended research versus targeted research, that also would be something we should know. So, yeah, it starts here really.

Host: Benjamin Thompson

Okay then, well Backchat, as I understand it, is all about the process of reporting as well, and do you find that there’s an element of luck to what you do? Do you fit into one of those four boxes at all? Can you give us some examples or is it literally you go the hard yards and get the results?

Reporter: Lizzie Gibney

I think it’s like your golfer chap whose name I can’t remember. If you built up all of your networks and you chat to people enough, the best stories will probably come about through a random conversation at a conference or something, but you’ve got to be at the conference, you’ve got to know that person, you’ve got to ask them a few open-ended questions or you’ve got to build up their trust enough for them to want to tell you, so it’s all about the situation you establish really.

Reporter: Davide Castelvecchi

And yet when I do go to conferences and a lot of times the best things I get out of them are not the ones that I expected. I went to a conference more than ten years ago, which is more like a popularising science kind of conference, where journalists usually don’t expect to find news or things to report, and that’s where I first learnt that there was this thing called the quantum internet, or that people were thinking about it, and this may be relevant to the later part of our conversation.

Host: Benjamin Thompson

Oh, absolutely and listeners, more on that coming up. But before that, a bit of a debate has grown about what the first flower looked like. Heidi, this is something that you’ve covered for Nature – could you tell us a little bit about the opening rounds of this discussion?

Reporter: Heidi Ledford

Well, first of all, it’s a long-standing question, right. This is something that people have tried to answer through analysing fossils but the fossil record is quite scarce for flowers – they’re so soft and delicate. But a few years ago, some people decided to take more of a big data approach to it and they assembled a team together to characterise the traits in a lot of different species of plants, and characterise their flowers. So, they analysed over 20 traits in about 800 different species. They formed this big database and then they tried to match that up with some of the molecular data that we have on how these different species are related to each other evolutionarily. From that data, they tried to crank out what would be the first flower, what features it would have.

Host: Benjamin ThompsonAnd what did they sort of posit that it might look like?

Reporter: Heidi Ledford

Oh, you know, a lot of the things that they were looking at are things that we don’t typically look at when we’re looking at a rose, for example. So, they didn’t have data on, for example, how big the petals were or what colour they were, but they had some data on the arrangement of the petals. They looked at the arrangement of the male and female reproductive parts, are they on the same flower was one. And they also looked at things like how many different reproductive parts did they have, how many different female reproductive parts did they have, how many male, and so forth.

Host: Benjamin Thompson

Well, recently then though, Heidi, some researchers have come back and said well, hang on a minute.

Reporter: Heidi Ledford

Yeah, I mean it’s a neat story I think because you had this team working together and they had this sophisticated modelling they were using, and they come up with image of a flower and they did actually have a sculptor put together a model and then they scanned that model and made that an image that they put out with the paper. But despite all of this analysis there was this plant morphologist who looked at that paper and he looked at that flower and he said I don’t think so because he had looked at so many different flowers and he understood the morphology so well and he’s like yeah, I don’t think that kind of geometric combination of parts really exists in nature and I’m not sure that it ever could, and so that was the debate about it.

Host: Benjamin Thompson

You mention there the lack of historical record for flowers. Does this debate get us any closer to what it might have looked like?

Reporter: Heidi Ledford

I mean I think so. I think they’re still working to expand their database and to continue their analysis. The debate really is in the form of a commentary that was published in the American Journal of Botany and the person who received it and who was working on this project, this e-flower project it’s called, he said that when he first got that letter he was just so nervous, oh no what is this going to be, but when he opened it up and he read through it he thought oh no, okay yeah this makes sense. And so, he went back and he looked back at some of the data that they had that suggested that you could have this particular combination of features in the male and female reproductive parts of a flower and he realised that maybe he had mis-scored a couple of the flowers that they had in there. So, he didn’t 100% agree with the criticisms that were lobbied about that work, but some of them he did think looking back at it maybe they’re right. And he redid the calculations and he came up with an answer that was I guess closer to what people might have expected to see, which was that these male and female reproductive parts shared a similar geometry around the central axis, and that was what was sort of triggering the debate.

Host: Benjamin ThompsonWell this seems I guess like quite a fruitful discussion. Listeners, I apologise in advance, that was a mistake. Well, this is backwards and forwards and they seem to be getting on relatively well, but I think we can all say there are examples where we get two groups of very entrenched views, you know I say x and I say y and we’ll never kind of meet in the middle, how do we get around the he said, she said arguments?

Reporter: Heidi Ledford

I think really what you have to do is just present the argument to the readers and then present some of the evidence on either side and let them decide which side they’re going to be on. And you try go on beyond - if you’ve got a he said, she said - you try to go beyond he and she and gather others, but you do often find cases where you’ve got teams entrenched on one side or the other and it can be difficult to sort out.

Reporter: Lizzie Gibney

And the problem is the things they disagree on are important but really boring. So, like they go really, really into depth and it’s like each little nitty gritty point, and you just think I can’t put that in a story because nobody is going to delve that far. The overall conclusions they disagree on, fine, but the whys and wheres of it can be very hard to actually weave in to a story, and then so it’s often, as you say, about getting people completely independently talking about where the community is right now, and maybe sometimes that is the story as well. You don’t have to come down on either side, but the story is debate, confusion, ongoing discussion about this hot topic.

Reporter: Davide Castelvecchi

I do agree with Heidi. In my experience, it vastly depends on the situation and in some cases, there is like maybe one group that is obviously promoting a bizarre and less credible view of things, but in a lot of scientific debates we need more data and good people can disagree.

Host: Benjamin Thompson

Well on that sort of we need more data point then of course science is super iterative and each time we go round, we have another set of results. What duty do we have to maybe keep up with the debate? If we stop covering something, does that mean if I were to Google it the last answer is the correct answer? How do we keep up with it and do we need to keep up with these things?

Reporter: Lizzie Gibney

I think it’s hard because if you start covering a story to then just stop if then it evolves a long way after that, it does feel a bit irresponsible, but it’s also hard because readers get tired, editors get tired of a particular story, you often get just one bite of the apple so you want it to be at the right time, so sometimes you kind of store it up. So if you last wrote a story a couple of years ago, and there’s been a lot of things going on in the interim, you might then wait a year or two and do your follow up.

Reporter: Heidi Ledford

I think it’s fair to say we spend a lot of time talking about when is the right time to weigh back in on this particular topic.

Reporter: Davide Castelvecchi

What helps, in my opinion, is that wherever we stop, whatever is the latest episode we’ve covered, to always kind of give sense that science never says that the last word on anything, and that there’s always, to the best of our knowledge, maybe we’ve come to this conclusion but who knows.

Host: Benjamin Thompson

I mean do you think that the – listeners, you can’t see but I’m doing inverted air quotes here – the importance of a subject makes inferences on how often we cover it, like if it was a vaccine story would we be more likely to cover it than the flower story, no disrespect to the researchers involved.

Reporter: Heidi Ledford

Yeah.

Reporter: Lizzie Gibney

Yeah, although what defines important depends on who you ask. So, a story we covered blow by blow was when the BICEP-2 experiment thought that it had seen gravitational waves in 2014, and there was immediately a big press conference about it, it was huge news, these kind of ripples from the early universe, but it soon emerged it was challenged by lots of different people and we covered that as it went along every week or so for a little while perhaps. But it was a really huge scientific story.

Host: Benjamin Thompson

Well, finally then in this episode of Backchat, let’s talk about the internet, but perhaps not the internet as we know it. Davide, you’ve written a feature very recently about the quantum internet. Could you give us a little overview about what it is?

Reporter: Davide Castelvecchi

Well first of all, it isn’t yet. It’s still a hypothetical thing that people are working on. They’re working on designing it, they’re working on demonstrating the physical principles of it, but the main idea is that there’s this notion of quantum information which is more complex than normal classical bits that are in normal classical computers, and it’s very hard to transport it, to download it, to send it anywhere, but that’s what the quantum internet would do.

Host: Benjamin Thompson

I mean you’ve been reporting on this now for quite a while. I mean you first wrote about it in 2008.

Reporter: Davide Castelvecchi

Yeah, it was in a previous life, in a previous job. So, I wrote about it first in 2008, and people were very optimistic that we would have the first working quantum internet device within a few years, and here we are ten years later and we still don’t have it, but people are still confident that they will have them within a few years.

Host: Benjamin Thompson

Well actually Davide, I happen to have the quote here that I think you’re talking about, and this is from Mikhail Lukin of Harvard University and this is from your 2008 piece: “I’m optimistic that within a few years we’ll be able to build at least a lab demonstration of a quantum network.” And this is something that I wanted to talk to you about. We’ve all seen it, I’m sure, in ten years’ scientists hope to have done x, which I find that an awful tease, and it’s so rare that you get that follow up in ten years, but you’ve actually been able to do it, right. You started in 2008, it’s 2018 now. How has the story evolved?

Reporter: Davide Castelvecchi

Well, they haven’t quite done it but the progress has been really substantial, maybe slower than people expected. Now the word from the Dutch scientists that I talked to is that they are hoping to have a working demonstration by the year 2020, connecting four cities in the Netherlands. This could turn out to be – if it works out well – kind of the quantum version of the famous Arpanet, which was kind of the progenitor of the classical internet, which also connecting four cities in the western United States in the 1960s.

Host: Benjamin Thompson

So that’s another two years away then I guess, so the needle has moved further way again, but for me sort of as a regular punter, what can the quantum internet maybe do for me?

Reporter: Davide Castelvecchi

It’s very controversial whether it will have consumer applications with one important exception which is privacy. The one thing we know about quantum physics when it comes to communications is that you can use quantum physics to communicate data securely. Then there are the more science fiction things and many have been proposed like networking quantum computers together, also quantum computers are things that are a few years away and always will be maybe, but to make them communicate you will need a way to transfer quantum information.

Host: Benjamin Thompson

What do we have to do to make sure that we don’t get overexcited about this? How do we avoid getting caught on the hype train and getting swept up in the excitement?

Reporter: Davide Castelvecchi

So, as a reporter who’s been around the block a few times by now, I think experience helps and it also helps to see how people have covered very promising, very exciting research in the past, and to see how that has stood the test of time. So that way you can see if we can avoid falling into the same traps as people have in the past, or you can see maybe there’s people who have done a very good job and who’s coverage has stood the test of time and you can try and learn from them,

Reporter: Lizzie Gibney

I think in this case, so in physics we do get asked quite a lot of the time, but okay but what’s it for, what’s it going to do, what’s the why behind it, and you do have to come up with some kind of example. And I think it’s a classic case with Davide’s story, there are loads of very speculative, great ideas for what you might use a quantum internet for. Are they going to be actually what we use it for? Probably not. But at the moment we don’t even have this thing to play around with, but once that exists, then somebody else, coming back to serendipity, is probably going to say oh hey like we’ve now got this bit of kit, this is what I could now do, this is something I’ve been trying to do from a completely separate field for ages and it’s now been enabled. So, you just don’t know what the future holds and therefore you kind of have to grasp at straws a little bit. And obviously, the way to do it is to say these are great ideas, these are possible, but not promise too much.

Host: Benjamin Thompson

Well then everyone, there you have it. Many thanks to my guests Davide Castelvecchi, Heidi Ledford and Lizzie Gibney for joining me today. You can read their work and more stories from the world of science over at nature.com/news. If you’d like to get in contact with us you can reach us on Twitter – we’re @NaturePodcast – or on email – podcast@nature.com. This has been Backchat. I’ve been Benjamin Thompson. Thanks for listening, see you next time.