Nature's Take: How Twitter's changes could affect science

27th February 2023

Host: Nick Petrić Howe

Welcome to Nature's Take. This is the show where we dive deep into the stories that matter in science. In each episode, we pull in some of Nature's finest, present them with a topic, and see where the discussion leads us. In this episode, we're talking Twitter. This social media platform has over the decades become indispensable to many scientists. It's a place to share findings, engage the public, it’s a source of data and it has given a voice to people who might otherwise be excluded. It's not all roses, though. Twitter has also helped foster the spread of misinformation, abuse and harassment. And in recent months, it's fair to say that Twitter has been going through a bit of a tumultuous time. Since Elon Musk took over the platform, many Twitter employees have been laid off, there have been all sorts of malfunctions on the site, and some potential policy changes have concerned scientists. To discuss the fallout of this as it pertains to science, I'm joined today by three very knowledgeable guests from across Nature. First up, we've got Anne Marie Conlon, Chief Editor, Audience and Engagement.

Anne Marie Conlon

Hi, I'm Anne Marie, I manage the @Nature Twitter feed.

Host: Nick Petrić Howe

Next, Mary Elizabeth Sutherland, senior manuscript editor for cognitive neuroscience, behavioral and social science.

Mary Elizabeth Sutherland

Hi, I'm delighted to be here. As Nick said, I'm the manuscript editor for most of the papers that have to do with people actually doing things at Nature.

Host: Nick Petrić Howe

And Michael White, Senior manuscript editor for climate science.

Michael White

I handle anything to do with atmospheric sciences, oceanography, and the cryosphere, and I've been an intermittently heavy Twitter user, to being currently, almost a non-Twitter user.

Host: Nick Petrić Howe

So just to start us off in this discussion, maybe it's useful to talk about how each of you actually interact with Twitter. Anne Marie, why don’t you start seeing as you're Nature's Twitter person.

Anne Marie Conlon

Part of my role is to manage the Nature social media feeds, and that includes our Twitter feed, which has a few million followers. In a personal capacity, I use Twitter compulsively, because I have my own networks. You might notice my accent that I'm Irish, so I have like Irish Twitter, I have media Twitter, I have just random nerd things Twitter. And, yeah, it's a way it's a way of connecting with people outside of my own small circle in, so I yeah, I use Twitter an awful lot.

Michael White

I think I've used Twitter for a couple of main reasons. One of them is to find scientists who are also interested in baking, that's been a big Twitter outreach effort on my part, but I more substantially, I think what I've tried to do is to take what I've learned from editing for the past 15 years, and to make that available to the community in a way that might be useful. So, you know, after you see thousands of papers coming through a review, you kind of get a feeling for what reviewers object to, so I've tried to have long threads that are sort of highlighting the things that I think would maximize a paper’s chance of getting through review. And then I've tried to promote content at Nature that I think might not have been immediately seen by the community, so I tweeted extensively about our recent racism in science special issue. So it's been kind of more for me, like a service role, trying to facilitate the community's interaction with Nature and trying to bring some of our higher level thematic content back to the community that I worked with.

Mary Elizabeth Sutherland

So, I feel like I am Mike in training. I, I'm, basically my role was created to increase our content in the behavioral and social sciences and cognitive neuroscience, so one of the ways that I've been using Twitter a lot is for outreach. I've been engaged in just letting people know, you know, saying, hey, great paper, we're interested in this, or tweeting things that I've seen published elsewhere, saying, this is a great paper, you know, consider and then tagging the people and say, you know, consider submitting this type of research to Nature. So that's my personal use of it, and then as the editor in the behavioral and social sciences, I see a lot of papers that are based on Twitter data, so a lot of questions about polarization, about misinformation about the spread of information, you name it, everybody talks about things on Twitter. So, it's been a great source of data just in the research that I read on a daily basis.

Host: Nick Petrić Howe

So, you will clearly have your uses for Twitter, and lots of scientists do as well, but not every scientist is on Twitter. So, I'm interested in whether you think it is actually that important as a tool?

Michael White

I mean, I think it can be a really important tool for a wide variety of people. Like in my domain right now, there's a ice drilling campaign to get ancient ice from Antarctica, and that drilling team has been posting to not great numbers of views, but it's been really fascinating to kind of just watch their progress and watching things break down in the field and the repairs that are going on, so I love it for that kind of programmatic update in particular.

Anne Marie Conlon

It's kind of where the conversation is for a lot of scientists. Like same as a lot of individuals use it for their own communities with work or hobbies. There's a community, a strong community of working scientists that follow Nature and engage with Nature and that I see every day talking to each other, talking to us just about what's going on in their industries and seeing what's happening in different pockets of science that they might not know about if it weren't for seeing other communities on Twitter.

Michael White

I also feel like it's become a way to get out feelings about mental health in academia. I feel like when I left academia back in 2008 to come to Nature, I was sort of viewed as being completely insane for leaving a tenured job, but now I feel like it's happening just routinely, and that people are openly talking about their mental health struggles on Twitter and why they have left to go to industry, or to, in my area, a huge universe of climate startups, I think that's been a really positive development.

Host: Nick Petrić Howe

Well, Mike, I'm quite interested as well, you said you were a heavy user of Twitter, and now you're not so much. So what's sort of changed for you?

Michael White

I mean, I think after the change in ownership, huge numbers of people in my area, left Twitter to go to Mastodon, and the feed changed enormously. So I tend to get things showing up in my feed from just a few people repeatedly. And all the people I used to follow and be interested in, large numbers of them are gone. And then just the junk ads that are polluting my feed have become kind of epidemic. So it's become much less interesting and much less, sort of, science focused in my area, and more like a marketing tool for things I don't care about.

Anne Marie Conlon

The ads and their replies is a new feature that is definitely frustrating to a lot of users from what I've seen, and frustrates me, if you're just normally you would have an ad in your feed, but now you get it in a reply to a tweet. So, it could be like person talking about serious mental health issues, and then ad for something nonsensical, whimsical and silly —

Michael White

Yeah, completely!

Anne Marie Conlon

— just interrupting the kind of serious tone, or that's one thing that has started to bother me as well.

Michael White

You know, Mary Elizabeth, you were talking about the use of Twitter for research purposes, and one thing I've always wondered about, like, I've heard talks where people are using Twitter to understand people's attitudes about extreme weather or heat waves or things like this, and I know that probably historically, that's been done through like interviews and surveys. In your work, I mean, do you see that you get really radically different answers from Twitter than you would from more traditional social science data gathering approaches?

Mary Elizabeth Sutherland

That's a really great question. You often don't get very radically different answers, if you're looking at a question that has been asked before, but what you do get as you get the ability to ask bigger and different questions than you could have asked in the same way, because one of the problems that comes with a survey, in a way, is if let's say, you know, I come to you and I say, I'm just going to pick a university, there's no affiliation between Nature and this and I just say, hey, you know, I'm Mary Elizabeth Sutherland. I'm a tenured professor at Columbia University in New York City, and I want to know what you think about this. There's some response bias that comes in, right —

Michael White

Yeah, right.

Mary Elizabeth Sutherland

— and so people do do survey designs that take this into account, but it's difficult. What's really cool about Twitter, is people are just choosing to share this information, right? People are like, this is what I think —

Michael White

Yeah.

Mary Elizabeth Sutherland

— there's nobody prompting me this is just my own behavior. So in a way, you're collecting spontaneous behavior, and you can get to different populations, because as we were saying, the nice thing about Twitter is that anybody can join, and so what's been great in my field is that I've started seeing not just English language analyses of Twitter data, but also in other languages. So you can say, Okay, let's go look at something else, you know, does this work in a different language —

Michael White

Oh wow.

Mary Elizabeth Sutherland

— and you get these really cool converging or not, findings across different cultures that are just pure spontaneous behavior, which I think is really valuable. So that's another topic, right? The upcoming policies that we're not sure for how accessible those data will be, is definitely something that is difficult for my community, because it really has been such a resource.

Host: Nick Petrić Howe

So, this is in reference to potential policy changes, because it's still unclear exactly what's happening here, where scientists may have to pay $100 a month, maybe more to access all of this data. People might have actually heard about this, this is something called Twitter's Application Programming Interface or its API, and that's how people can access this data. If this were to happen, do we have a sense of what the impact will be? You've mentioned how useful this is, as a way to get the sort of sense of what people are thinking. If this was to change, how would scientists react?

Mary Elizabeth Sutherland

I mean, everyone is worried about it. I know a lot of the authors I work with have just been trying to get as much data as they can in advance, because we don't know what the policy is and it's really difficult if that's your research, that's your data source, and somebody says, access to this may or may not change in a way that you can or cannot afford, right? It's a great degree of uncertainty, but I think that it would be very damaging if it made it difficult for many people to access it because something that's cool about these data is that you can, well, they were accessible so you could address interesting research questions without a huge budget. So, it sort of equalized things a bit more right across the —

Michael White

Right.

Mary Elizabeth Sutherland

— the haves and have nots.

Anne Marie Conlon

Yeah.

Mary Elizabeth Sutherland

So it just sort of polarizing is the wrong word, but it increases the disparity?

Michael White

Well, I mean, it kind of will foster helicopter science remotely, essentially, where you'll have people who will be able to do research on parts of the world that can't afford to pay for the access, but who aren't actually living there, like in my area, you can always imagine that, a really interesting way to use Twitter data would be for extreme weather events, and trying to understand how communication and response efforts are actually being taken out by the people who were affected, which could be measured via Twitter data, but in many parts of the world where these extreme events are gonna have the biggest impact few people there will be able to pay for access, I imagine.

Mary Elizabeth Sutherland

What worries me, and this is full speculation, but then once you start getting this pay for access for the data is it also going to be, you know, paying for the different things that you can do, and is that also then going to skew who is on the platform? I mean, already, Mike, you've talked about Elon’s takeover has skewed who from your community is still on Twitter, and I'm sure the same is true for all of our different communities. So, I feel like that is also changing, right? If you look at the data, you say, okay, people have just joined, but then if there's a major exogenous event, like somebody taking over, and then you have people's own decisions of what to do, that also impacts the data.

Host: Nick Petrić Howe

And one thing this is sort of like associated with, we’ve talked about sort of inequality, is Twitter has been sort of a way for many people to have a voice who wouldn't have a voice and have sort of engagement with their scientific peers when they may find it difficult; maybe they're a PhD student, maybe they're from certain communities that have been overlooked in the past. So, do think, and you know, Mike, you've touched on this a little bit, with sort of changes to Twitter we're going to see people sort of losing the voice that they've fought so hard to gain on this platform?

Michael White

I think it's certainly possible. I think after George Floyd and related social movements, there were groups on Twitter that came up in my area like Black in Geoscience, there just are strikingly few black geoscientists in tenure track positions. And so Black in Geoscience was a, I think, a great group that sort of self-organized, and it led me to discover many people I thought were not at all on my radar, and then there's GeoLatinas, for example. So, groups that are less representative in geosciences have found Twitter to be, I think, a place to both kind of collect people within those communities and also to raise the visibility of those scientists.

Anne Marie Conlon

Well, anecdotally, from my being very online, it does feel like some people have left Twitter because they felt like the environment created by Musk's takeover was more hostile to them than it had been previously. I think, actually, it's not just anecdotal, there was some studies in the months following the takeover that said that hate speech and spam and things have increased like a significant percentage more hate speech against marginalized groups, people, and obviously, we can tell Elon Musk's views on all of this, he's been very vocal on it. But then there's a lot of people, definitely in the communities that I follow and in Nature’s communities that don't agree with that. So, people are leaving Twitter and people that we want to hear from are leaving Twitter. But, at the same time, not as many people have left as, as was expected by the kerfuffle when the takeover happened and, and I know Mike was talking about Mastodon, the move to Mastodon felt like everyone was leaving all of a sudden, but now, using Twitter every day, as I do, I don't I still see our community as a very vibrant place with full of discussion and science community, and even though a lot of people will say I’m on Mastodon, they're still engaging. So, I don't know, it's hard to say exactly what the impact will be, who's been lost, and who remains, you know, but I suppose time will tell.

Mary Elizabeth Sutherland

I've had researchers describe it to me as a virtual conference in a way, so you know, when you go to a conference, you're exposed to things that you are interested in, but also a lot of things that you didn't know, because you know, anybody from that area who comes is there, and also you're exposed to all of the things right? It's not just a paper that somebody's written, it's also potentially a resource, so somebody will say, oh, cool, look, look at this cool new technique, you know, that can be used for this, so you start understanding what different techniques can be, you can connect with people, you have the potential for jobs to come up, you can ask people questions, write polls, like, hey, what would you do in this situation? So it's just there's a, it makes it really easy for people, I think, to connect one with people that they might not know that are within their network to share information, and to share information more than just, you know, a publishing table of contents because you can understand, sort of, you can get the pulse of the field much more easily.

Michael White

Yeah, for sure.

Mary Elizabeth Sutherland

Yeah. And I just feel like I myself, don't use Mastodon, but it seems like there just isn't another really equivalent way of doing this kind of, all of these things, right? LinkedIn is like a professional network. But it doesn't give you all of the other things, there just isn't another platform that appears to have this easy virtual conference style, right? And so, I feel like, though people may be against it in some way, it's, at least I see as well that lots of people are still there, because there's just this, just, there's so much to gain from it.

Host: Nick Petrić Howe

Another thing I wanted to sort of touch on as well, that we've kind of mentioned, I mean, some scientists have left Twitter, but it has been a really useful place to share information, but it's also been a place for a lot of disinformation, too, and a lot of the employees at Twitter who are responsible for dealing with that sort of content and moderating it, are now gone. So, what do you think is gonna happen with the sort of rise of Mis and disinformation?

Mary Elizabeth Sutherland

That's a tough question.

Michael White

It's gonna get worse!

Anne Marie

Yeah, bad things!

Mary Elizabeth Sutherland

Yeah, it's gonna get worse, but my hope is that AI is potentially going to be able to help, right? I mean, the other thing that we have on the rise is large language models, right, with the, I guess the most popularized is ChatGPT, but the ability of AI algorithms to do things that humans could do, but very quickly, right? So, to sift through information and flag like yes or no, right? Does this fit with what we know or what, what other information is there another there's, of course, a lot of room for things to go wrong, but I'm hoping that there's a way to start integrating some of these large language models into these online forums to hopefully give some indication as to whether or not something would be true or not, are some prompts.

Michael White

I, I agree with you on the potential, but, but let me just give you like a couple of examples in climate science. So, the sort of climate denialism has always been active on Twitter, so for example, climate variability is a normal part of climate, or you could say that most of the greenhouse effect comes from water vapor and not carbon dioxide, both of which are true, but both of which are absolutely not the main story that we're talking about now. It's like, how would AI, I don't know how AI would deal with that, because they're not factually incorrect.

Anne Marie Conlon

And these nuances that AI would miss, they're only as good as the information they're fed, so if they're scrolling, the, you know, a language database or a fact database for facts like Michael said, they could collect the wrong fact and there’s too much room for nuance and misinterpretation, I think, but I hope you're right Mary Elizabeth, I'd love that, I would, but my, my, kind of, my thinking is always just that, this is my like, journalism hat on, is that like media literacy is the kind of biggest casualty of, of the internet, and this is what perpetuates a lot of misinformation online, because people just don't check their facts at all or look, look at what a reputable source is. And like as someone who is promoting content from a trusted publisher, I get so frustrated when I see someone just citing a fact that sounds like it could be good and then in a few clicks, like, oh, no, this is totally nonsense. I feel like there's like a multi-pronged attack against misinformation, and one of it is media literacy, but that has worked in conjunction with some type of moderation of content online.

Mary Elizabeth Sutherland

We published a paper March 2021, showing how shifting people actually using Twitter, using Twitter as a field experiment, speaking of things that one can do with Twitter, showing that shifting people's attention to accuracy can actually reduce the spread of misinformation, which is really nice. So that's, that's actually another good point. So, Twitter is also used, not just for collecting observational data, but you can also do these types of studies interacting with users and see if a prompt makes a difference, and it does, when you get people to actually think about the accuracy than they don't just share willy nilly.

Host: Nick Petrić Howe

And in the context of the rise of misinformation and disinformation on the platform, is it maybe not a requirement, but a good thing for scientists to be there to sort of combat this misinformation?

Anne Marie Conlon

I mean, I think it's really important for, for Nature, as a publication to be a voice for reason and science on the internet, and that's my whole job, basically, and on Twitter, we wrote a lot about COVID, at the very start of the pandemic, and we had incredible traffic to articles because people are desperate for information from a trustworthy source, and Nature because it's a known trustworthy source, people could go to us to get that. And then, part of that is sharing that content on Twitter, and then also having that content re-shared by other scientists, or also addressing things that scientists are asking on Twitter. I think it's important from that point of view. If we all left all of the scientists, we’d be worried what would happen to the people who remain who aren't as informed or wouldn’t get their information from a trustworthy source.

Mary Elizabeth Sutherland

I mean, I was drawn to career in editorial rather than research because I really feel that it's important to communicate science as best as possible, and that's something that you can definitely do as a researcher, but it's really great to be able to shape a broader sense of that as an editor. And so it feels like, you know, that's part of my, my own personal mission statement in what I do is having as accurate a scientific record as possible, and transmitting that to as many people in the world as possible, right, doing as much as I can to help good science get out into the world and be used properly in the decision making, by properly I just mean, being used with full understanding.

Michael White

I think those are all great points, Mary Elizabeth, and I think that I totally agree with Anne Marie in the sense that, like institutions, I feel have a larger responsibility than individual scientists, because for individual scientists in my area in climate, trying to beat back the tide of misinformation would be a full-time job. So it's just orders of magnitude more work, to combat misinformation than it is to spread information and to expect that there's going to be like a responsibility of individual scientists to engage on Twitter to fight against that, I think is fairly unrealistic. I think if there are people who wants to do that, and there certainly are people who want to do that, that's fantastic. And they should be supported, but I think it's, it's probably more appropriate to view institutions like Nature as a tool for doing that.

Host: Nick Petrić Howe

And so what do you feel is Nature's role in this space? Should we be doing more to combat misinformation?

Michael White

I mean, it's always going to be a question of resourcing, right? So, you have to have more people like Anne Marie or in Anne Marie’s division, who have the necessary domain knowledge and maybe the ability to interface with the Nature editors who handle those papers to try and push back on misinformation. So it'd be, you know, it's like a fairly big expansion of our effort to have that sort of tracking down misinformation about our specific work or work within the field. So I think in principle, yes, in practice, it's a fairly big ask.

Anne Marie Conlon

Yeah, we have almost 2.5 million followers on Twitter, and not every person that shares something from us gets seen by hundreds of thousands, or who knows how many other people, so you just can't catch every single thing. From a content point of view, we're very aware of our role of making sure that nothing that we put out there could be misinterpreted, so we're very careful on our headlines, and the pictures and everything that we use that will go on Twitter. To make sure that it's not going to be easily, like for example, we’re forever writing about vaccines, we're very careful that the headline is not something that a vaccine denier, or someone that thinks vaccines don't work could misconstrue and re-promote, to promote their agenda, because we want the facts to stand on their own, so we're very careful in that area. So that's, in my mind Nature's role on Twitter.

Host: Nick Petrić Howe

And so, I'm gonna end with the sort of broad question for you. Do you think that science as an institution needs like an open platform such as Twitter in order to operate well?

Michael White

I do. I'm just going to give you an example that's been happening over maybe the past decade, which is discussions about academic misbehavior. So, it's happened a lot in climate science, but in the past year, there's been this eruption of discussion in economics. So, economics is a field that’s historically been dominated by men, and it's been viewed as being hostile to women, and so there have been a lot of issues that have been discussed through Title IX, or through university human resources departments that have gone from the perspective of many women in economics nowhere. And so like, fairly recently, a very public campaign began on Twitter naming individual male economists who were accused of inappropriate behavior in many ways. And it just opened up this gigantic storm of controversy about the appropriateness of doing this, the consequences they might face. And so I think that, you know, Twitter is a way to try to make academia more like what it should be, and I'm not saying that these accusations are going to be borne out in any way or particularly, but it's an important way of having this discussion, you know, people have been removed in academia in my area, because of discussions that were initially disclosed on Twitter. So, I think for that one particular reason, it's something like this is really critical, and wasn't being accomplished through traditional academic enforcement mechanisms.

Mary Elizabeth Sutherland

I think we're human, and we need to communicate. I mean, that's, that's what we're doing here in this podcast, and that's what so much of our lives are about, and with research, especially all of our authors, all of the people we engage with, they devote their lives to this particular subject, and they get so deep and so into it, and it's so hard, then to communicate with everybody else, you know, unless you live with somebody who does the same thing. You're not able to and so the ability to have a platform where you can continually communicate about the things that you spend your day doing I think is also very valuable. I know that you can do that, you know, within a department and not but this idea to be able to connect to people who really understand what you're doing and share your excitement for it, I think is also just really good for our mental wellbeing as a society and for researchers, so not just speaking about mental health, which is really important, but that it actually helps scientists’ own satisfaction and purpose to be able to share with people who really understand, which can be hard when you're such an expert, and the other expert, you know, is a continent over.

Anne Marie Conlon

I think, from a science journalism point of view, Twitter is quite, I don't know if I'd say essential, but extremely useful. Because in that conference, Time Square kind of capacity, it's what people are talking about, and the journalists job is to kind of know what people are talking about, and to give answers to the questions people are asking. So, it's part of, here's what the scientists I talk to today, my contacts are saying, here's what's happening on the news, here's what's happening on Twitter, here's what's happening. It's one strong source that is many sources all together that can help you kind of get a finger on the pulse.

Host: Nick Petrić Howe

Yeah, right. Well, I'm hoping that this as well is going to be a useful way for people to get their finger on the pulse on how recent changes to Twitter are sort of affecting science. But I think that's all we've got time for this time on Nature’s Take. We'll be back with another episode in the near future. But for now, all that's left to do is to thank you all once again for joining me.

Michael White

Thanks Nick!

Anne Marie Conlon

Thank you.

Mary Elizabeth Sutherland

Thank you so much Nick, this was great.