He believes that there are major economic and political interests that fuel the denial of science. In his opinion, we should focus not only on the symptoms but also on the deeper underlying structures that allow denialism to thrive. To maintain trust in science, scientists should actively address issues like science publications that become more focused on ranking and commercialization rather than the content of the research or industry-sponsored research.
ChatGPT, the Effect on Teaching and Science
When talking about ChatGPT, he advocates soul searching about what is the purpose of the texts we produce in science, asking ourselves, why do we produce it in the way that we produce it? Why is it valuable to write, understand, read, to critically reflect? And what is it that the machine can do and what is it that it can’t do?
(…) we should separate the issue of research and teaching. So first and foremost, universities and institutes of higher education that are involved in teaching, they will face an issue of students handing in AI-improved texts and materials. And there’s a real soul searching about what is the purpose of the texts we produce in science if we have such tools. I think that’s actually a good debate. We should be asking ourselves, why do we produce it in the way that we produce it? Why is it valuable to write, understand, to read, to critically reflect? And what is it that the machine can do and what is it that it can’t do? So that’s a good discussion. But that’s going to hit scientists more from the teaching side. From the research side, there have been some studies done that at least if you give people abstracts and small summaries of papers, then you can fool some people even in their fields of expertise. But then it becomes pretty clear quite soon onwards that there is no real knowledge behind it, it’s just the prediction of what would be probably a good paper. So I do see, however, the speed at which these “stochastic parrots”, as they have been called, produce any type of output as potentially dangerous. So put simply, the speed at which fake science can be produced is enhancing, and if this speed comes to a system that prioritizes people having lots of output so people get rewards for writing lots of papers, here you have a tool that helps people write lots of papers. You put these two dynamics together, that’s a recipe for a lot of problems.
About Science Papers behind Publishing companies’paywalls
About paywalls for scientific publications, he stresses the need for everyone to participate and contribute to the global pursuit of science and not solely rely on Sci-Hub and the work of people like Alexandra Elbakyan, who founded it and who did a lot of good for science while doing illegal things.
(…) there’s a huge mismatch between the science that we need and we need to read accessibly and we also we need to publish it accessibly. It’s not just about reading, it’s also usually authors have to pay in order to get their articles printed, even in the most prestigious journal especially that’s a problem if you’re from the Global South or if you’re from an institution generally that doesn’t have funds to publish research. So generally, of course, knowledge should be free, or at least it should be possible for everyone to participate and contribute to the global pursuit of science. But that’s not possible, and we have to rely on these illegal sites like Sci-Hub and for instance, have to rely on the work of people like Alexandra Elbakyan, who founded it and who did a lot of good for science while doing illegal things, right? And that says more about how our global system of science is problematic and less about Sci-Hub or these illegal sites.
Extended Peer Review of Science papers needed
(…) I’m more a fan of this concept of “extended peer community” that comes from the post-normal science theorists. Basically what it means is that we have problems today from digitalisation to climate change that even transcends our traditional research methods, because decisions are urgent, values are being debated, worldviews conflict, even traditional research methods are not enough. So we need broader communities, we need to increase the size of who counts as a relevant person of critique, of knowledge, involve all stakeholders, involve different people who hold different world views, and involve a broader community than just the scientific peer community. (…)
Listen to the podcast Interview with subtitles on the TrustTalk YouTube channel
Transcript of the Interview
Like always we provide a transcript of the interview on the TrustTalk podcast. If you appreciate this and would like us to continue doing that, we would welcome your donation (see below for the link). Any donation, no matter the size, helps us continue our effort to bring you meaningful contributions on trust.
Help TrustTalk make more podcasts on trust, a small donation helps us a lot
TrustTalk is a non-profit project. Making a podcast requires sizable amounts to produce. To continue delivering the podcasts, we would very much welcome your help. Your donation helps a lot. Thank you very much for your support!
