Lecture 'How to trust science? A realistic assessment'
- For whom
- Alumni , Employees , Private individuals , Students
- When
- 23-04-2025 from 14:00 to 16:00
- Where
- Auditorium 1 Jan Broeckx, Blandijnberg 2, 9000 Ghent
- Language
- English
- Organizer
- Department of Electronics and Information Systems - Faculty of Engineering and Architecture
- Contact
- tijl.debie@ugent.be
The need for a more realistic account of the trustworthiness of science.
Traditional defences of the trustworthiness and reliability of science involve an account of how various gatekeeping mechanisms act in concert to keep science focused on finding the truth. They include motivations of researchers to pursue the truth disinterestedly, rigorous use of scientific methods, peer review, and self-correction.
In this talk, I will review recent direct and indirect evidence from meta-science and studies on research integrity, which suggest that, in large swathes of contemporary science, these gatekeeping mechanisms are not functioning as they should – and that attempts to fix this have not (yet) had major impact
. I will also consider how the emergence of publicly available generative AI-tools is likely to make the situation even worse. The upshot of all this is that we need a more realistic account of the trustworthiness of science.