BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:AI\, automation\, and the problem of error in science - Stephan Gu
 ttinger (University of Exeter)
DTSTART:20260226T153000Z
DTEND:20260226T170000Z
UID:TALK244123@talks.cam.ac.uk
CONTACT:Ahmad Elabbar
DESCRIPTION:Science is rapidly moving towards a more automated future. Whi
 lst this is not an entirely new phenomenon – automation in science has b
 een around for at least 150 years – recent years have seen an increase i
 n talk about\, and implementation of\, automated workflows. This drive for
  more automation is particularly prominent in laboratory-based disciplines
  such as material sciences\, chemistry\, and the life sciences. The core a
 ims that drive this push for more automation in the laboratory are 1) to i
 ncrease research efficiency and 2) to improve the replicability of experim
 ental outputs. In this talk I will analyse how these two aims are affected
  by the adoption of powerful new AI tools in science\, with a focus on the
  life sciences. I will argue that whilst new AI capabilities\, in particul
 ar the 'reasoning' abilities of large language models (LLMs)\, have the po
 tential to boost the efficiency of research\, they could have a negative e
 ffect on the replicability of scientific results. More specifically\, I wi
 ll argue that AI-driven automation (ADA) can reduce scientists' ability to
  troubleshoot the experimental process by diminishing their ability to ide
 ntify\, understand\, and correct flawed research outputs. Two key implicat
 ions of this analysis are 1) that the move from traditional automation to 
 ADA needs to be managed in a context-sensitive manner which ensures that a
  laboratory's ability to deal with experimental error remains intact\, and
  2) that we need to develop a better understanding of error (and error-rea
 soning) in science in order to tackle implication 1).
LOCATION:Hopkinson Lecture Theatre\, New Museums Site
END:VEVENT
END:VCALENDAR