BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Bias in AI - Derek McAuley
DTSTART:20201008T140000Z
DTEND:20201008T150000Z
UID:TALK151690@talks.cam.ac.uk
CONTACT:Srinivasan Keshav
DESCRIPTION:“Blame it on the algorithm” is a common refrain these days
  when systems that include algorithmic components deliver very clearly una
 cceptable results.  Examples range from Amazon’s gender biased hiring  (
 machine learning) algorithm that learned from biased examples\, though rac
 ial bias in AI methods used for predictive policing\, to the recent UK A-l
 evel grade “normalization” of 2020\, which was biased by design. These
  scenarios are often succinctly referred to as “algorithmic bias”\, an
 d there are many more examples\, some of which are not so obvious. Moreove
 r\, this phrasing often leads people to think that this purely a technolog
 y problem when in fact one must look at the complete system\, including th
 e human elements\, in which the algorithm is created and operated\, to und
 erstand how these systems make mistakes. The talk will highlight the ways 
 in which “algorithmic bias” can arise\, and how we can identify and mi
 tigate the damage.\n\nBio:\n\nDerek McAuley is a British academic who is P
 rofessor of Digital Economy in the School of Computer Science at the Unive
 rsity of Nottingham and director of Horizon Digital Economy Research\, an 
 interdisciplinary research institute funded through the RCUK Digital Econo
 my programme. He acted as a Specialist Advisor to the House of Lords Europ
 ean Union Committee into online platforms\, and Chief Innovation Officer d
 uring the founding of the Digital Catapult. He is a Fellow of the British 
 Computer Society and member of the UKCRC\, a computing research expert pan
 el of the Institution of Engineering and Technology and BCS.\n\nhttps://en
 .wikipedia.org/wiki/Derek_McAuley \n\n
LOCATION:https://meet.google.com/ehj-dwaz-rea
END:VEVENT
END:VCALENDAR