BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Embedded Agency - Adrià Garriga Alonso (University of Cambridge)
DTSTART:20190130T170000Z
DTEND:20190130T190000Z
UID:TALK119545@talks.cam.ac.uk
CONTACT:Adrià Garriga Alonso
DESCRIPTION:Most current theories of agency deal only with _dualistic_ or 
 Cartesian agents. That is\, the agent that makes the decisions is itself o
 utside of the world that it takes decisions in\, and is not affected by it
 .\n\nBut technically this is not the case\, as the algorithm that makes de
 cisions is implemented using something (a brain\, a computer) in the world
 \, and modifications to that something can change the algorithm.\n\nAccord
 ing to some views\, understanding the theory behind _embedded_\, non-Carte
 sian agency\, is key to solving some important problems in AI safety (prov
 ably aligned self-modification\, wireheading). Others think it's not so im
 portant. We shall learn about current attempts to build embedded agency th
 eories and discuss how important it is to continue work in that area.\n\nR
 eading list:\n\n- Embedded Agency sequence from MIRI: https://www.lesswron
 g.com/s/Rm6oQRJJmhGCcLvxh/p/i3BTagvt3HbPMx6PN
LOCATION:Cambridge University Engineering Department\, CBL Seminar room BE
 4-38
END:VEVENT
END:VCALENDAR