Observatoire des droits humains (anglais seulement)

Forensics tool ‘reanimates’ the ‘brains’ of AIs that fail in order to understand what went wrong

(Version anglaise seulement)

par David Oygenblik, Ph.D. Student in Electrical and Computer Engineering, Georgia Institute of Technology
Brendan Saltaformaggio, Associate Professor of Cybersecurity and Privacy, and Electrical and Computer Engineering, Georgia Institute of Technology

AIs are notoriously opaque, even to the people who build them, which makes it hard to know why they fail. A new tool aims to reveal their inner workings at the moment they went off the rails.

Lire l'article complet