Hyper-Personalized AI and Its Dangers

#16 - Responsible AI and Building Trust in AI - Liu Feng-Yuan

Topic 14 of 16

Published Nov 23, 2020 · Clip 43:20–45:51 (2:31)

Feng-Yuan discusses hyper-personalization as tailoring recommendations from overwhelming choice, like e-commerce and Netflix. He warns of dangers: path dependency feeding people more of the same, optimizing engagement sliding into addiction, and echo chambers where people only see confirming content and label differing views as fake news. He suggests hyper-personalization may increase unknown unknowns by narrowing exposure.