
What AI Can and Cannot Replace

#16 - Responsible AI and Building Trust in AI - Liu Feng-Yuan
Topic 4 of 16
Published Nov 23, 2020 · Clip 17:30–20:15 (2:45)
Feng-Yuan explains where AI is strong—narrow tasks with abundant data like language models, reinforcement learning in games, and computer vision—and where it fails: general reasoning and un-instrumented data, exemplified by HR attrition prediction since emotional dynamics aren't captured in email or Slack. He demystifies AI as brute force over data rather than conscious beings.