
Challenges of Adopting AI in Industry

#16 - Responsible AI and Building Trust in AI - Liu Feng-Yuan
Topic 9 of 16
Published Nov 23, 2020 · Clip 30:14–33:47 (3:33)
Feng-Yuan identifies two adoption bottlenecks: getting the right data in place (generic cloud APIs don't cover domain-specific problems like manufacturing defects), and the translation/ideation conversation between business, data science, and technology teams. He notes businesses often bring science-fiction problems like stock prediction or research-level problems, and advocates solving 80% of a process with AI and leaving the last 20% to humans.