Key takeaways
- How the panel's real world AI work differs from movie AI: narrow, task specific systems versus the general intelligence Hollywood portrays
- Favorite AI movies discussed, including Terminator 2, Ex Machina, Star Wars, and Black Mirror, and why "Be Right Back" struck the closest to reality
- The Turing test and how close, or far, today's chatbots are from truly passing as human
- Why hard coded ethical rules for robots, in the spirit of Asimov's laws, break down once a system can reinterpret its own objectives
- The ethical dilemmas already built into self driving cars, including whose safety the vehicle is designed to protect
- A closing detour into simulation theory and whether enough personal data could ever reconstruct a person after death
Overview
In this session, H2O.ai CTO Arno Candel and Kaggle Grandmasters Dmitry Gordeev and Gabor Fodor (Beluga) trade favorite AI movies, from Terminator 2 and Ex Machina to Black Mirror's "Be Right Back," and discuss what each gets right and wrong. The recording covers the gap between the general intelligence movies imagine and the narrow AI the panel builds day to day, plus the Turing test and Asimov's laws of robotics. It closes on a philosophical note, touching on simulation theory and whether enough data could ever recreate a person after death.