In this episode, Byron talk about holding artificial intelligence accountable, human fallibility, and self-driving cars.
Transcript
How will we hold artificial intelligences accountable for the mistakes that they make. The mistakes can kind of come from two different sources. One can actually just be coding errors made by the people who are creating the system. The other, of course, are deliberate decisions made by the people coding the system.
In either case, as these systems come to inform and advise more and more of our lives, their propensity to make errors will only go up. Self-driving cars would be the obvious example, but the number of places that machine fallibility can cause enormous damage just boggles the mind. Of course human fallibility can also cause the same amount of damage in all of these areas, if not more, and we managed to make progress and advance civilization even with highly fallible people therefore presumably we will work it all out with artificial intelligence as well.
- Subscribe to Gigaom AI Minute
- iTunes
- Google Play
- Spotify
- Stitcher
- RSS