Chapter
The Responsibility of Creating AI Systems that are Fueled by Self-Doubt
58:43 - 1:10:37 (11:53)
As AI systems become more intelligent, it's important for them to have self-doubt to maintain human control and supervision over their actions. Society should also create social norms that encourage doing good and being perceived well by peers to ensure AI is ultimately used for good.
Clips
Stuart Russell believes that as AI systems become more intelligent, they should be programmed with self-doubt to maintain human control over the technology.
58:43 - 1:03:03 (04:20)
Summary
Stuart Russell believes that as AI systems become more intelligent, they should be programmed with self-doubt to maintain human control over the technology. The flexibility and versatility of even a human toddler is still vastly superior to current AI capabilities.
ChapterThe Responsibility of Creating AI Systems that are Fueled by Self-Doubt
Episode#93 – Daphne Koller: Biomedicine and Machine Learning
PodcastLex Fridman Podcast
The use of technology should be properly monitored to prevent adverse effects.
1:03:03 - 1:06:53 (03:49)
Summary
The use of technology should be properly monitored to prevent adverse effects. The podcast discusses the potential dangers of machine learning, social engineering, and gene editing, and how society must regulate them.
ChapterThe Responsibility of Creating AI Systems that are Fueled by Self-Doubt
Episode#93 – Daphne Koller: Biomedicine and Machine Learning
PodcastLex Fridman Podcast
The speaker believes that most people mean well and creating social norms that correlate doing good with being perceived well by peers is important for our society.
1:06:53 - 1:10:37 (03:44)
Summary
The speaker believes that most people mean well and creating social norms that correlate doing good with being perceived well by peers is important for our society. The question of whether machine learning and gene editing will lead us to a better world relies on the underlying belief that most people are good and tend towards doing good.