Artificial Superintelligence: A Futuristic Approach
By Roman Yampolskiy
In this comprehensive overview of AI safety engineering, Yampolskiy addresses the key question of how to ensure that superintelligent systems will prove beneficial to humanity. He discusses how to test for human-level intelligence, what it means for AI to self-improve, why humans may not remain the dominant species on Earth, and more. An excellent resource for anyone seeking a more in-depth understanding of the threat posed by AI and the work being done to ameliorate it.
Catch Roman on the FLI Podcast here.
Buy the book here.