Artificial Superintelligence: A Futuristic ApproachA day does not go by without a news article reporting some amazing breakthrough in artificial intelligence (AI). Many philosophers, futurists, and AI researchers have conjectured that human-level AI will be developed in the next 20 to 200 years. If these predictions are correct, it raises new and sinister issues related to our future in the age of |
Contents
AICompleteness The Problem Domain of Superintelligent Machines | 1 |
The Space of Mind Designs and the Human Mental Model | 21 |
How to Prove You Invented Superintelligence So No One Else Can Steal It | 41 |
Wireheading Addiction and Mental Illness in Machines | 57 |
On the Limits of Recursively SelfImproving Artificially Intelligent Systems | 81 |
Singularity Paradox and What to Do About It | 107 |
Superintelligence Safety Engineering | 135 |
Artificial Intelligence Confinement Problem and Solution | 145 |
Efficiency Theory A Unifying Theory for Information Computation and Intelligence | 167 |
Controlling the Impact of Future Superintelligence | 185 |
Back Cover | 191 |
Other editions - View all
Common terms and phrases
ability AI-Box AI-Complete AI-Hard algorithm analysis answer approach Artificial General Intelligence Artificial Intelligence artificially intelligent system Behavioral Biometrics Ben Goertzel Bostrom capable CAPTCHA Chalmers cognitive communication complexity computational resources Computer Science confinement environment confinement problem efficient Eliezer Eliezer Yudkowsky encoded Eurisko example experiments formalize friendly Friendly AI future Gavrilova goal Goertzel hardware Hibbard Hugo de Garis human human-level intelligence IEEE improvement input intel intelligent agent intelligent machines International Conference knowledge ligent limited machine ethics mind design Nick Nick Bostrom Omohundro optimization oracle Paper presented philosophical possible potential proof proposed protocol question Recognition recursive self-improvement reverse Turing test reward function robot Roman RSI process RSI systems safety engineering Schmidhuber scientists Shahaf simulated singularity solution solving source code specific Springer string SuperCAPTCHA superintelligent machines technological singularity theoretical theory tion Turing test University utility function Waser wireheading Yampolskiy Yudkowsky