Interview
Garry Kasparov: Chess, Deep Blue, AI, and Putin | Lex Fridman Podcast #46
- Artificial intelligence is expected to become a defining factor in civilization's future, with the Deep Blue match historically inspiring a generation of researchers.
- In confined systems like chess, Go, Shogi, and Dota, machines are predicted to always prevail over humans by making fewer errors, validating the view that human competition in intellectual fields was fundamentally flawed.
- Machines will not reach 100% perfection but are expected to amplify existing societal biases rather than resolve global issues, while potentially struggling to reproduce unique human qualities.
- Correcting errors in advanced systems like AlphaZero may require hundreds of thousands of games when facing superior opponents.
- The greatest risk in human-machine collaboration is identified as humans attempting to interfere with superior machine knowledge.
- New generations of players possess significantly more game knowledge than predecessors due to accumulated information and access, though the speaker anticipates their own ability to contribute new knowledge to chess is diminished.
- Undemocratic and totalitarian systems relying on central planning and citizen control are forecast to fail as drivers of innovation and are predicted to end suddenly.
- Specific concerns include the Putin regime facing internal difficulties with an uncertain timeline for its conclusion, and a prediction that Russia will interfere in the 2020 U.S. election to ensure Donald Trump's survival.
- Donald Trump's reelection is expected to ruin NATO and cause damage to America and the free world beyond imagination.
- Political and personal plans include a return to Russia sooner than anticipated and a continued commitment to making a difference as long as energy, strength, and passion remain.
- Future actions are tempered by the recognition that changing past events could negate other accomplishments due to the butterfly effect.