Godfather of Artificial Intelligence" Geoffrey Hinton on the promise, risks of advanced AI

geoffrey-hintonai-safetyneural-networksai-consciousnessturing-award60-minutes

Abstraction: Geoffrey Hinton 60 Minutes interview warning of AI existential and societal risks

Key points:

Hinton believes current AI systems "understand" and are "intelligent," and will eventually have consciousness and exceed human intelligence
He warns AI could "escape control" by autonomously writing and executing code to modify itself
Immediate risks cited: unemployment, fake news, unintended bias in employment/policing, and autonomous battlefield robots
Hinton won the 2019 Turing Award with Yann LeCun and Yoshua Bengio for work on artificial neural networks
Neural network learning algorithm was designed, but the resulting networks are not fully understood — "like designing the principle of evolution"
Hinton has no path that "guarantees safety" and calls for global regulation and a treaty to ban military robots; compares AI moment to Oppenheimer and the atomic bomb

Connections: Geoffrey Hinton · Google · Chatgpt · Openai · AI Safety · Neural Networks · Large Language Models