Dr Ben Goertzel - Why The Godfather of A.I. Quit Google - Warns Of Danger Ahead
London Real
00:00
The Problem With Reward Maximization in AI
OpenCog founder Ted Mark says aji can't be achieved by reward maximizing agents. He argues that human values are messy and diffuse, so it's hard to design good rewards for AI systems. OpenCog is not really about reward maximization, he says; its more about making a statement.
Transcript
Play full episode
