The pace at which Yudkowsky and Ngo can go back-and-forth with each other without losing context is fascinating and intriguing. Just reading the first 90 minutes of the September 5th conversation took me an hour to absorb. These two have very divergent perspectives with their arguments for each problem presented. Yet, they are able to fluidly jump from topic to topic, while still making meaningful progress toward a shared goal, even if not explicitly defined. I’m curious to know Yudkowsky’s current position on his former suspicion of GPT-6 being a true AGI.
The pace at which Yudkowsky and Ngo can go back-and-forth with each other without losing context is fascinating and intriguing. Just reading the first 90 minutes of the September 5th conversation took me an hour to absorb. These two have very divergent perspectives with their arguments for each problem presented. Yet, they are able to fluidly jump from topic to topic, while still making meaningful progress toward a shared goal, even if not explicitly defined. I’m curious to know Yudkowsky’s current position on his former suspicion of GPT-6 being a true AGI.