Robin Hanson

In my experience, prediction markets have the potential to aggregate diverse information effectively, making them superior to traditional methods for forecasting future events. Their real-world accuracy showcases their efficacy.

Simply expanding large language models (LLMs) is not enough to achieve artificial general intelligence (AGI). AGI requires a deep understanding of cognition, reasoning, and world models, which goes beyond mere scaling of existing technologies. We need innovation in AI frameworks and integration with diverse types of knowledge.