artificial
artificial

Science as a superhuman recursively self improving problem solving system

I'm watching this interview with Francois Chollet where he talks about science as an example of a superhuman recursively self improving problem solving system and how we can use it to reason about what a superhuman artificial general intelligence m…

Have a doctor explain to a patient that the diagnosis was made by an AI doctor twice as intelligent as, and vastly more knowledgeable than, the top human doctor in any medical specialty

Certainly, let's imagine how this might unfold. The doctor sits down across from the patient, maintaining eye contact and a level of directness. "Look, your diagnosis came from an AI medical system, and this isn't just any AI. Imagine the …

Using Multi-Agent Reinforcement Learning results in better urban planning outcomes

Urban planning is tricky – governments push top-down changes while locals want bottom-up ideas. It's hard to find compromises that make everyone happier. A new research paper proposes using Multi-Agent Reinforcement Learning (MARL) to vote on land …