Rodney Brooks, on whether or not anything useful about AI can be learned from robot experiments. More abstractly, how does one solve the problem of demarcation in AI (what line separates AI-relevant research from everything else)?
Rodney Brooks, on whether or not anything useful about AI can be learned from robot experiments. More abstractly, how does one solve the problem of demarcation in AI (what line separates AI-relevant research from everything else)?