26 ms·
Moravec's Paradox
- zero-sharp 3y agoI've thought about this in the past. When we can spell out the rules & formalize the task, machines can usually outperform people at that task. But our cognition doesn't just operate on simple rules, like in a chess game. It has the capacity to recognize similarity, for example. And that's a lot harder to define. Like when is a face a human face? Obviously technology has come a long way, but I don't think computers always had an answer to those sorts of vague, but intuitive, questions.