If AI is supposed to mimic – and one day even surpass – human intelligence, scientists first need to settle the debate about the nature of human cognition. Are humans born as blank slates or with certain cognitive predispositions? As Clive Thompson lays out in an essay for Wired, how computer scientists answer this fundamental question will determine the direction of future AI technology.
Deep learning is the driving force behind today’s most popular AI applications, including speech recognition and vehicle driving software. This machine learning method uses artificial neural networks that mirror the brain. Such networks allow AI to process vast quantities of data and start recognizing patterns independently. Some scientists believe that deep learning, on its own, won’t be able to take AI to the next level. Self-taught machines lack the ability to reason and make logical inferences. They also don’t have the “common sense” humans take for granted enabling people to master new situations based on past experiences.
Comment on this summary or 开始讨论