In these articles and podcast artificial intelligence pioneer Yoshua Bengio explains why he thinks that true artificial intelligence will only be possible once machines have something babies are born with: the ability to interact with the world, observe what happens, and adapt to the consequences of their actions. He also explains how something called deep learning will help machines act to understand the world in a similar way, and how machines can already extract meaning from language using clever mathematical representations of words.
The agent perspective — How do we learn and understand the world? What separates us from machines? Yoshua Bengio believes it all comes down to our agency.
Building meaning: Dog or ostrich? — How do we build our models of the world? Yoshua Bengio explains how deep learning will help machines act to understand the world in a similar way.
AI, babies, and agency: The podcast — Listen to Yoshua Bengio talk about his work and ideas on artificial intelligence, and find out how machines can be taught to only pay attention to what matters.
You may also want to read...
The following article, originally published on the FQXi community website, looks at how our ideas of cause and effect, which we learn about by interacting with the world, could be revised in the light of quantum mechanics to build better artificial intelligences.
Is causality fundamental? — Untangling how the human perception of cause-and-effect might arise from quantum physics may help us understand the limits and potential of AI.
