Deep learning versus the rest: the future of AGI.

This is an excellent set of interviews with 23 of the world’s leading AI practitioners, entrepreneurs and investors. More than anything else it exposes the gap between two schools of thought. There are those who believe that deep learning is enough to solve all the remaining barriers to AGI (eg Hinton, Bengio) and those who don’t (eg Pearl, Marcus). The argument at the heart of it is that deep learning cannot currently master facilities such as common sense and causal reasoning. Rather, it excels in pattern recognition. The non-believers assert that the AI world will need to look to techniques beyond deep learning to achieve common sense and the like. There appears to be some interaction between the two fields of endeavour but only time will tell who is right.

Link to book:

You may also like to browse other AI books: