CACM Jun 2019 - 2018 ACM Turing Award

Once treated by the field with skepticism (if not outright derision), the artificial neural networks that 2018 ACM A.M. Turing Award recipients Geoffrey Hinton, Yann LeCun, and Yoshua Bengio spent their careers developing are today an integral component of everything from search to content filtering. So what of the now-red-hot field of deep learning and artificial intelligence (AI)? Here, the three researchers share what they find exciting, and which challenges remain. In this video, Yoshua Bengio, Geoffrey Hinton, and Yann LeCun discuss their breakthrough work and the path that led the three of them to receiving the 2018 ACM A.M. Turing Award. This interview is also printed in “Reaching New Heights with Artificial Neural Networks,” in the June 2019 issue of Communications of the ACM.

[@superxingzheng]

Geoffrey Hinton: If you have an idea and it seems to you it has to be right, don't let people tell you it is silly. Just ignore them.

[@aitechsystems5233]

This seems like piece of history being built. This video may be looked at like a defining moment of AI growth.

[@sarahchen8692]

Amazing! Just do what you love and inspire yourself for more interesting ideas~

[@placebooracle]

Very cool and deserved, though it is so not true that people did not work on neural networks between McCulloch&Pitts and Hinton et al in the 80s.... Many Neural network architectures were developed in 40s,50s, 60s and 70s under the Cybernetics movement of which McCulloch&Pitts were founding members. Many people worked on them: Shannon, Wiener, Rosenblatt, Ashby, Von Neumann, Von Foerster, (see for instance his Numa-Rete in the 60s and 70s ), Pask, many others.

[@nsomazr]

The first Perceptron which was a modefication of McCloch and Pitts work by Frank Rosebratt.

[@deeplearningpartnership]

Wow.

[@roberttlange8607]

Great video. Faster = more accurate?

[@casperes0912]

Well, somewhat. If it takes a million year to improve your accuracy, it's possible, but not feasible. Having a faster system allows for more accuracy not because it's not possible to compute with more accuracy on a faster device, but simply because the time needed to do so stops being as big a bottleneck

[@JazevoAudiosurf]

it's funny that we stopped the evolution from man to machine like it's nothing too important and eventually decided to go on evolving

[@Sirius_2780]

cool he is using thinkpad