(T) If you did not have the chance to attend Google IO this week at the Shoreline Amphitheater in Mountain View, most of the videos of the talks are now online. Among the many talks, one that interested me was Google Sr Fellow Jeff Dean, talking about AI with a few other guests.
Following are a few statements that are worth thinking about:
Things easy for humans are hard for computers, and the reverse is true
The quality and performance of machine learning algorithms have significantly improved in particular for deep neural networks applications, speech, and vision in particular
We still do not know how to learn from small data sets – children learn from small data sets
We do not know how to transfer learning from one domain to another one
Systems need to have the underlying algorithms and framework in place before to start learning – we do not know how to design a system which can choose the best algorithms and framework as it is learning
For more on our present transition (?) from a mobile-first to an AI-first world…
Note: The picture above is the Citroen 2CV of an old couple in Palo Alto.
Copyright © 2005-2016 by Serge-Paul Carrasco. All rights reserved.
Contact Us: asvinsider at gmail dot com.