(T) Artificial intelligence, neural networks, quantum computing, bioinformatics…those are exciting fields for computer scientists. But while quantum computing and bioinformatics are still new, artificial intelligence and neural networks have been studied for a very long time. So why artificial intelligence (AI) has never created a wave of successful start-ups as did the Internet? Why neural networks (neural nets) has never created a wave of successful start-ups as did the Internet?
AI was hot in the mid-80s, simply said, trying to make cognitive systems, and hit a sweet spot with expert systems which helps in making decisions when a knowledge base could be modeled. Neural nets started to gain popularity in the early 90s, simply said, trying to make systems that process information as a network of human neurons, and hit a sweet spot for systems that can predict patterns, and suggest new outcomes from learning past ones.
Probably, the insane success of the Internet is simply due to its simplicity to exchange reliable information between a growing number of computers (although HTTP is a simple (and maybe not so well designed) protocol…TCP/IP is not!). While both AI and neural nets have ambitious goals using complex technologies, AI and neural nets have provided wonderful fields for academia but not for venture capitalists.
However, the number of techniques and algorithms that derived from AI and neural nets, and have made it into consumer, business and government computing applications are phenomenal for a broad range of systems that require to analyze, recognize, decide, predict, and optimize.
There is probably not a better proof to that statement that the brainpower that Google has invested in AI: Peter Norvig, co-author with Stuart Russell, of the popular textbook Artificial Intelligence: A Modern Approach; Sebastian Thrun, a well-known machine learning and robotics professor and scientist; and recently hired inventor, entrepreneur, and futurist Ray Kurzweil.
Peter Norvig directed the core search algorithms group at Google, Sebastian Thrun led the development of the Google self-driving car, while Ray Kurzweil is working on natural language.
And in addition this week, the University of Toronto has announced that “Google has picked up a ground-breaking start-up out of (its) Department of Computer Science…University Professor Geoffrey Hinton and two of his graduate students, Alex Krizhevsky and Ilya Sutskever, incorporated DNNresearch Inc. in 2012, and the company has been acquired by Google for its research on deep neural networks. Hinton is world-renowned for his work with neural nets, and this research has profound implications for areas such as speech recognition, computer vision, and language understanding.”
It goes without saying that Google will certainly leverage Professor Hinton’s pioneering work in deep learning, a machine learning technique, that can among many things recognize objects, and translate speech in real time.
For more on AI:
To refresh your memory about Artificial Intelligence, the course on Udacity by Peter Norvig and Sebastian Thrun:
A great talk about the Future of Robotics and Artificial Intelligence, from Professor Andrew Ng, Director of Stanford Artificial Intelligence Lab:
To refresh your memory about Neural Networks, a talk by Professor Hinton:
For the recent developments in Deep Learning by Professor Hinton:
Note: The picture above is a slide from Professor Hinton.
Copyright © 2005-2013 by Serge-Paul Carrasco. All rights reserved.
Contact Us: asvinsider at gmail dot com.
Categories: Artificial Intelligence, Deep Learning