The Godfather of Deep Learning
Geoffrey Hinton is one of the geniuses who helped create modern AI
What is Deep Learning?
Recently Deep Learning has become the hottest subject in the field of artificial intelligence, thanks in particular to breakthroughs in image and language recognition in recent years that have approached or surpassed human levels of comprehension.
It is basically a subfield of machine learning algorithms inspired by the structure and function of the brain called artificial neural networks. Deep neural networks uses multiples layers or artificial neurons to progressively extract higher level features from the raw input.
Deep Learning VS Traditional Machine Learning
Most of Deep Learning's core ideas were proposed during the 1980–1990s, but the lack of digital data to feed these algorithms and their higher CPU consumption made them impractical at the time. As the computer’s processing power increased and the nature of the world information became more and more digital, the algorithms began to perform better and better, surpassing traditional methods.
They require high-end machines contrary to traditional Machine Learning algorithms, that's why GPU has become an integral part now to execute any Deep Learning algorithm.
In traditional Machine learning techniques, most of the applied features need to be identified by a domain expert in order to reduce the complexity of the data and make patterns more visible to learning algorithms to work.
Deep Learning algorithms try to learn high-level features from data in an incremental manner. This eliminates the need of domain expertise and hard core feature extraction.
But how did it start and who are the main figures behind it?
Who is Geoffrey Hinton?
There are several deep-learning heroes to mention , but in general the English Canadian Geoffrey Hinton — together with Yoshua Bengio and Yann LeCun — are referred to as the “Godfathers of AI”. Those three researchers have won the 2018 Turing Award, known as the ‘Nobel Prize of computing,’ for conceptual and engineering breakthroughs in artificial intelligence (AI). In this particular article, I will focus only on Hinton.
Dr. Geoffrey Hinton is VP and Engineering Fellow of Google, Chief Scientific Adviser of The Vector Institute and a University Professor Emeritus at the University of Toronto. Hinton received a Bachelor’s degree in experimental psychology from Cambridge University and a Doctoral degree in artificial intelligence from the University of Edinburgh. He coauthors several relevant fields including applications of Backpropagation, Boltzmann machine, Deep learning, and most recently Capsule neural network. Another interesting fact about him: Hinton is the great-great-grandson of logician George Boole whose work eventually became one of the foundations of modern computer science.
In the next sections I would like to share some very relevant and informative interviews with him.
Bloomberg Interview
In this non-technical interview, Dr. Hinton provides an overview of how things have evolved over the years since his first ideas were published in the 1980s/1990s.
Heroes of Deep Learning Interview
In this more technical (as well as my favorite) interview conducted by Dr. Andrew Ng, they discuss various internal topics in Deep Learning, not just restricted to Dr. Hinton’s work, but several other researchers work. Almost all the papers mentioned in this interview can be found here, thanks to my fellow student Darryl Wright.
This is a must watch interview for any data scientist.
Google I/O 2019 Stage Interview
In this stage interview at Google I / O 2019, Geoffrey exposes the conceptual and engineering breakthroughs that have made deep neural networks a critical element of computing.
At some point Nicholas Thompson, Dr. Hinton’s partner in the chat, challenged the idea that machines could learn to perform any and all human brain activities: “There is no emotion that couldn’t be recreated? There is nothing of humans that couldn’t be recreated by fully functional neural networks? And you are 100% confident on this?”
Dr. Hinton replied that he was “99.9% sure.”
“What about that 0.1%?”
“We might be in a big simulation,” joked Dr. Hinton.
Informal Conversation between Nick Bostrom and Geoffrey Hinton
This is a very interesting informal conversation between Geoffrey and AI risk researcher Nick Bostrom in November 2015, conducted by the New Yorker's journalist Raffi Khatchadourian, about the future and risks of AI.
Nick Bostrom is Swedish-born philosopher and polymath with a background in theoretical physics, computational neuroscience, logic, and artificial intelligence, as well as philosophy. He is best know as the author of Superintelligence and Simulation Hypothesis.
Recent Publications
Still very active, over the last years Dr. Hinton has co-authored several paper publications, among my favorites CapsNet, a new approach to Computer Vision and Pattern Recognition.
Resources
If you wish to learn some Deep Learning concepts directly from Geoffrey Hinton, the Youtube channel Artificial Intelligence — All in One provides a playlist with the lectures of “Neural Networks for Machine Learning” taught by him.
