‘Godfathers of AI’ honored with Turing Award, the Nobel Prize of computing

The 2018 Turing Award, known as the "Nobel Prize in Computing," was awarded to a trio of researchers who laid the foundations for the current boom in artificial intelligence.

Yoshua Bengio, Geoffrey Hinton and Yann LeCun, sometimes called the "AI sponsors", have been recognized with the annual $ 1 million prize for their work in the development of the IA deep learning subfield. The techniques developed by the trio in the years 1990 and 2000 allowed great advances in tasks such as artificial vision and speech recognition. His work underpins the current proliferation of AI technologies, from self-driving cars to automated medical diagnostics.

In fact, you probably interacted with the descendants of the Bengio, Hinton, and LeCun algorithms today, whether it's the facial recognition system that unlocked your phone or the model of AI language That suggested what to write in your last email.

Since then, all three have occupied prominent places in the IA research ecosystem, between academia and industry. Hinton divides his time between Google and the University of Toronto; Bengio is a professor at the University of Montreal and started an AI company called Element AI; while LeCun is the chief artificial intelligence scientist at Facebook and a professor at New York University.

"It's a great honor," LeCun told The Verge . "As good as it gets in computing, it's an even better feeling that is shared with my friends Yoshua and Geoff."

Jeff Dean, AI's AI chief, praised the trio's achievements. "Deep neural networks are responsible for some of the greatest advances in modern computing," Dean said in a statement. "At the heart of this progress are the fundamental techniques developed by the winners of this year's Turing Award, Yoshua Bengio, Geoff Hinton and Yann LeCun."


godfathers of ai honored with turing award the nobel prize of computing

The deep learning techniques that Bengio, Hinton and LeCun developed support many technologies, such as self-driving cars.
Photo: Aeva

The achievements of the trio are particularly remarkable, as they maintained their faith in artificial intelligence at a time when technology prospects were discouraging.

AI is known for its boom and bust cycles, and the subject of exaggerations is as old as the field itself. When the research does not meet the inflated expectations, the funding and interest known as "AI winter" is frozen. It was at the end of one of those winters in the late 80's when Bengio, Hinton and LeCun began to exchange ideas and work. Related problems These included neural networks, computer programs created from connected digital neurons that have become a key building block for modern AI.

"There was a dark period between the mid-1990s and early to the mid-2000s when it was impossible to publish research on neural networks, because the community had lost interest in it," says LeCun. "In fact, I had a bad reputation, it was a bit taboo."

The trio decided they needed to rekindle interest and raised funds from the Canadian government to sponsor an interrelated research center. "We organized regular meetings, regular workshops and Summer for our students, "says LeCun." That created a small community that […] around 2012, 2013 really exploded. "

During this period, all three showed that neural networks could achieve solid results in tasks such as However, the rest of the research world did not pay attention until 2012, when a team led by Hinton adopted a well-known artificial intelligence reference point called ImageNet Until now, researchers had only implemented incremental improvements in this challenge. object recognition, but Hinton and his students broke the next best algorithm in m more than 40 percent with the help of neural networks.

"The difference there was so great that a lot of the people, you could see a big change in their head" clunk, "says LeCun." Now they were convinced. "


  Facebook Prineville Data Center

The deep learning framework that created the trio was then fueled by abundant processing power and data.
Photo: Vjeran Pavic

Cheap GPU processing power (originally designed for games) and a large amount of digital data (issued on the Internet in the same way that a car emits gases), it offers fuel for these small cognitive engines and since 2012, the basic techniques that Bengio, Hinton and LeCun promoted, including inverse propagation and convolutional neural networks, have become ubiquitous in AI and, by extension, in technology in general.

LeCun says he is optimistic about artificial intelligence perspectives, but he also knows that much is needed. s work before the country fulfills its promise. The current systems of artificial intelligence need a lot of data to understand the world, they can be easily deceived and they are only good at specific tasks. "We just do not have machines with common sense," says LeCun.

For the field to continue on its upward trajectory, it will be necessary to discover new methods that are as fundamental as those developed by AI sponsors.

"If we can use new methods to create intelligence on a human level, well, there are probably another 50 mountains to climb, including those we can not see yet," says LeCun. "We have only climbed the first mountain, maybe the second."

Please Note: This content is provided and hosted by a 3rd party server. Sometimes these servers may include advertisements. igetintopc.com does not host or upload this material and is not responsible for the content.