Presentations by a pioneer of deep learning Geoffrey hinton and University of Toronto alumnus and entrepreneur Allen Lau were among the highlights on the opening day of this year’s influential collision technology conference.

Hinton, a professor emeritus at the University of the University of Toronto who now works at Google and has been dubbed the “godfather of deep learning,” spoke on Tuesday of his efforts to develop a new type of artificial neural network capable of recognize and interpret images better than existing images. systems.

One of more than 600 presenters at the annual conference which runs through April 22, Hinton noted that most systems today use convolutional neural networks that don’t recognize objects the way people do.

“They use a lot of texture information, which people are immune to, and they use a lot of shape information that people are sensitive to,” said Hinton, an AM Turing award winner who is also the chief scientific adviser of the Vector Institute for Artificial Intelligence.

“They are also not very good at extrapolating to new points of view. People can see an object from one point of view and then extrapolate to many different points of view. But convolutional neural networks need to see objects from many different perspectives to understand them. “

To illustrate this flaw, Hinton showed images of a school bus slightly obscured by visual noise. While humans can still easily recognize the images as showing school buses, Hinton revealed that convolutional neural networks mistake buses for ostriches.

His solution? A system called GLOM – short for “agglomeration” – which is inspired, in part, by the relationship between cells, organs and DNA.

“What I want to do is design neural networks so that they have different ways of seeing the same thing, just like people do, and that they see things the way people see things,” he said. he declared. “It would make them a lot more interpretable and a lot less likely to make crazy mistakes – like that school bus.”

For the second year in a row, Collision – which features global speakers from technology, business and media – is being held virtually due to the COVID-19 pandemic. The event, which counts the U of T as an official partner, was to be held in Toronto for three years starting in 2019.

The topic of AI was also at the heart of the panel discussion with Lau, whose Wattpad storytelling platform was recently bought by South Korean internet giant Naver for US $ 600 million. Lau joined senior Naver executives for a discussion on democratizing storytelling through digital technologies.

Lau, who founded Wattpad with another U of T alumnus Ivan Yuen in 2006, said that using AI to study feedback and information provided by Wattpad users provides “information that no one else has.”

“Due to the sheer volume of content and the ideas, comments and other data surrounding it, leveraging AI is a very obvious choice for us to understand the ideas of content programmatically,” said Lau, alumnus. from the faculty of the University of Toronto. of Applied Science & Engineering. “By taking advantage of this, we can adapt the content to other forms such as movies, TV shows, and print books, as we started doing a few years ago.”

He added that the combined 160 million users of Wattpad and Naver – and the growing reach of businesses – would serve to continually diversify their content and meet changing public tastes.

“A few people held the key to the entire system in the past – now we are expanding it to seven billion people on this planet. Everyone will have their say, ”Lau said.

“So in this context, I think the content will be more diverse than before because the contributions to the system will not come from a handful of people – they will come from the world’s population.”

Jon French, Director of U of T Entrepreneurship, and Derek Newton, Assistant Vice President of U of T, Innovation, Partnerships and Entrepreneurship, moderated a session to help entrepreneurs, industry leaders, investors and government to engage in U of T’s innovation and entrepreneurship ecosystem.

The U of T also hosted a session on Wednesday to help entrepreneurs, industry leaders, investors and government engage in the university’s thriving ecosystem of innovation and entrepreneurship. . Over the past decade, U of T entrepreneurs have created more than 500 businesses and leveraged over $ 1.5 billion in investment.

During the session, Jon French, director of the U of T Entrepreneurship, noted that the university’s campus-linked network of accelerators – spread across its three campuses – provides a range of resources and supports to entrepreneurs at all stages of their journey, and with different backgrounds and areas of expertise.

“We always say that there is no single door, or bad door, in the entrepreneurial ecosystem, and that our students, researchers, professors and alumni who start businesses can very often participate in several linked accelerators. on campus to get to where they need to be. ,” he said.

Derek newton, Assistant Vice-President, Innovation, Partnerships and Entrepreneurship at the University of Toronto, added that the University of Toronto has worked hard to combine its research strengths with the resources to protect and commercialize these ideas. This includes the protection and management of intellectual property (IP). Since 2016, investment and M&A activity involving university IP-based startups has grown from $ 175 million to over half a billion dollars in 2021, he said. .

“This is an exciting trajectory that we seek to accelerate and continue – with your help, with your partnerships – to help the University of Toronto launch even more businesses that will continue to grow and evolve here in Toronto and around the world, ”Newton said.

Newton also praised the potential impact of the Schwartz Reisman Innovation Campus, which is being built in St. George with the help of a $ 100 million donation from Gerald Schwartz and Heather reisman.

“This is truly the next generation of the University of Toronto’s support for our innovators, entrepreneurs and businesses that rely on made-in-Canada intellectual property,” said Newton.

U of T’s involvement in Collision continues Thursday with a conference by Mike Murchison, CEO of customer support company AI ADA. He will talk about emerging trends in the customer experience space.