Indian-origin professor is driving innovation in artificial intelligence in Canada
Professor Ajay Agrawal, a prominent figure in the field of machine learning, believes the next five years will witness a surge in narrow applications of artificial intelligence.world Updated: Apr 01, 2017 19:27 IST
Professor Ajay Agrawal feels like he’s back in 1995. That year, the first major commercial Internet service providers like AOL went online and Yahoo’s search engine became available to the public. It was an inflection point for the Internet and, Agrawal believes, artificial intelligence or AI may now be nearing that stage.
Agrawal, 47, is a professor of entrepreneurship at the University of Toronto’s Rotman School of Management, but he also happens to be the founder of the Creative Destruction Lab (CDL). He has emerged as a prominent figure in the field of machine learning, with CDL’s AI stream.
The programme launched in 2015, and the 2016 batch has 50 companies engaged in AI-related enterprises. “To our knowledge, having 50 AI companies represents the greatest concentration of AI companies on any programme on earth. As far as we know, that’s still true,” Agrawal said in an interview with Hindustan Times.
Agrawal, who was born in Vancouver, is also the co-founder of NextCanada, which started NextAI last year. NextCanada’s mission is to “increase national prosperity by providing an ecosystem to support the country’s most exceptional entrepreneurs and innovators”.
He said AI, as of now, “is a prediction technology, it’s a way to taking many forms of data, voice, text, all sorts of data captured by sensors, all types of cameras, heat sensors, lights sensors, and converting that into predictions of things, whether it’s predicting traffic or weather or what website you’re going to click on or what movie you’re going to watch.
“I’ve been interested in intelligent machines for a long time. I became interested in the economics of artificial intelligence about four years ago and then launched the AI stream of the Creative Destruction Lab two years ago and then launched NextAI in late 2016.”
Playing the part of guides at the CDL course are leaders in AI technology, such as Cambridge, the England-based William Tunstall-Pedoe, founder of Evi, which was acquired by Amazon and became “an integral part” of Alexa; Barney Pell, who managed an 85-strong team at NASA that flew the first AI system in deep space; and Russ Salakhutdinov, Apple’s director of AI research.
As Agrawal plays a pioneering role in building AI talent in the country, CDL will increase its intake to 75 companies in 2017. The “big thing” that will occur is that while 50 will come in “classical machine learning”, they will “launch the world’s first seed stage programme in quantum machine learning”, featuring 25 companies.
This programme will be in partnership with Vancouver-based D-Wave, believed to be the maker of the world’s first commercial quantum computer, which has been purchased by the likes of Nasa and Google.
The 2015 batch of the AI stream, the only one to have “graduated”, has already seen nearly half of the companies funded. Investors, including those from Silicon Valley, have been flying into Toronto. As Agrawal said, “We’ve brought intelligent investors into the Canadian eco-system.”
This programme will be formally announced in about five weeks, and it will be “open to people from around the world,” Agrawal said.
While the first two batches have no Indian companies, Agrawal hopes that will change. “I imagined when we launched this new quantum machine learning, that students at places like IITs might be interested. For many people, it will be the only place in the world where they will be able to go to in order to get access to a quantum computer.”
Rather than using binary bits as in conventional computers, quantum computers use quantum bits or qubits to encode information, enhancing the power of the system.
This is a nine-month programme that “pairs start-up founders with experienced technology entrepreneurs and investors in a unique setting that is laser-focused on designing a series of objectives with measurable deliverables.”
Agrawal expects the next five years to witness a surge in narrow applications of AI. After that, he said, “It will become a lot more interesting. Rather than having narrow prediction applications, we’ll start seeing things that feel like real intelligence embedded into inanimate objects.”
In a decade, according to Agrawal, we could see the artificial gaining ground on the real. In that period, he will be among the leaders in Canada in harbouring and harnessing a catchment of intelligence required to mine that potential.