National Post (Latest Edition)
Three Quebec teams in Top 10 on AI XPrize
PSYCHIATRY, TUTORING AND REMEDIATION FIRMS
In 1919, New York entrepreneur Raymond Orteig offered a US$25,000 prize to the first pilot to fly non- stop between New York and Paris. The Orteig Prize, which would be worth US$ 360,000 today, is credited with spurring huge advances in aviation before it was finally won in 1927 by a cocky daredevil named Charles Lindbergh.
In 1996, when NASA’s space- exploration efforts were stalling, reading about Lindbergh i nspired U. S. aerospace entrepreneur Peter Diamandis to launch the XPrize: a US$ 10- million award to the first team to fly a three-passenger vehicle 100 kilometres into space, twice, within two weeks. After SpaceShipOne won the competition, Diamandis’s XPrize Foundation began offering multiple challenges, catalyzing innovation in fields ranging from tricorders ( Star Trek- inspired medical scanners) to adult literacy and oilspill cleanups.
Even with seven XPrize contests now on the go, few people noticed last December when the IBM Watson XPrize for Artificial Intelligence issued its first ranking of the 150 teams competing for US$5 million in prize money. But as the world rockets warily into the AI economy, Canadians can take pride that three of the Top 10 teams are from Quebec.
The AI XPrize aims to spark audacious projects that “solve societal grand challenges.” The Quebec teams’ early success proves not only that AI can be a powerful force for good, but that Canadian entrepreneurs — with their smarts and conviction — can play a big role in the AI revolution.
These are the three Canadian AI XPrize leaders.
Founded in August 2017 by neuroscientists at McGill University, this Montreal firm aims to leverage the collective intelligence of the global medical community to more quickly diagnose and treat patients with mental- health issues. Aifred ranked second among all entries in December, based on the company’s potential impact and its progress so far.
G oal: “T he quality of psychiatric care needs to be better,” says co- founder and CEO David Benrimoh, a psychiatry resident at McGill. “If you’re a patient, we don’t know what treatment will work for you. One patient took nine years to find the right remedy. Aifred will compress that nine years into nine minutes.”
Product: Aifred Health’s deep- learning platform will access millions of clinical trials to recommend the best treatments for each patient. An app sold to doctors will collect real- time patient moods and conditions, and supply more data for the predictive engine. Initially Aifred is focusing on depression, which costs governments, patients and employers billions of dollars a year. It will then move on to treatments related to anxiety, addiction and PTSD.
Plan: A “lite” version of the product will be released by summer to help physicians follow best practices and feed anonymized data to Aifred. Clinical trials for the predictive platform should begin late this year.
Current state: With just one paid employee (and lots of volunteers), Aifred is looking for financing. Benrimoh says the XPrize Foundation has provided valuable exposure and facilitated conversations with potential investors.
Impact: “This is a humanitarian- focused work,” says Benrimoh. “AI is going to change the way mental health is practised.”
This Montreal firm is developing a free math- tutoring platform that uses AI to match students who need specific help with peers who have already mastered those concepts.
Goal: Montreal tech entrepreneur Patrick Poirier founded Erudite i n 2012 to help people learn. He’s a 44- year- old high- school dropout who later caught up, earning five degrees ( in- cluding computer science, psychiatry and molecular biology). Poirier founded Erudite in 2013, at first to develop math- based games, then as a tutoring platform. “Tutoring is very efficient at helping people improve their grades,” he says. “It’s a US$56-billion market. But at $40 an hour, it’s very expensive.”
Product: After experimenting with automated learning platforms, Poirier decided the best tutor for high- school students is another high-schooler. His free service, ERI, matches struggling math students with stronger ones (who donate their time). AI will match up learning partners, preserving their anonymity while monitoring t heir online chats and pointing out errors.
Plan: Fo l l o w i n g pilot projects in Montreal and Kenya, Poirier hopes to go live in September, and host 200,000 students by year end. By mid-2019 he hopes to earn revenue by selling a version of the platform to commercial tutoring firms, to help them speed up teaching time and reduce costs.
Current state: With seven full-time employees, Erudite currently runs on Poirier’s sweat equity and $ 2 million from the Canadian Media Fund, the federal IRAP program and a New York City incubator. It’s now working with a U. S. broker to raise money through equity crowdfunding.
Impact: With AI’s power to learn, Erudite hopes to extend its tutoring services beyond algebra to geometry, and then the sciences, in two years. “The AI will continue to improve,” says Poirier. “In five years I hope we will be helping 50 million people.”
This Quebec City firm, founded in 2015, is a partnership between golfing buddies Marc Paquet, a chemist with experience in environmental site assessments, and former IBM marketing executive Daniel Fortin. WikiNet is developing a system that learns from past environmental cleanup efforts to provide automated, expert recommendations for treating contaminated lands.
Goal: Sites contaminated by industrial uses are huge problems for land develop- ers, and for the governments that often get stuck with the cleanup. CEO Paquet hopes to make remediation faster and cheaper with an AI system that amasses global data on site remediation, and predicts the best ways to treat each site.
Product: Paquet contacted major engineering consultancies to confirm their interest in a predictive remediation product, while Fortin got assurances that IBM would support their efforts. WikiNet’s WatRem remediation assistant gathers information from documents, journals, conferences, consultants and governments to offer the best cleanup solutions. Its machine- learning algorithm will continually improve the quality of its predictions.
Plan: WatRem should hit the market this summer. WikiNet j ust announced a contract with the city of Montreal for a related product, Traces, that will monitor the transportation of contaminated soil to detect illegal dumping.
Current state: WikiNet has 10 employees and was mainly self- funded, with some help from the Business Development Bank and Quebec’s Desjardins Group. Paquet says the company is now talking to angel investors: “We have enough cash to get to commercialization, but we need to keep investing in our applications to keep them evolving.”
Im p a c t : According to Paquet, 33 per cent of the world’s agricultural lands are contaminated, creating severe consequences for consumers everywhere. By getting solutions into the field faster, “We can have a huge impact on society,” he says. And once experts have a common platform to share remediation solutions, he predicts, global collaboration and innovation will soar.
Asked if WikiNet can win the XPrize contest, which ends in 2020, Paquet reveals just a hint of Lindbergh-like confidence. “Of course,” he says. “Our name’s already on the cheque.” Rick Spence is a writer, consultant and speaker specializing in entrepreneurship. rick@ rickspence. ca Twitter. com/ RickSpence