Against a backdrop of public markets falling and venture funding tightening, Canada was the only region to see healthcare AI investment increase in 2022’s first quarter. Across the board, healthcare is the top-funded sector in AI, according to CB Insights’ global State of AI report. When it comes to the intersection of AI and healthcare, Radical Ventures has maintained two core beliefs that have shaped our fund’s thesis since inception. First, we believe AI will impact every area of our lives, but it will have a disproportionate impact on healthcare delivery and consumption, helping to move us from reactive and generalized to proactive and personalized healthcare. Second, positioned at the nexus of world-class health networks, machine learning researchers and a data-rich single-payer system, Canada is uniquely positioned to be a leader in healthcare AI innovations.
Radical Ventures has made some exciting new investments in this space in the last two quarters, adding to a strong portfolio of healthcare AI companies which include Aspect Biosystems, PocketHealth, Synex, and Genesis Therapeutics. Earlier this year, we announced new investments in Unlearn.AI, a company applying AI to improve clinical trials and Birch AI, a leading NLP start-up targeting healthcare customer support. Last month, Radical Ventures Co-Founder and Partner Tomi Poutanen launched Signal 1, a Radical-backed health startup on a mission to transform hospitals with AI. Tomi wrote about his motivations for building Signal 1 in a recent issue of Radical Reads. Some additional investments will be disclosed soon.
Deploying healthcare AI in clinical settings is not easy, and a lot of work is being done to build trust amongst healthcare stakeholders. At Radical Ventures, we have seen a new generation of startups emerge under leaders who combine clinical, technical, and commercial expertise. These versatile teams are successfully deploying healthcare innovations in the real world and we want to amplify their stories and spur more innovation in the space. In the coming weeks, stay tuned for a new Radical Reads series, “Demystifying Healthcare AI” where we will hear from leading founders, researchers, and clinicians working at the intersection of AI and Healthcare. Stay tuned!
Sanjana will be participating in an Investor Roundtable on Thursday, May 26th at 1:30 ET alongside some of Canada’s top investors at the MaRS flagship healthcare event Impact Health. Moderating that panel is Eric Hoskins, Partner, Healthcare Innovation at Maverix Private Equity and Former Minister of Health in Ontario.
5 Noteworthy AI and Deep Tech Articles: week of May 22, 2022
1) Five Canadian startups crack CB Insights’ 2022 AI 100 list (BetaKit)
Six Radical Ventures portfolio companies were named in CB Insights’ sixth annual AI 100 list, which aims to highlight the most promising and innovative private AI startups in the world. Three of our listed portfolio companies are based in Toronto, including Cohere, Untether AI, and Waabi. Other Radical Ventures portfolio companies that made this list include San-Francisco-based Twelve Labs and Crossing Minds, who opened an office in Toronto in January, as well as an announced company.
2) How AI startup Cohere landed $170M of funding in a year (StartUp Here)
“We started off the year with around 20 team members and we currently have about 75, so we’ve had pretty dramatic growth.” Aidan Gomez, CEO and Co-Founder of Radical Ventures portfolio company Cohere, was recently interviewed on the company’s rocketship growth over the past two years. A year ago, the company emerged from stealth mode and quickly secured two funding rounds. Among Cohere’s investors is AI Godfather Geoff Hinton, in whose Google Brain Toronto lab Cohere founders Aidan Gomez and Nick Frosst met. Cohere has recently expanded. First into the US with an office in Palo Alto headed up by Bill MacCartney, former director of proactive intelligence at Apple, and more recently in London, led by Phil Blunsom (who ran NLP at DeepMind and is a Professor at Oxford University) and Ed Grefenstette (who worked with Phil at DeepMind before opening the Facebook AI Research office in London). Cohere’s growth reflects NLP’s growing use cases and a rapid desire for adoption from businesses of all sizes.
3) Scientists identify characteristics to better define long COVID (National Institutes of Health)
NIH-backed scientists developed machine learning models to identify patients with suspected long COVID. The models find differentiating factors between patients with long-COVID from those without. To train three machine learning models, scientists drew on data from the electronic health records of nearly 100,000 patients who tested positive for COVID, including nearly 600 diagnosed and treated for long COVID. The AI system confirmed risk factors for long-COVID include preexisting chronic diseases such as diabetes and chronic kidney disease, as well as long-term symptoms such as chest pain, malaise, and sleep disorders post-infection.
4) Is DeepMind’s Gato AI really a human-level intelligence breakthrough? (New Scientist)
Researchers took a step toward achieving the goal of human-level intelligence by delivering one model that performs many different tasks. Named Gato, the general-purpose AI can complete 604 different tasks, including captioning images, engaging in dialogue, and playing games. It performs 450 of those tasks better than an expert more than half the time. The question remains if amortizing what 600 distinct, smaller networks can do into one model constitutes as human-level or general intelligence. The model still has the same weaknesses as other AI models, such as having biases and an inability to remember context. AI luminary Yann LeCun is quoted: “We still don’t have a learning paradigm that allows machines to learn how the world works, like human and many non-human babies do.”
5) The future of baseball might include robot umpires (Discover Magazine)
The Automated Ball-Strike system (ABS) tracks pitches using radar to provide a reliable, precise, rule-book strike zone for pitchers and hitters. Far from a sci-fi image, the robot looks like “a large black pizza box” mounted behind the plate. The controversial robot has pushed a debate on what fans want from their sports entertainment: “for a hundred years, you go to a game, you have a hot dog, you yell at the umpire. Who’s going to yell at a computer?” Cornell researchers have been studying an AI system in the context of chess that better understands that humans make mistakes intentionally. An argument is being made to employ a robot intelligent enough to make calculated mistakes.