Big news for AI in Canada: Geoffrey Hinton, one of the "godfathers of AI," has been awarded the Nobel Prize in physics for his pioneering contributions to AI systems. In other AI headlines, St. Michael's Hospital in Toronto is using AI to save lives, while concerns are growing over the rise of AI-powered scams targeting vulnerable populations.

Big news for AI in Canada: Geoffrey Hinton awarded the Nobel Prize

University of Toronto AI scientist and one of the "godfathers of AI," Geoffrey Hinton is being awarded the prize in physics alongside John Hopfield, for their pioneering AI work that brought us the systems we have today. AI was a bit of a theme with the Nobel committee this year, with the prize in chemistry being awarded to the group of scientists involved in creating Google's AlphaFold2 that was able to predict the structure of 200 million proteins.

The AI "grandparent scam"

The insidious use of AI is something we all need to be vigilant about, especially as malicious actors become more creative in how they target vulnerable populations. Timmins police recently warned about a popular scam targeting grandparents. Scammers (though that word feels too soft for these types of scams) are using AI to mimic voices of loved ones to con the elderly into paying ransoms. It is good practice to make sure your elderly family members are aware of these dangers. I've heard of some folks opting to come up with a codeword that parents and grandparents can ask for if they receive a suspicious call that sounds like a family member.

One Toronto hospital reduced unexpected deaths by 26 per cent

However, in a good news story about AI, St. Michael's Hospital's AI-powered early warning system, Chartwatch, is saving lives. The technology will alert healthcare providers that there may be a patient issue to investigate before it is obvious, leading to earlier interventions. Dr. Bo Wang, Chief AI Scientist at UHN, said in a recent interview said the biggest areas for AI impact in hospitals are in imaging, natural language processing (clinical notes, documentation etc.), and personalization, such as analyzing genetic information.

A recent NBER working paper looking into the adoption of generative AI, finding that nearly 40 per cent of survey respondents had used generative AI at work or at home. Surprisingly, slightly more respondents (32.7 per cent) used GenAI outside of work, compared to for work (24.1 per cent) in August 2024. Given the timing, perhaps workers were using ChatGPT to help plan their summer vacations. Writing communications and interpreting/translating/summarizing were in the top 3 highest-ranked uses both at and outside of work. Other top uses at work are performing administrative tasks and coding software, which outside of work personal assistance and entertainment recommendations are deemed helpful.

OpenAI's new models releasing models aimed at scientists, researchers, and coders

Premium users of ChatGPT now have access to two new models with ChatGPT. The o1-preview model is slower, but it uses Chain of Thought (CoT) logic to "think" through an answer more deliberately allowing it to solve harder problems, especially with science, technology, engineering, and math (STEM) questions. It became the first model to beat PhD-level scholars on the hardest questions on a benchmark that tests the performance of large language models. The new canvas model turns ChatGPT into more of an assistant where you can work together on improving code or a writing piece. In this mode, a panel opens up on the side of your screen with the text or code you want to work on together. Google also released a new tool aimed at researchers in September with the release of NotebookLM, an AI research assistant that lets you upload a document or PDF and generate summaries, briefing notes, FAQs, and even a podcast on the material.