Last week, Canada announced its long-delayed national AI strategy, encouraging the development and safe use of this important technology. There is one problem. The strategy as written will ensure only one thing: a weaker, less prosperous Canada.
Why is that? As a starting point, Canadians should know four things about AI. First, today’s AI is the worst it will ever be. Second, most top AI researchers believe it will be able to do anything a typical human can do on a computer, at a minimum, by 2030. Third, as AI contributes to science, robotics, and self-improvement of AI models themselves, national defence and industrial strength will require access to, and use of, the best models. Fourth, there are only two countries with organizations even trying to make those frontier models: the United States and China.
