Photo: courtesy

By Pooja Sainarayan

Local Journalism Initiative

AI technology allows machines to learn from experience and adapt human-like intelligence. The reality of Artificial Intelligence (AI) is all around us, from its use in banking, GPS guidance, smart home devices and generative AI tools like chat GPT. Humans have been toying around with AI for several decades, however the implication of the technology is evolving now like never before. AI can be traced back to the 1950s, from the design of chess-playing computers to the first artificial neural network. There are two major subgroups of AI – weak and strong. Weak AI, also known as artificial narrow intelligence (ANI) is trained to perform very specific tasks. This is the type of AI that is most common in day-to-day tasks today, some examples include Amazon’s Alexa and self-driving vehicles. Strong AI on the other hand, is a theoretical form of AI where a machine would have an intelligence equivalent to or even greater than humans. It would be self-conscious with an ability to learn, solve problems and strategize. Strong AI only exists in science fiction for now, but research on its development is ongoing. AI technology often goes hand-in-hand with deep learning, which is closely modeled after the adaptability aspect of the human brain, to develop AI algorithms in learning from accessible information and perfect its ability in making predictions over time.

Several cities in Canada have been implementing deep learning AI technology for various projects. For example, Edmonton has integrated AI with remote cameras to monitor wildlife coming into the city. Since 2022, Alberta has been using AI tools to analyze data points and foresee where new fires are most likely to occur the following day, giving firefighters a head start. A powerful tool worth research and improvement, as tech partners predict this investment could save up to 5 million dollars a year. Montreal is experimenting with Fujitsu, an AI tool used to analyze the traffic flow of over 2000 traffic lights in order to help the city take proactive measures in decreasing traffic-related issues. Apart from increasing the flow of traffic and reducing air pollution, it can also help the city plan maintenance routes for snowplows or other service vehicles more competently. In addition, Montreal’s transit agency is planning to use AI to monitor the CCTV footage to recognize any signs of public distress in efforts to prevent suicide in the subway system.

Recently, municipalities in Quebec City have been adopting AI tools to track everything from trees, cars and backyard pools. The Communauté métropolitaine de Québec (CM Quebec) which encompasses Quebec City and its suburbs, states that this project will help municipalities monitor urban growth, parking availability and environmental goals. The geomatics development manager of CM Quebec, Frédérick Lafrance mentioned the organization has worked with deep learning AI technology using aerial photos of the city to identify buildings, swimming pools, backyard trampolines, cars and various other features. As expected, the AI would be able to analyze larger data sets of the aerial photos in a shorter time frame compared to humans, to get more work done in less time. This AI-generated data analysis can be used in several ways, such as measuring urban greening and tree cover versus how much of it has been converted to asphalt over time, said Lafrance. Tracking backyard pools and such features can help the city coordinate inspections. However, the use of AI as a surveillance tool is very different from having an inspector perform the duty, so it remains to be seen how the public reacts to this change. In this case, the AI is using already generated aerial footage to differentiate objects, and not digging into further information such as licence plate numbers or the make and model of any objects.

Interestingly, on the other hand, the impact of AI can go beyond measuring trees and backyard pools. In a 2017 U.S study, an AI deep learning tool was used to characterize the make and model of cars in millions of pictures from Google Street View. Researchers found that in cities where sedans were the majority over pickup truck vehicles had an 88 percent chance of voting Democrat whereas cities with more pickups had an 82 percent chance of voting Republican. Findings like these serve as an important tool for policymakers and pave the way for ethical and sociological questions.

Despite having a relatively brief history, the technology has shaped our lives like nothing ever before. As AI technology grows more and more powerful, we can only expect its impact to increase with the years to come.