Skip to main content

Chris Dyck, program co-ordinator for Georgian College’s Big Data Analytics program, is passionate about exploring the implications of human-machine interactions.SUPPLIED

Machines already cover a wide variety of physical work tasks. They are also gaining more and more capabilities that allow them to outperform humans in areas of the cognitive realm. Just consider a computer’s ability to dominate at chess, translate spoken language into text or analyze big data sets.

As AI technologies mature, their potential for replacing humans for certain types of work is growing. Yet people will always have a role – and responsibility – in guiding technology development and machine learning, says Chris Dyck, program co-ordinator for Georgian College’s Big Data Analytics program. “The future is shaped by technology capabilities, but to make the technology work, people have to train it.”

Any demonstration of functional AI typically comes out of many years of training with data sets. For example, when people are asked to identify objects for online security features, they unwittingly contribute to machine learning, says Mr. Dyck. “If you are directed to a form that asks whether there is a car in the picture, it is part of a training algorithm – and billions of feedback forms then enable AI to get better at identifying cars.”

Similarly, a large number of examples would be required to train technology to identify bird calls, an application Georgian College is exploring in partnership with the Ministry of Natural Resources and Forestry, says Mr. Dyck. “The project involves placing devices in conservation areas to pick up bird sounds. This can provide information about which and how many birds are in the area and for how long.”

This exciting use of technology also illustrates the kind of training AI requires, he explains. “There are thousands of species with unique sounds in the area. Rather than using one example of a bird call, we need thousands, since each bird sounds different.”

With the amount of human- and machine-generated data growing exponentially, the question is how to leverage it for the most meaningful outcomes, says Mr. Dyck.

Data science, which uses automated methods to analyze massive amounts of data, can provide knowledge for stakeholders to enable informed decision-making, predicting trends and understanding customers better.

These variety of uses mean that Georgian College students enrolled in the nine-month Big Data Analytics graduate certificate program gain skills that are applicable – today and in the future – in a the wide range of fields, including government, applied research, human resources, health care and marketing, says Mr. Dyck. They also benefit from another key collaboration. “We were the first in Canada to start a program with Microsoft. That’s a big leap forward in our ability to adopt cutting-edge technology and train students so they can hit the ground running,” he says, adding that Microsoft AI technology has been found the “most human-like.”

Mr. Dyck shares his passion for exploring the implications of human-machine interactions with enthusiasm. He invites his students to use data analysis to probe the connection between certain topics and sentiment in news articles, and to investigate why AI often fails to recognize people’s gender. “This raises questions of how people identify gender and why a computer can’t identify certain things well,” he says. And if AI tends to identify gender as neutral, could that lead to media and society moving towards a reduced emphasis on gender as well?

“I am always intrigued by how we influence data and how data influences us,” says Mr. Dyck.

Advertising feature produced by Globe Content Studio. The Globe’s editorial department was not involved.