Applying AI to business and management processes―across all levels of an organization―offers huge potential. Potential to create more agile and adaptive ways of working, enable better decision making, ramp up operational efficiency, and even to identify and capture completely new opportunities.
Realising this potential is now a priority for companies across most sectors—as they look to stay relevant and stay competitive.
Before developing an AI strategy and implementing it, leaders need to identify where AI, data analytics, and machine learning can fit into their organization, and how they can use those tools to create value. The first step is to assess the current landscape of AI.
The Lay of the Land
Matthew Mitchell, Professor of Economic Analysis and Policy at Rotman School of Management, offers his assessment: “It’s all over the map. There are companies that are the leaders on earth in understanding what these technologies can do, and who are already implementing them. This includes the usual suspects of tech companies and companies in the marketing sphere, and companies that have a big customer service aspect to what they do—like banks and airlines. We also see a lot of companies saying, ‘Our company needs an AI strategy.’ Often they don't really know what that AI strategy would be there to accomplish, and they can be sold a lot of snake oil about what these technologies can do for them, because they don't know.”
Join Matthew Mitchell on Rotman’s executive program: Put AI to Work: Managing with Machines to transform complex business challenges into new opportunities
Dates: June 16–17, 2020 │ Format: In-class study │ Location: Toronto, Ontario
Considering the current state of development of the technologies themselves and if they are at a mature stage now, Mitchell says: “Again, the answers are all over the map. I think you could use the word ‘mature’ in the sense that people know how to deploy these technologies, and the technologies do improve efficiency for lots and lots of firms in some arenas. In other arenas, I think it's really, really far away.”
The key to getting started with AI and machine learning is to have some people at the firm that understand broadly what AI algorithms do. That these algorithms, as Mitchell points out, “are mostly trying to solve categorization problems—to decide, based on data, what category something comes from―whether it’s a customer, or a product, or some other data-point comes from. If someone contacts your call centre, the algorithm can figure out from the words they've heard and computed, what this person's problem is—and decide how best to proceed, whether to ship them through to a human, or to another kind of solution.”
Call centres are a great example of what the technology can do. It can categorize people, and tasks, and problems—and help determine where resources should be deployed. Who, in the case of a call centre, are the people that really need to talk to a human?
“In targeted marketing this technology is useful in categorizing what kind of a buyer this is. In some firms, AI is helping with HR decisions too―screening candidates. Understanding what problems AI solves well will help you understand where it can add value in your organization.”
“There isn't a one size fits all solution. You can’t simply drop this technology into your HR department, or into your customer service department. It starts with having some people at a high level knowing what this technology can do.”
Start with the Problem
Mitchell believes rather than seeing this through the prism of ‘new technology’, you need to start from the business problem you want to solve. “It has to start with someone with a problem, and that person having an understanding of how these algorithms work, so that they can communicate with someone with practical knowledge, so that together, they can implement these processes.”
How to acquire this knowledge can depend. While a big bank may have an entire department of technical people with a thousand employees who have data science and machine learning backgrounds who are attacking all business problems in the firm with their technical training, for lots of other firms these technologies are things you don't even necessarily have to implement in-house. There are things that you can outsource and buy from firms that have the technical knowledge. “The key thing is to figure out when you have a match between a business problem and this kind of technology,” says Michell. “If that match isn't there, then it's not going to succeed no matter how technically sophisticated your data people are. And the truth is, if you have a great match, you might not need to employ your own technical people at all.”
This concept is central to the thinking behind the program at Rotman led by Mitchell: Put AI to Work: Managing with Machines. The idea is to touch on, in a non-technical way, some of the ideas about how modern AI works and when it doesn't. Participants will be people who need to figure out how AI can fit into their firm. “We're not really looking for super-technical people that want to learn the absolute nitty-gritty of how these algorithms function. We’re looking for the individuals at a decision-making level that I mentioned earlier—the leaders who are looking for an understanding of the technologies, the landscape, and with problems to solve.”