This month, the Centre for Social Impact ran its final brown bag lunch of the year. We had the pleasure of hosting Dr. Stephen Thomas, who is currently an Adjunct Professor right here at the Smith School of Business, as well as the Director of the Master of Management Analytics (MMA) and Master of Management in Artificial Intelligence (MMAI) programs.
Dr. Thomas spoke on his views of the roles of artificial intelligence (AI) & analytics in business, as well as how they will impact us moving forward. As a vastly complex topic which touches nearly every industry, he noted it was very difficult to pack a semester’s worth of information into a very short 45 minutes.
It is no secret that AI is exploding. While it has a long history spanning back to the mid twentieth century, in the past decade huge investments have made to introduce rapid growth in the field. Everyone from local grocery stores to federal governments have carved out their stake. The technology touches our lives in ways that we don’t even notice. Yet, as individuals the majority of us possess very little technical knowledge on how any of it works. It seems to be one of those facts of life that is too frightening to even contemplate, like where socks go when you put them in the dryer.
Simply put, AI itself is when computers attempt to do human things. The most pertinent images that we have are of self-driving cars, robots that beat world champions at chess, and “deep fakes” that possess the ability to place Nicholas Cage in any movie.
Of course, it all goes much deeperthan that. With six central disciplines, that each manifest themselves in countless ways, AI really is everywhere. There is software that makes it possible to do away with the manual work of programming, push computers to understand images, and much more. There are also physical robots that we can get to do things for us, which are already vastly popular in the manufacturing world.
On the other hand, a much less flashy and infinitely more important area is the overlap of AI and analytics. This is when computers use data to attempt the most complex human task that there is: decision-making. With machine learning at the centre of it, this kind of AI ends up being the focus in a business context. In fact, it is rare to find any company in any industry that isn’t currently taking the steps to invest in this kind of technology, if they have not already done so.
We live in a world that simply has access to loads more data than we ever have before. Thanks to cloud technologies, and the fact that no one is safe from the internet, computers have more to work with than they can even handle. This is what has amplified machine learning.
Firstly, at the simplest level, computers can learn supervised by building off of comprehensive past data. This makes possible things like web ad placement, fraud detection, and much more. If you have ever ended up buying more than you planned for based on online recommendations, or been saved from having your credit card declined just because you were on vacation, then you have AI to thank for it. Take that how you will.
When data gets a little more complex, computers can learn unsupervised to find patterns in the data. While it is both more expensive and difficult to engage at this level, it can have incredible results. Using the success story of credit card fraud, the use of machine learning led to the development of fraud detection programs that have created 60 per cent in total savings for an industry that was once losing $80 billion annually to fraud. On a smaller scale, and in a non-business context, a hospital in Singapore even used unsupervised machine learning to save 100 lives each year when the software found patterns in patient treatment.
At the highest level, machine learning can be based entirely on reinforcement. With almost no data to work with, this is when a computer can develop by learning from its own mistakes. This is where you see technology like self-driving cars. Of course, it can appear as almost dystopian. I don’t see anyone being too comfortable with getting into a car that will accelerate at any moment because it doesn’t yet know that it shouldn’t. However, on a grander scale, while trusting corporations to do their due diligence, this kind of machine learning gives rise to the idea that we can learn together. For instance, as an individual driver when you improve your skills, no one else improves with you. However, when your self-driving car learns how to drive more safely in the winter, the data that it collects will be available to an entire network of cars who will learn alongside it with each movement that it makes.
While it would take far more than 45 minutes, or even 45 semesters, to learn everything that there is to know about the ever-changing world of AI, one thing that is clear is that it is becoming essential to how business is done. AI can create significant value, and there are now increasingly negative consequences to neglecting the technology. As business students, it is important to engage with this topic as much as possible, and critically consider the fact that the addition of technology to commerce has set a precedent for rapid innovation. You have to be a lifelong learner above everything else, because from now on things will not be the same from one year, or even one week, to the next.
Just as important to fostering innovation is to consider the challenges that come alongside it. Instead of accelerating, as managers we must push the brakes every so often and ensure that we fully understand the decisions that we are making. AI has already landed some corporations in hot water, like when the adoption of resume-reading software by Amazon made it impossible for female applicants to be considered. Like all technology, and like everything in business, AI is a continuous and engaging process. You never stop learning, and you never stop asking questions.
We greatly appreciate Dr. Thomas’ time and efforts in setting the Smith School of Business on the fast track to innovation and we wish him all the best in his journey!
Written by Aysha Tabassum, BCom 2022