Renowned Data Scientist Avi Goldfarb to Keynote BMC Exchange Chicago
We are thrilled to present Avi Goldfarb, Ellison Professor of Marketing at Toronto’s Rotman School of Management, University of Toronto, and co-author of Prediction Machines: The Simple Economics of Artificial Intelligence as keynote speaker at BMC Exchange Chicago. We spoke to Avi for a sneak preview of what he will discuss with our audience—read on for some key insights.
BMC: Your book has earned great responses from people in all industries who are navigating digital transformation efforts thanks to its clear and coherent advice. You mention four aspects of decision making that are critical to all businesses today.
Goldfarb: Right, the four aspects of decision-making are prediction, action, judgment, and data. Prediction is the ability to fill in missing information. It helps businesses make better decisions. Action is the ability to do something with your decisions. There is little value in making a decision if you can’t take an action. Judgment is the ability to know what decisions to make. It is the process of determining the payoffs to various actions. Finally, data is the key input to the entire process. Without data, better prediction and better decisions would not be possible.
BMC: What role does artificial intelligence (AI) play in prediction today that it hasn’t in the past?
Goldfarb: So even though there’s a lot of recent excitement about AI and what it means for business and society, we remain a long way from truly intelligent machines. We have been 20-50 years away from truly intelligent machines since 1956. The reason we talk about AI today is that one specific aspect—machine learning—has itself gotten so much better. It allows far better predictive ability now than in the past. It enables you to fill in missing pieces—that’s the data—and it helps make better decisions.
BMC: You’ve talked about Amazon’s predictive ability as an example of this—as well as an ongoing opportunity.
Goldfarb: Amazon’s business model is essentially no different than a Sears catalog right now. Based on your buying history, they make a guess at what else you might like and they present it as suggestions when you log on to your account. With millions of items in their catalog, their suggestions are pretty good —say, they’ll suggest 20 items and you might buy 1, but not good enough to transform the business beyond being an extraordinarily well-run catalog company. If the machine learning continues to improve and the recommendations improve through new data, they could redesign how they operate. If they were certain enough about what you want, they could ship the item to your door before you order it. It moves from shopping then shipping to shipping then shopping. Better predictions reduce uncertainty, and that creates opportunities.
Another example is the airport lounge. We cater to frequent flyers in airport lounges, give them snacks and drinks to reward them essentially for uncertainty—we don’t know how long they’ll have to wait for their flights, and we want to keep them happy. But if we could remove that uncertainty and keep them from having to wait at all, how much happier could they be?
It all comes down to better prediction. As long as the cost of prediction—and that includes the full burden of work associated with it, from data-gathering to manpower to software costs—as long as that cost continues to go down, the ability for innovation will go up. And that’s how industries are transformed.
BMC: You’ve discussed Steve Jobs on stage in 2007, introducing the iPhone… and who could have guessed that out of that pocket innovation would come the rideshare economy and the decline of the taxi industry.
Goldfarb: Right, and we don’t know what the next frontier will be. We can only guess, based on the innovations that are unfolding right now, what we see and use every day that may become obsolete or fundamentally transformed.
BMC: So what about the last two components to decision-making: action and judgment?
Goldfarb: The most important things for most businesses to remember are simple: identify where uncertainty is a bottleneck. What could you do better with more information? What are your constraints because of bad predictions? And then think from an AI-first perspective: what are the new opportunities for prediction, and what are you willing to take on at the expense of other things?
BMC: Sounds like something people who are interested in cloud migration and security might be interested in…?
Goldfarb: Security is fundamentally a prediction problem. Solve the uncertainty, predict the possibility of failure, and security is ensured.
With cloud, there are concerns about speed, about latency. If you secure the data to predict what you will move and where you will move it, you remove the uncertainty and you speed up the migration process and probably lower costs, too.
To hear more of Goldfarb’s insights and information about his current research, don’t miss his exciting keynote at BMC Exchange Chicago on June 6! To learn more about BMC Exchange, a global IT management event series that brings practitioners and experts together to share technology insights and product knowledge, visit https://exchange.bmc.com.
These postings are my own and do not necessarily represent BMC's position, strategies, or opinion.
See an error or have a suggestion? Please let us know by emailing email@example.com.