From Jeopardy to sports: IBM’s cognitive computing vision
- An interview with IBM exec Noah Syken on the ways cognitive computing tech can be used in the world of sports.
Read about the latest sports tech news, innovations, ideas and products that impact players, fans and the sports industry overall at SportTechie.com.
IBM’s Vice President of Global Sponsorships (Sports and Entertainment) and Client Programs, Noah Syken, was kind enough to share some of his time with us to speak about all things cognitive computing and Watson. Syken is probably best known for his role in “Jeopardy: The IBM Challenge,” in which he took over the marketing side of things for the television program where super computer Watson competed against some of Jeopardy’s best players.
Nowadays, Syken is focused on the world of sports, and finding ways in which the technology displayed on Jeopardy can assist sports organizations in helping players and staff. With big time deals doing analytics work for teams like the Toronto Raptors and huge sporting companies like Under Armour, the power of Watson technology cannot be ignored.
Below is an interview with Syken where we learn how intricate this cognitive computing technology is and its current and potential future impact on the sports industry.
What is cognitive computing? How does it work?
Without getting too long into the history—if you think about where computings come from, we’ve come from the early 1900s mechanical computing systems where punch cards were used for the calculations moved into the programmable space … Now we have this notion of cognitive computing systems that understand reason and learn—computers that actually learn from their experiences, computers that are actually taught. So you need to actually teach these systems a base level of knowledge.
These systems can handle massive amounts of data, really huge volumes of data that never existed before. We need these new types of cognitive systems that can handle this type of data, and new kinds of data. Data that is both structured and unstructured—unstructured being videos, sound, photographs, tweets and everything everybody is doing out there.
The need is to understand the structured world of data and the unstructured world of data, and also being able to interact with the data, in ways that people communicate using natural language. In the past, in order to operate systems, you needed to be a programmer or needed to be able to code, and the evolution here is to be able to handle this volume of data and engage with systems in a way that works for humans.
Why is it big news when Watson works with a sports organization?
To some extent, it takes some of the subjectivity out of the world. One of the keys is that these systems aren’t really designed to replace humans. They’re designed to assist humans. When you think of the general manager of a sports team, there are so many different sources of information, whether it is the league stats, the salary and front office information related to a team, not only your team but all of the other teams in the league, or all the teams and players that are in the minor leagues coming up.
There is all this other information that can add powerful insights, so a system like Watson or cognitive computing can read and interpret all of the press conferences of players over the course of this year, last year or the past five years, and help general managers understand the mindset of that player. Is that player becoming more positive in how they talk to the press? Becoming more optimistic about their game and what is going to happen on the court? That is important for a GM to know. All that stuff can be brought together in a really simple, user-friendly interface, like what we have for the Raptors.
As far as a GM or scouting role, could Watson replace a human in the future?
I don’t think so, that’s not really what we see. It’s more augmenting decision-making and not so much replacing it. The ability to take information, organize it and present it in a way that humans can make an assessment is really important.
Humans bring a natural bias to certain kinds of attributes, we all have our favorite players, we all have our favorite teams and we all grew up in a particular city. All of those things color how the human looks at the objective data. But if a cognitive system can really scour all of that data and place an objective frame on interpreting that data, and then present it in a very consumable way to a human, that combination of a human and a cognitive system really are going to lead to better outcomes.
If you think about the Ryder Cup coming up, the Ryder Cup captains are going to have to put their players in pairings. Those captains have been on tour with the tour pros for a long time. They know who their friends are, they know what folks they may not have an affinity for, they played with some of these guys in the sun and the rain, etc. These captains will bring a certain bias to their pairings, but if they are helped by a cognitive system, then they can erase all of those things.
How can cognitive systems enhance fan experience?
We are in the very early days of this space. But certainly we think that these systems can add pure customer service. Today, if you are going to an event, you might go to a team’s website and scour through the FAQ’s to try to find what train you need to get to the stadium on or what time does the event start or what you can bring in … well again, cognitive systems can interact with humans in natural language.
So, why can’t I just go to an engagement advisor for the local baseball team or the upcoming U.S. Open event, which we will actually be doing this at, and provide a set of cognitive systems to help from a customer service perspective.
The stadiums and the facilities themselves should also be cognitive. Some of this stuff is happening just based on mobile technology, but you shouldn’t find yourself standing in a long food line or bathroom line, or not knowing where the best parking is.
Teams should have an understanding of how their consumers choose to engage in events. We are starting to use Watson in the programmatic advertising space. So, the sports properties themselves should be able to deliver more relevant messages and more relevant engagements through cognitive systems that know more about their fans.
Could Watson be used to actually train elite athletes?
One of the things we are thinking about is the equipment itself that athletes use, whether it be a tennis racquet or a golf club. Babolat makes a great connected racquet that starts to kick off information about forehands, backhands, spin-rates, etc. That is all great information, but if you can bring cognition to that information and actually relate it to the rest of the world … if you can model how an up-and-coming athlete compares to those that are best in the world, that is a big opportunity for training. In the same way, being able to understand the connected devices and equipment is important. This can allow up-and-coming athletes to model their game based off of world-class athletes.