Cognitive Computing is rapidly changing how people interact. As a result, this creates opportunities and challenges for brands across all industries. So we asked IBM’s Director of Product Management for IoT, Luis Rodriguez, back to discuss.
Rodriguez first joined the Brand Lab Series™ in episode 43 to talk IoT disruption. Therefore, he was a logical choice to talk cognitive technology. So what excites him about cognitive?
I’m super-thrilled to have Luis back. As you know, his first episode was absolutely fascinating as we talked all about the Internet of Things, how it’s disrupting business models and entire industries. He’s just a wealth of knowledge, perhaps the smartest guest we’ve had on our Brand Lab Series, and we’ve had some fabulous guests before. I’m thrilled to hear what he has to say today about cognitive computing.
Yeah, it’s good to have him back. Actually, he made his first appearance with us on episode 43. You can go back and listen to that episode and to all of our other episodes also.
And yes, and you can find episode 43 and all of those past episodes in iTunes, Google Play, as well as iHeartRadio. Luis, I’m super excited to have you back on the Brand Lab Series. I was joking with Natalie before we started recording that you might be the smartest guest we’ve ever had on our show.
[chuckle]
Not one, not two, but three degrees from MIT. You have a great career over the last almost 30 years, including some time obviously, most recently with IBM. What has been the most interesting change you’ve seen personally as it comes to technology over that period of time?
Yeah, and thank you for that. I think really the biggest change that’s happened is the speed at which people can innovate. And when I started my journey, there weren’t databases, you had to build your own. There wasn’t cloud, you had to build your own system. And now we’re in a position where anybody can just sit down, get free accounts on a cloud system, drag and drop systems, whole systems, not just databases or apps, whole systems, build mobile apps with ease, with little to no programming. Kids are doing things that it took armies to do before, and that ability to just innovate on demand, rapidly, anytime/anywhere, is changing the dynamics of our entire planet. And to me, that’s the most exciting change that’s happened is the kind of the democratization of innovation.
Yeah. As a parent myself, I’m stunned, and I know Natalie is too, how tech savvy and how quick our kids seem to understand and can leapfrog us on some days in terms of their understanding and uses of technology. I know in our last episode, we talked about how the IOT is disrupting business models and entire industries, so why don’t we start, before we get into our conversation on cognitive computing is, what are some of the misperceptions around the Internet of Things?
Yeah, there’s a couple. One of ’em, which I find pretty pervasively is that people say, “Oh, it’s just a fad. It’s just a term that’s out there.” And the reality is from what we’re seeing, we’ve got hundreds of use cases, and we’ve got customers that are vetting these use cases, that are implementing the use cases, or already actually using these use cases. So it’s not at all a fad. Another one tied to that is people say, “Oh, whatever it is, it’s hard.” It’s not hard. You can do it at home today with little or no programming experience. You can start doing Internet of Things on your home, small businesses. And then, the third one which I think is really critical, is the misconception that, “Oh, you can just wait.” And in some industries like manufacturing, that can be a real problem cause manufacturing physical things, there’s a life cycle there.
Whether it’s cars or durable goods or aircraft engines, if you understand how your customers are using your products, and you’re feeding that back into your production cycle, your design cycle, that can get you ahead. And if you’re sitting there thinking, “I’ll just… Internet of Things. I’ll just… It doesn’t matter.” Your competitors, who are listening to your customers in ways that you’re not, are gonna be able to get ahead of you. I think those three things are three of the most, I’ll call it, misconceptions, mis-receptions around IOT.
See Natalie, so you can start the IOT at home.
There you go. Good to know.
Yeah, like I joked, we’ll have a quiz for you after the end of today’s episode.
[chuckle]
Why are data and analytics critical to IOT success?
It’s one of the most important questions here, and so for IOT to be successful, you have to have insight, and so let’s start with that point. You have to have insight, whether it be insight of how your products are being used or insight about what’s happening inside a particular building, insight when even things… We have customers that are asking, when and where will the next crime happen? Or how much damage is my kid’s brain taking from the football practice? Or my eating habits what’s that leading to? How many supplies? How much Stack should I bring on board this week? Cause of my small business in a corner coffee shop. So that insight is what drives the value around IOT, and so what data does is it increases the amount of insight that you can have. And so we’re seeing things on data, and it’s not just structured data, we’re seeing things where you’re bringing unstructured data in as well; so unstructured data are things like video or audio or written text, things that aren’t time stamps or measurements. And when you bring all that together you can get better insight into what’s happening.
If you’re in a building and some emergency happens and you have audio feedback, you can triangulate what happened. Let’s say some strange noise happened in a building and you’re starting to smell smoke, or where did that come from? Now you can start triangulating it, you can start listening or reading to what’s happening on the street and figure out is some bad situation escalating or is something good happening? And so that’s the data piece of it and analytics is what changes that data into the insight, and so we have an evolution that’s happening with analytics, used to be very simple and still there’s a lot of simple-use cases where you say, “If this then that.” And that’s one of the simpler views of it, but then you start having things like predictive analytics. So you start looking at a motor and start understanding how much energy it’s using, listening to what’s happening, how it sounds, looking at vibration, and then being able to predict when that motor’s gonna go bad.
With the right analytics you can have big savings, there’s a customer that we have that has a billion dollars of food spoilage every year, and a large part of that’s due to not being able to predict easily when refrigeration units are going down. And if you can predict that then it causes a lot less spoilage, which is always a good thing. And then you have cognitive, so you have cognitive analytics. So instead of analytics that are programmed they have analytics that can learn, and that’s where the next big frontiers happening around analytics, and that’s what’s gonna take IOT really to a whole new level.
Well let’s build off that last point related to cognitive computing, which I think is a fascinating topic on its own and I know one half of your very impressive title at IBM is all around Watson as well. I recently just saw the movie Hidden Figures, which was about the space program back in the day, and I’m always reminded at the incredibly rich history IBM has been kind of at the forefront of computing and technology. Talk a little bit more about Watson and cognitive computing, and how that’s really evolved and changed so much over the last couple of decades.
Yes. So Watson is IBM’s entree into cognitive computing and hopefully the listeners here have seen at least the initial history of that where the use case that was built out was, can Watson compete against Jeopardy champions? So we all know in Jeopardy you have to understand not just English and be able to parse English, but you have to understand English in context; you have to be able to determine when something is a pun versus something should be taken literally. And so our scientists spent years designing a system that was a cognitive system to compete there, and what’s the difference between a cognitive system and a regular programming? Well cognitive system is not something that you program to generate some number like 42, cognitive system is something that is a system that you teach and that can reason.
So when it comes back to you with an answer like on Jeopardy, when it came back with an answer, the answer was… It actually came back internally with several answers, with probabilities of whether those answers were correct or not. And one way that I like to position that is imagine that you go to a meeting, you’re sitting in some space there in a conference room and you hear a ding, how does your brain interpret that? Well the brain depending on how the ding sounds it may say, “Hey that sounds like an elevator ding and I’m near the elevators, so that’s probably what it was.” And if somebody asks you, “What was that noise?” You’d probably say, “Oh those are the elevators.”
Are you 100% sure? No, there’s a probability that you have in your mind that you’re right, but you don’t have 100% certainty and you’re not gonna go and get up and go try to figure out whether that’s true or not. Well Watson is a system like that, it takes in context, where I am, what’s near me? It takes in data including unstructured data like audio, and then it makes inferences about what’s happening around it, or it tries to answer questions that you ask it and then it learns over time. So if it’s in a different situation instead of you’re sitting in an office building, you’re sitting somewhere else and you hear a ding, well maybe that ding is because there’s a microwave nearby cause you’re in a kitchen; so now it learns, “Well if I’m in a kitchen, then it’s more likely to be a microwave ding than an elevator ding.” So these systems that learn, and so Watson is one of them, they evolve over time and they get better over time and they learn over time. And so this is what the evolution of Watson has become. It’s gone from being a Jeopardy champion to becoming something that learns specifically about certain areas like medicine or oncology.
So, building off this conversation around cognitive computing which is absolutely fascinating, who are either some companies or maybe some industries that you see are starting to do this in a really interesting way?
Yes. There’s some really interesting use cases. So, let’s start with medicine. So, medicine is one of those areas that is based on averages. So, you go out there and you have a new drug that you want to apply for example, and so you test in on hundreds, maybe thousands of people and then you say, “Oh, if it works on those several thousand people then it should work on hundreds of millions of people.” And instead, what’s happening with cognitive is we’re starting to have entities like Watson, be able to read all the different studies and be able to understand people’s genetics and be able to get much more hyper-personalized, and be able to recommend treatments that… Or be able to diagnose things that are much more personalized to a specific person. That’s a game changer. That’s a game changer for humanity. And it’s anything from spotting cancer, being able to look at a spot and say, “Is that really melanoma or not on a skin?” To being able to recommend treatments that are specific to your genetics and your lifestyle and so on. That industry is already starting to change based on Watson or cognitive ability to consume all the data that’s out there that no one doctor or team of doctors is gonna be able to consume and to reason about overall.
Hospitals. So that’s a related one. So there’s another interesting industry. So one thing is having Watson in the walls. So, if you have cognitive computing in the walls and you’re a patient, you can start interacting with your room. Is it too bright? Is it too hot? Is it too cold? Are you hungry? Who’s visiting you? Is something wrong? And Watson is listening. You don’t even have to say a word. Maybe just by listening, Watson can tell that something isn’t right and summon a nurse or a doctor. Let’s take that to hospitality. So hotels, being able to have a concierge that when you show up, immediately interacts with you, knows what temperature you like your room at, knows what kind of food you want available to you, is able to guide you to local restaurants or activities, that kind of thing. Financial. So, being able to trade. You’re doing trading or you’re trying to understand a risk insurance, being able to understand what’s happening out in the world in a more real time way. So, having a cognitive system interpret events. So something happens, some accident happens in a potassium salt mine somewhere. Well, there’s not that many potassium salt mines out there, so what’s that gonna do to the price of potassium chloride, and who’s gonna be impacted by that? Or some press release goes out and is that real or is that not? So, being able to understand that.
Retail. If you’re the little guy on the corner with a coffee shop, what if cognitive can understand what events are happening, that reads the local newspaper and is able to recommend that you’re gonna need more people on the drive-thru lane because there’s gonna be something that comes about that, where it predicts that people are gonna want more coffee at a certain time period, so you have the right staffing. Law enforcement and being able to predict crimes. Listening and looking. There’s a lot of unstructured data out there. And then one that’s interesting as well is in sports. We had a customer start asking about, “Well, can we use cognitive to help us understand who we should pick, who we should prioritize for a draft? You always know who the top players are, but what about when you get down the list? Who are diamonds in the rough, and how do we find those more easily? So, it’s almost like every industry has something that cognitive is gonna come and help to change.
Could a point be made? Do you think that artificial intelligence is accelerating too fast?
Well, it’s interesting that you ask that because one thing that’s clear is that we can’t control the rate and pace of that. So, you’ve heard some scientists, prominent scientists and others talk about the dangers of AI, but they also admit the same thing that the genie’s out of the bottle and this is not gonna slow down. And I think the better question is, is how quickly are we as people keeping pace with understanding the art of the possible. And I think that’s where we as people are not looking at this fast enough. So how many people do we know who are willing to still wait for transportation to show up? They call a cab versus summoning an Uber. I know that in some cities, that’s becoming more similar because of the pressures there, but we still have… I know people that still are willing to wait an hour versus willing to wait a few minutes. How many people pay a lot of money for a bad hotel room, instead of paying less for something through something like Airbnb? So at the end of the day I think it’s really on us. It’s really on people to understand what’s happening, what the art of the possible is, and then to join in the narrative around what we need to do to make sure that AI or the cognitive doesn’t go in a way that’s detrimental to what we’re trying to achieve as humanity.
So just like our prior episode on how IOT is disrupting business and industries, this conversation around cognitive computing and changing the way we interact is again both fascinating and semi-terrifying at the same time. Luis, I loved all your insight as it breaks down across different industries where there’s some opportunity in that. Just kind of a quick final question. In a simple fashion, where do you see this having a strong impact on marketers and using cognitive computing to create some stronger consumer relationships?
Okay, so first thing is that it’s really important for people to understand cognitive. It’s a tool, it’s not the end all and be all, but I think it’s important that marketers understand what it is, how to use it, and what its limits are. So you’ve gotta have that kind of at least basic understanding so that you can figure out how it might be applicable. And then just following up on that, I think there’s an imperative for a marketer to use that tool to innovate. We’re at the very entry stage of cognitive. There’s a lot of green field out there; there’s a lot of innovation that’s gonna happen, so anyone who’s marketing in a marketing firm, I think there’s an imperative to be able to innovate with that tool. The other thing is if you feel, and if a marketer feels that this is too technologies focused or, “I don’t know what to do with this,” there are firms out there that are innovating with cognitive.
IBM is certainly, we’re innovating but there are others out there as well that are innovating, so partner up with someone, challenge that firm to help do what you do, or change what you do, or start a firm that innovates around marketing. But to get real concrete, one of the things that I’ve found and that I’ve gotten some interesting feedback on as a way to start looking at cognitive, is start with chatbots. So a chatbot is something that you can type to, you can talk to, and it responds back, and there are courses out there that you can just search on, including ones at IBM, where they’ll teach you how to build simple chatbots. And then you can start bringing in cognitive capabilities into that and start playing with that to really understand what the art of the possible is. And one of the more interesting ones I’ve seen recently around that is a student who created a chatbot that would go and work with local authorities to try to reverse traffic tickets, parking tickets, and it’s all legal because this is the chatbot interfaces with the system that tries to do that. So there’s a lot of innovation that’s gonna happen just even starting with chatbots that are connected to cognitive systems, and I think marketers should really dive in very quickly into that.
Some great advice and insight there to help marketers get their arms around cognitive computing and some of the strategic advantages. I like the thoughts there on chatbots as well. So how can people learn more about you? And how can they learn more about cognitive computing and Watson?
Yeah so I’m on Linkedin, Luis H. Rodriguez is my profile if people wanna connect with me, but the thing about cognitive computing that I think is you can search… So IBM obviously has Watson; there’s others out there as well, but I would say go do a search, the one that I’m most familiar with is the IBM systems. Get a free account and log in and start working with it. The best way to predict what’s gonna happen is to actually do it, and that’s where we are with cognitive computing.
Well, that’s a great way to end this episode, so Luis, thanks so much again for being on the Brand Lab Series. I know you’re busy executive, and I appreciate all your insight and I know our audience will as well; so thanks so much for being on the Brand Lab Series again.
bsolutely, my pleasure.
Tags: Technology, Brand and Marketing, Technology
Enjoyed this episode? View all of our episodes now.
