For this week's edition of Fast Forward, I'm talking to Mark Simpson, VP of Offering Management and Strategy for IBM Watson Marketing.
Continue Reading Below
We'll be discussing Watson, of course, but also artificial intelligence, machine learning, and—most importantly—how businesses use these tools to better understand their customers, partners, and how they operate their businesses. Read or watch our full discussion below.
Dan Costa: A lot of people have seen the IBM Watson commercials on TV, and they know that Bob Dylan had something to do with this in some vague way, but how would you define Watson as a product?
Mark Simpson: Watson's a cognitive computing product, which can learn like humans. It learns, it understands, it reasons in the same way that humans do and can be taught over time. So that computing can be applied into many different areas ... Essentially it is a trusted advisor that we can give to humans that can take in masses of data and help them in the decisions that they make, augmenting their intelligence.
I think that's the phrase that has stuck with a variety of people. It's not necessarily artificial intelligence, it's augmented intelligence. It works with humans to really help them do things that they wouldn't be able to do with the brains that God gave us.
Yeah, that's the key differentiator that we use within IBM. A lot of companies have artificial intelligence and what we're not trying to do with augmented intelligence is... just replicate human intelligence or replace human intelligence. We're really trying to give that human a way to make better decisions through having more information. So really having an advisor sitting by their side that can help them in making smarter decisions or faster decisions depending on their situation.
So Watson itself, as you mentioned, it does need to be taught to perform in a variety of these different environments. How does that process work?
Yeah, as you would teach a human, really. You point Watson in the right direction and you give it data and information that it needs to ingest to be able to make, to be able to learn and understand and reason in the ways that it needs to do. It continually learns, so Watson initially will make some wrong decisions as well and it needs guiding in the right areas, but that's all part of the learning process. So really, as a human would learn, you think of it in the same way for Watson.
You specialize in the marketing field. How is Watson applied to the field of marketing?
Watson can be applied or I see Watson applied in three big ways. Firstly, you can think of Watson as a set of computing APIs, which really anybody can have access to. So we see customers that will use those APIs for any number of reasons, whether that's looking at reaching customers in a different way, so if you take the example maybe of 1-800-Flowers, who have now a concierge service on their website called Gwyn who would ...ask you questions about what you're looking for. Say you're looking for a present, a gift for your mother. [Watson tech will ask you] what the occasion is or about what your mother's interested in. It will then start giving suggestions as to the right gifts to give your mother.
You can think of Watson in sort of very broad terms. The second way of thinking about it is what we're trying to do with embedding Watson into our marketing platforms. We think of that really in how can we save marketers time and how can we make marketers make smarter decisions.
Thirdly, you could look at Watson within the marketing platforms as a completely different way to interact with our marketing platform. So Watson, when we are embedding Watson as an assistant to the marketer, it can analyze and monitor campaigns on an ongoing basis. You can ask Watson questions about those campaigns and how those campaigns compare to each other in order that you can make different and the right decisions in a faster way to take action on what you're seeing happen with the campaigns that you're running.
In this world of big data, where we are collecting so much information about our businesses and about our customers, that information overload seems like it's probably the biggest problem that marketers have. You see them reaching for new dashboards all the time to try and make sense of all the information that's coming in, but dashboards aren't enough. Dashboards a lot of times aren't dynamic enough to really make sense of those oceans of information.
You're absolutely right, and I think as well if you layer onto that 80 percent of the world's data about customers and data generally is unstructured and how a human can get its head around unstructured data at that sort of scale and that sort of quantity is very hard to imagine. You can think of it across many industries and marketing what people are saying on social media and in healthcare or doctors' notes that are being written and many other applications in other industries, being able to take in that unstructured data, being able to make sense of that unstructured data just gives individuals, and in this case marketers, the trusted advisor they need to start making much smarter decisions.
Can you give us, you mentioned 1-800-Flowers, can you give us other examples of how Watson's being used to sort of make sense of these datasets?
So we have partnered with Staples with their Easy Buttons. People have seen the Easy Button where you just press the button to order more pencils or whatever you're short of at the time. Watson is partnered with Staples to help enable and help power that, so whether you're short on supplies and you press the Easy Button or whether you go into an app on your phone or whether you use Facebook Messenger or Slackbots or what have you to order new supplies through Staples, Watson will make sense of what you're doing and it will enable that.
It can also start predicting your needs. So there's different applications. We've also worked very closely with North Face who again have a similar assistant to Gwyn with 1-800-Flowers where you can talk to the North Face website through your phone or on your computer. You can also type and interact with it, but you can tell the North Face about your trip. Say you're, I don't know, hiking in the Himalayas in September and the North Face can go off and look up the location, find out the conditions there and it can come back and make recommendations around the gear that you're going to need to have a more pleasant trip.
There's an interesting transition that happens there. The North Face is a great example because in some ways it's a conversational interface that's powered with AI that's very natural where you're just giving it a little bit of information, but then there's this massive data analytics back end where it takes this unstructured data and then turns it around and makes actionable recommendations for you.
Yeah. Look, understanding human language is not an easy thing, which is kind of the first stage when you're interacting with it, even through speech or text or however. Then you have the masses of data behind it looking up location and weather conditions and matching that against product categories and skews and those sorts of things. It's a really exciting way of shopping compared to going onto a website and clicking on category pages and looking through lists and trying to use filters. It's a much more enjoyable way of shopping.
It seems to me that brings us to that concept of cognitive commerce where you've got a cognitive computer that's enabling commerce transactions up and down the front.
Yeah, absolutely. Absolutely, and as I say, Staples is a great example of that, trying to predict when somebody is getting short of supplies based on the history of consumption and enabling that ordering process to happen however someone needs and wherever they are and have that sort of more immediate and just-in-time sort of fulfillment is, it is changing commerce significantly, yes.
How hard is it for a company to get started with Watson? These are pretty sophisticated concepts. You can imagine Staples has got a team of people working on this. IBM is, obviously, a very large company. How hard is it for a retailer to spin this up?
If you look across not just retail as well, it's accessible to anyone, but there are obviously certain skills that you need to be able to develop products around. The Watson APIs are open to everyone. We've seen teenagers using it to build some really cool apps like fighting parking tickets [up] to very large enterprises using it to augment and sometimes replace their own AI. Really there's a huge scale of use cases there.
How can Watson help you fight a parking ticket?
It's a fantastic way of looking at the parking ticket and looking at all the parameters around it and how to maybe fight that parking ticket and how you're interacting back to that particular authority. As I say, developed by teenagers, absolutely phenomenal. The success rates as well are much higher than they are if you use humans to do it.
Stand in line and go to the courthouse.
Very good. I think another key thing is the ability to make sense of unstructured data, like once you've got a structured dataset it's relatively easy to do some analysis and be able to do some visualizations and so forth, but Watson's ability to make sense of unstructured data is one of the things I think that sets it apart.
Yes, absolutely. In marketing, if you look at a couple of the key trends that are happening in marketing at the moment, one is there's an explosion of data. I think we're projected to have about 20 times the amount of data by 2020, which is only two and a half years away. That's a huge amount more data, but as you say, the shape of that data is changing as customers start revealing themselves less, they're actively going out a way to give brands less of their information. Technology has sort of advanced to enable that as well.
A lot more of that data is really about behavior and trying to extract intent from individuals' behavior. A lot of that data is completely unstructured so I think the older sort of CRM style of marketing, while it has its uses is going to become less and less effective and we need to find new ways to pull out that intent and use that to market to individuals better.
How long will it be before this becomes a proactive process where the tools actually know more about your future needs and your future wants than you may even know about yourself?
Yeah, I think in simple cases it has already. There are many cases where you can see that from my wants as a marketer, when I go into our analytics platform I want to look at a campaign, I want to see if there are any problems with it or anything like that. Watson's actively pushing that to me as an individual now. I think there are varying degrees of that as you go through different industries and down to a personal level as well, so it's a hard question to answer, but I think that that is already happening at the moment. Really I think look, we're at the very early stages of cognitive computing. I think the exciting thing is where it can take us when you look forward because I think the future is very hard to sort of fathom all the effects that cognitive computing can have.
Let ruminate a little bit on that, that future in just a second. I think we've got a question from the audience.
We have a viewer who wants to know what's the latest innovation in the field? What are you most excited about?
Yeah, that's a good question because the field is moving so fast all the time and I think you're seeing new innovations pop up all over the place. I think that for me some of the work we're doing in cancer and cancer care is really phenomenal. You look at those applications of curing disease and helping people who are suffering, and I think that those are the elements which I think you really sort of stand by this and think this is just a phenomenal piece of technology and computing that we really need to grasp with both hands. So I think there are many different uses across many different industries.
Can you think of an industry that's not going to be transformed by either cognitive computing or artificial intelligence? Because we're seeing it being used in marketing applications, we're seeing it being used in transportation applications, commerce applications. It seems like it's one of those fundamental technologies that disrupts everything.
Yeah. I think the exciting thing is it's hard to imagine an industry that won't be affected. I think if you see the examples that are in the industries at the moment, we've partnered with H&R Block for tax advice and to file your tax returns, taking into account all the changes in tax laws over the last 12 months and being able to help H&R Block's tax advisors make better decisions and save their clients more money. You move that to some of the work we're doing maybe with Kone, who produces elevators and escalators and partnering with our internet of things business to enable elevators to be fixed really in real time and before they go wrong. No one likes getting stuck in an elevator, right?
It's really exciting the different areas in which it can be applied. One of the applications I saw the other day was with Whirlpool where the washing machine with Watson embedded in it is able to speak to the tumble dryer to advise the dryer of how long the drying cycle should be, which sounds really small and obvious, but if you think of it, if you get the drying time right how much energy that can save across the globe, it gets really exciting. The scale of some of the problems.
Small efficiencies at scale can make huge differences.
That's pretty fascinating and it's always interesting to see these super sophisticated high-end technologies come down to that most prosaic level of you know what? You're going to spend less money on your electric bill because your washing machine is going to be more efficient.
Absolutely, yeah. Your washing machine's going to be more efficient and hopefully you'll get better customer service because your washing machine can speak to the customer service operative and they can know what's going wrong and send the right parts with the engineer to fix it.
Great call. I want to move to the questions I ask all my guests, a little bit about the future. What technological trend are you most concerned about? What keeps you up at night?
Yeah, it's hard to keep me up at night. I sleep quite well. I'm a fairly sort of positive person, but I think when you look at all sort of technology evolution, I think the thing that most concerns me is that the way in which it's being used. The concerns are in when it's being used in an unethical way, particularly with the pace of technology advancing. I think it's very important that we do as much as we can to use it ethically and I think that we do as much as we can also to nullify the unethical uses of these technologies too. I think that's probably if I was looking at one or the biggest overarching area I would pick that area.
Do you think that individual companies should take that on themselves as they're developing technologies and products to think about the consequences of these technologies before they bring them to market?
I think it's the responsibility of everyone.
Yeah. We've got another question from the audience.
Is Watson going to be used for medicine or law?
There are already applications of Watson true to medicine. I mentioned cancer, the work we're doing with cancer care and that will go broader as well, so Watson can ingest doctors' notes. It can analyze the thousands of medical journals that are produced every single day to sit there as an advisor. Remember, we are advising the doctor on decisions that they need to make. It's not that Watson is going to take over. It's really Watson is going to work alongside in medicine. Equally, in law I absolutely can see that Watson will sit alongside lawyers and those in the legal profession to enable them to make better decisions, yes.
We've seen that issue of augmentation play out specifically with Watson where Watson can beat a grand chess master.
Yeah. Look, that's how we foster everything we do. As we say, augmented intelligence rather than artificial.
Since you're a positive person, on the upside, what are you most optimistic about? What are you most hopeful about?
Look, as you can see I get very enthused by AI and it's hard not to be really enthused by that. I think there are a few sort of big events that happen through people's lives, and I think that AI is one of those big events. When you look at the application of AI and when you look at the application of cognitive technologies, the [possibilities are limitless]. It's hard to fathom where it's going to reach to. I think that thinking about how we can apply that in a better way, how we can help people more by using cognitive computing is a real exciting area. If I look at sort of, I know you always ask about sort of people's favorite apps.
If I look at sort of the, my life over the last 40 or so years, there were some big major events that happened through it, which kind of excite you and you use devices or whatever it is. You think when the PC really sort of becoming mainstream is one. When the smartphone launched 10 years ago, only 10 years ago is another and I think AI is another one of those. I look at the work we're doing, I now have a couple of small kids, the work we're doing with companies like Sesame Street who have had 45 years working with kids and we're building apps to help children's learning. I know it's coming up to school holidays, I know my wife will be keen on finding new ways to entertain and educate the kids through school holidays, but the work we're doing there I think is really exciting.
Yeah. There is a generation of kids, and I think it's because our generation of the proper age where we're invested in these technologies in a new way and the same way I was left at home in front of the TV, that's not going to happen with this new generation. They're going to have software tools that help them learn on every screen that they carry.
Absolutely. Absolutely. I would love to get my eldest away from the TV more and you can already see that shift. You can already see the shift to the iPad and the computer to investigate and to find out more. I think it's an exciting future.
If you had one app or product or service that you would have to point to and say, "This thing changed my life," what would it be?
Yeah. I look at, as I say, there are some major events that have happened I think. You look when I was young, getting our first computer in the house. You look at the smartphone, having everything at your fingertips wherever you are is another one of those. As I say, AI being one where I think it can really sort of take that to another level as well. I think through people's lives there are some really major ones where you will continually find that you get new, sort of new opportunities to rely on something that is really big. I think AI is probably the next revolution.
All right. If people want to follow what you're doing, what IBM is doing, what Watson is doing, is up to these days, how can they find you online?
Yeah, so I'm on LinkedIn. I'm on Twitter. I have a new Twitter account actually. My last one I can't get ahold of.
You lost your Twitter account?
I did lose my Twitter account.
For was it bad behavior?
It wasn't bad behavior, no. Nothing to do with that. I lost the passwords and the email address it with, so my new account @SIMMOMJ.
All right, we'll try to get you some followers to sort of ramp it up quickly.
Exactly. It would be a help.
For more Fast Forward with Dan Costa, subscribe to the podcast. On iOS, download Apple's Podcasts app, search for "Fast Forward" and subscribe. On Android, download the Stitcher Radio for Podcasts app via Google Play.