0:02 CTO, IBM Watson
We're very glad to be here in Africa. As you know, late last year, we opened our 4th research lab in IBM, here in Africa. And Africa represents to us an incredible, very exciting set of opportunities. And that's for many reasons, okay, not the least of which is the African economy is expected to be about two and a half trillion dollars by next year, by the end of next year, which is bigger than countries like Brazil, Russia, India, Australia. So we have a remarkable economic base here. But another statistic, which I find very exciting...I worry about demographics around the world. And one of the great things about Africa is, you're still having babies, you're going to have 1.1 billion young people in Africa, more than India and China combined by 2040. These are remarkable numbers and this represents hope, and future for our planet, and huge set of opportunities. And one of the things we'd like to talk about today is the changes in computing, why those changes in computing are going to be so exciting, effective, and able to help us in solving the most outstanding and interesting issues for Africa. So what is this cognitive computing that we're talking about? So, in IBM, we actually started off as a company in tabulating systems and what are tabulating systems? These were these old systems, working with punch cards, etc. And you could program a little bit, you could have a patch panel, or maybe you could set a few switches, but you really couldn't program them. They were fixed buttons. They were built for things like doing a census. And almost 100 years ago, it was possible to do a census using these kinds of tabulating machines. And then a breakthrough occurred about 50 to 60 years ago, the period of program, where we built computers that had stored programs. And those stored programs were able to transform the world of information. They were able to process in a very refined, in a very flexible way, information. And yet, at the same time, those systems were limited, they had all the problems that we've become aware of aging and legacy software, they had a brittleness, and they didn't have an adaptability to deal with the changes that are going on in the world, and problems as they emerged. So very interestingly, in the early papers on programmable systems, and stored program control, the only papers by Don Norman and others, they spoke about the ability of computers to adapt. They even spoke of the ability of computers to modify the programs while they were online. But of course, when we were in school, this was very much frowned upon, you should never write programs--I had points deducted on my programs--Because I wrote assembly code that modified itself while it was running. And that was considered demonic to do things like that. But meanwhile, things were happening, which led us [inaudible] that new area, we at IBM refer to it as a period of cognitive systems. It's about systems that can in fact, have perception, memory, judgment, learning and reasoning. And this, this was a change. This is about systems that can adapt. Systems that can respond to data in real time, and probably a novel of systems that can learn. So why did this happen? Why is this the age of computing when there are so many computers around us that are still in the programmable systems era? What change? What happened? And why do we say now is the period of cognitive systems? Well, one thing happened. What was that? Data. Thanks to the Internet and connectivity, and breakthroughs in communications, which we all know and live with, with our mobile phones, etc, massive amounts of data became available. And also at the same time storage technology became available to in fact store almost unlimited amounts of data, doubling every two years. And we're now at the point in the world where we have about...this year, we will have about 40 zettabytes of data created. And by the way, it doesn't matter whether it's created now, or in the last couple of years, or in all of history because the data is increasing at such a rapid rate...remember, students if you remember the geometric progression, if something is doubling every two years, all of the data's history of the world will be created...All the data in the world will be matched by what was created in the next couple of years. Okay? 40 zettabytes of data. What's a zettabyte? That's one with 21 zeros after it. So write that down, one with 21 zeros put up your hand when you finish writing. It'll be a while before you have 21 zeros but that is a lot of data. And that large amount of data has allowed us to do certain things, some people associate cognitive computing with some of the techniques which 20 or 30 years ago, [inaudible] limited applicability, techniques of artificial intelligence, they were good techniques. And they worked okay but they didn't work in a transformative way. And a transformative way that we need today. The massive amounts of data allowed us to use techniques that were not known before, techniques of machine learning, statistical pattern recognition, various techniques, and we finally have enough data to make these things work. And the instance of that, of course, was the IBM experiment, Watson, which we did a couple of years ago, we were able to outperform the two top champions in the world in answering arbitrary questions, arbitrary questions about history, geography, science, medicine, baseball, trivia, film trivia, we were able to outperform them. Why? Because we had access to massive amounts of data, and we had created new techniques, which allowed us to work with managing and learning from that data. So data is what's made the difference. Here's a little picture of the growth of that data. This is where the 10 to the 23 numbers come from [inaudible] 40 times 10, to 21...and you'll see that there's many interesting aspects to this data, it's coming from many sources, Fred Matiang'i before mentioned connected cars, that would be a huge source of data. It's illustrated here, the whole [inaudible] of Internet of Things, of course, mobile computing, we're all generating data, not just as we type. But as we move around. Okay, cloud, of course, can hold that data. And we're all social beings that as social beings we are all generating masses of data, that data has all kinds of fascinating, and new capabilities associated with that. Data is not all the same, there are different kinds of data. The data has huge volume, we've already spoken about that. But the data is coming at us like crazy, too fast to even process all of it. And we have to come up with new techniques. And we are coming up with new techniques. The data has huge varieties. It's not just textual data that people are typing, it's coming from sensors, it's coming from, it's coming from, from many connected devices, it's coming from radiological images, massive amounts of data being generated. And interesting, the data has varying degrees of veracity. It's not always accurate, it's not always true, we have to find ways of working with different kinds of data, different mediums of data, to in fact, extract valuable information. But one of the things that we'll see as we address the problems of Africa is the ability to relate that data together and to find patterns, and in fact, to find surprises. And as we begin to extract features, find patterns and find connections, we're going to find amazing things. And they will fall right into the bullseye of the kinds of things we're going to need to do in Africa. So before we get to that... Cognitive computing - a couple of new capabilities, obviously, we can do things like assistance, sub-assistance, that could be very powerful in a call center, when you call a call center, for example, there could be a few thousand pages of information that the call center operator may not be able to navigate through effectively. But with the help of cognitive computing, we can easily answer a question like that. That's a relatively simple example. What about in a doctor's office? Do you know how many medical papers are in the literature? The number of papers in PubMed that are either in PubMed or cited by PubMed is now over 21 million articles. And more interestingly, we're producing in our medical research about a million...over, now over a million papers a year, how can a doctor keep up with a million papers a year? How can he even read 100 papers a year, given how busy he is? We have to help him with a system to understand that [inaudible] that information to understand what's important for that doctor's practice or that doctor's particular patient so they can make decisions. And finally, what about discovery? So I'd like to take you back to your days, we've all had courses in statistics. And I, I don't know if you had the same experience, but I had pretty cranky instructors in statistics. And they told me that we should always form a hypothesis and then test the hypothesis. Don't jump the gun, don't start looking at the data before you've formed a hypothesis. That was nice. But then [inaudible name] came along. Professor [inaudible name] came along and said, well wait a minute. You know, if you torture the data, for long enough, you can get it to confess to almost anything. What about the story in that data that's trying to get out so what about exploring the data and looking at it before you have any preconceptions or prejudices. That's discovery. And in fact, cognitive computing is helping us to find out stories that are trying to escape from that data are harmonious. And in fact, we're finding in our research in cognitive computing, that amazing things are beginning to reveal themself to us by taking a very new and cognitive approach to information. Now we're in Africa. Now we come to Africa, much has been said about all the grand challenges in Africa. The problems of cities, food, water and energy, and government and healthcare and [inaudible]. The challenges have been well documented, that if you haven't seen IBM's film about the challenges of Africa, in fact, many of the people here in the audience have been interviewed for that film, I urge you to look on YouTube and go and see that film. These problems are well stated, well known. Many of them are almost cliches, the problems are hard problems, the difficult to solve problems in Africa. So how are we going to address them? Well, project was...now why have we called this Lucy? Not far from here, about 3 million years ago, a woman walked upright. The name would be... Lucy, she's been called by the anthropological community, the archaeological community, Lucy. We don't know a lot about her but we strongly suspect and believe that we are all, we are all every one of us is related to her. And she lived in an environment...in a beautiful environment, which we all enjoy here in eastern Africa, in Rift Valley. And she lived in an environment and she managed, and she managed to be part of something that would become today's humans. So we chose the name Lucy, because Lucy reminds us that we are all connected. And we're all connected to our environment. And we're all connected to those hard problems, those problems of energy, and food and health. All the problems that Lucy had to deal with in her environment. That's why we call [inaudible] and what's Lucy going to do? Lucy is going to help us to marry together cognitive computing and problems of Africa. So what is it and how's it going to work? Well, let me get a little technical here because I want you to understand how basic concepts in the field that we're trying to do. About I think it was less than a year ago, in this auditorium, IBM announced that we were going to bring the biggest most comprehensive computer system and Watson, the Watson technology, which you'll hear about some more in a minute [inaudible]. We were going to bring that to Africa. Why would we bring that to Africa? We had a strong belief that many of the hardest problems in our world today, particularly in Africa, are problems of information, for example, problems of health care, we know that worldwide, about half of spending alone in health care, 40 or 50% is wasted. Why is it wasted? It's wasted on incorrect treatments, or treatments that don't work. Ok? In Africa, and the problem with healthcare, we have a problem of far insufficient number of trained physicians but we have many other people that can deliver care, ranging from a trained trained medical personnel and midwives, even family members, even the individuals themselves can understand about what they have what they should do, if they're feeling ill. So this is all about information. It's not entirely about information, of course, molecules are important, drugs, pharmaceuticals, understanding the mechanisms of diseases are very important. And information has a huge role to play. So what we'll be doing with Lucy. We will build and our goal is to build here in Kenya to serve all of Africa, a cognitive hub. That cognitive hub will be based on Watson [inaudible] the most powerful computer system on the African continent, and it will carry out several functions. It will not...First of all, it will not do everything for Africa, it will be part of a three tier model, meaning that there will still be server computers all across Africa, some created by IBM, some created by our competitors, but there will be many computers which will, for example, host educational material, or healthcare material, or material related to energy. But those computers will occasionally connect to a cloud in most countries. And that cloud will provide a repository and a sharing of educational information, whether it's the uploading of students scores, or its the downloading of the curriculum material. And what we are offering in the cognitive hub concept of Lucy is a central point where in Africa they will carry out several functions. For example, as students perform, it will allow that data and that performance and how they respond to interventions to be uploaded to Lucy. Lucy will combine that data and do response curves, we'll finally learn what programs work and what programs don't work, particularly for challenges in Africa, like challenges of large class sizes. And when those response curves are aggregated, they'll be made available into the clouds, and they'll be downloaded. That's the two directional flow that you see here. Similarly, for financial inclusion, we'll find out what is the role of credit, how much does credit, really hold in the formation of small business. We'll upload those response curves and download and share information. But most importantly, we will be [inaudible] combine information between domains. So we will understand, we'll begin to understand and discover the relationship between water in the cities or the relationship between water and agriculture. Perhaps we'll better understand the relationships between energy and traffic and congestion and loss of efficiency. Because of the congestion that we have on our roads. Or we'll understand the relationship between financial inclusion and does it actually create jobs? So we'll begin to understand these things. So we'll create and discover cross domain knowledge which will then be available to these clouds, and these server systems and remember, it [inaudible] full connectivity. So this [inaudible] the ability of a standalone server in a school that may suffer energy or power outages or connectivity outages, the school will still run, the educational material will still be there. But we will be functioning, we will be creating interventions and curriculum for the students that benefits from our experience and best practices across the entire continent. So that in a nutshell, is our thinking about project Lucy, and why it's a sharing and cognitive capability that the entire continent will be able to benefit from. Just give you a few examples. Education, we all know the challenges of education, there are many children that never get a chance to go to school. But if we can disseminate the curriculum, using information technology in the periphery of life and information that works, and interventions that work in large class sizes, we can make a huge difference. In the field of health care. Africa is well, well is often quoted 25% of the global burden of disease, and yet just a few percent of the world's health workers, how do we enable those health workers? Let me just take you back to the Lucy picture for a minute. Do you notice that we're going to provide the service but just via SMS from a very basic phone, you will be able to ask questions and receive answers. Whether it's a healthcare worker or an individual, we will provide that as an SMS service across Africa, where you can benefit from some of the services in Watson directly without going through this cloud service. So in healthcare, there's much that we can do. Working with infectious diseases, diseases like cervical cancer, for which inflammation is a big part of the field and chronic diseases across the continent. Water. We mentioned the importance of water in fact, there are those who argue that water might be the single most...clean water might be the single most important technical innovation in engineering or technology. Water as you know about 50% of the people worldwide who are hospitalized are hospitalized because of waterborne infections. Water also has an incredibly important relationship with agriculture. And there are many threats to food security throughout the continent. So the use of water boreholes, and whether they are providing clean and safe water and whether they're providing enough water. This again is another cross domain linkage. And finally, human ability. We've been discussing here in the lab the last the last day or so, the impacts of congestion and poor transportation. Poor transportation is a source of problems for national productivity, productivity is lost. People are unable to get medical care or unable to get to places of education because of poor transportation. Another linkage another cross-domain linkage which we can address through Lucy. Genomics and agribusiness we tend to hear in the West about genomics and personalized medicine. But genomics is turning out to have an incredibly important role in effective storing food security and and optimizing both agricultural crops and also ... pasture pasture. pasture science, finally, economic inclusion. Africa has been a leader in economic inclusion where all the breakthroughs are right here in Kenya. Mpesa for example. And those systems can continue to be rolled out in fact as part of our project Lucy work we're working on new low cost methods of obtaining a very low cost, but reliable enough authentication. But at the same time and of course, the main thing is, you'll be able to understand the relationship between financial inclusion and other aspects of economic development. So to wrap up the story, I think you've you've understood that these these techniques of cognitive computing and effort, they're really made for each other, but we need to solve them, we need to address them through an adaptive and a learning manner. That takes into account indigenous knowledge, tribal knowledge, and knowledge of Africa. What will work, what interdomain effects are important in Africa, and which ones are dominant and which ones are not. So we must integrate this cognition across these many domains, and we will discover relationships. And I think that we will have the best shot anywhere, of really making a huge change to the outcomes in the planet by using cognitive computer. For everyone that's here, we are seeking world collaboration, we will have collaboration around the world, with the IBM labs, with many other institutions. Several major educational institutions are represented here today and we'll hear them participating later in panels, the government organizations across the continent, around the world, NGOs, etc. They'll all be participants with us. This will truly be a collaborative, and a cognitive engagement that I think we'll all be very very proud of in a number of years. And with that I want to thank you all for attending today. Thank you.
So we've heard a little bit about Project Lucy. Our next speaker is going to go just one level lower. So Project Lucy is a 10 year initiative, but underneath it is a technology. And you've heard the name Watson. So I'd like to introduce the next speaker. Let me just read a little of his bio.
AO: This is the transcript from an opening keynote presentation given by the Chief Technical Officer of IBM Watson at the public launch of the IBM Research Lab at their campus in Nairobi, Kenya in 2014. Find the official press release from IBM related to this visit here.