Remarks by Craig Mundie, Senior Vice President and Chief Technical Officer, Advanced Strategies and Policy, Microsoft Corporation
2005 Pacific Health Summit
Seattle, Washington
June 10, 2005
CRAIG MUNDIE: Thank you. It’s great to be here. I guess I am, as some other people have said, among the minority in that my background is as a scientist and engineer, but not in the field of medicine.
I was really interested in Bill Castell’s “Three B’s” that he introduced in the keynote speech the other night when he talked about biology, bytes and broadband as three things that have to come together or are coming together to fundamentally influence how the world will evolve in medicine.
I wanted to drill a little into that because I think it is an interesting insight, but I would contend that most of you that don’t come from our side of the world, that being the computing and communications environment, really probably don’t yet understand how pervasively information technologies are going to play a role in every aspect of ultimately solving these problems.
When I listen to people in this conference talk about computing, they generally speak broadly about IT; they need a better IT interface, they need more databases, et cetera.
There are really three types of computing that we’re all going to deal with in the future, and mostly what you I think know as computing is what we could call business computing; there we deal with data management. But even here the technologies are fundamentally changing. We have the ability now to not only deal with huge amounts of data, but to begin to automate, process, and workflow at a level that has never been achievable before, and to do that across organizational boundaries in ways that were never achievable before. We’re also at the beginning of an era of collaboration that is much more intrinsic in the way that people will operate than anything that we’ve known, today even with the prevalence of things like e-mail. And to some extent business computing is creating a whole new concept of what markets will be. And all that, I just want to set aside for a moment.
The next thing is technical computing, and in this environment, people think of science and engineering. But here, too, as even Eric said in the last panel, the fundamental technologies are moving beyond the traditional applications of building microprocessors or faster communications capabilities; we now can do modeling and sensing, we have new methods of doing data persistence and dealing with the provenance of data.
When I go and talk to people in academia, they recognize today that one of the fundamental issues in accelerating advances in science and engineering is the fact that in a world where “in silico” is as important perhaps as “in vitro” or “in vivo” in trying to understand what’s going to happen, they no longer have the ability to maintain the provenance of the experiment, and as a result, it becomes difficult to recreate it or determine its correctness.
That leads us to another area I think will fundamentally change, and that is the whole model of publication. Many people talk about the “Journal of Medicine” or the “Journal of Nature”; fundamentally, my prediction is journals as you know them, refereed as you have done them, will basically go away in the not too distant future. The whole notion of reputation and how things will be reviewed will be much more an intrinsic part of the collaborative process. This is accelerating things dramatically in a number of fields. Certain people, particularly in the computer science realm, have already broken with the traditional notion of the refereed paper journal, and in those fields you see a significant advance.
But there too I want to set that aside, and I really want to focus a little more on what I call consumer computing. Why? Because as I listened to the conference, and as colleagues like Eric mentioned, our ability to reach a part of the population through these technologies and to get them engaged in a variety of ways is dramatically better than most people really understand.
I went to Microsoft more than 13 years ago to start our efforts in non-PC computing, and now, more than a decade later, these things are really coming to light. Of course everybody knows our work and Intel’s and others work in the personal computer, but today we have huge investments and worldwide rollouts of intelligent televisions, smartphones will become the standard cell phone, which will basically be the most popularly used computing device on the planet, probably already has achieved that or certainly will in the near future. We have watches that are becoming computers and two-way communications. Many of us are old enough here to remember the Dick Tracy wristwatch; basically, they’re here, and we have to figure out how to use them.
The whole notion that consumer electronics are getting smart and connected, and, as Eric said, there’s a great deal of interest in building personal instrumentation and medical electronics, and we clearly see a huge market, as do many of the consumer electronics companies who work in that area.
We also are at the cusp now of being able to take things that were the province of factories and moving them down to help people with everyday tasks. So the notion of things that can help with monitoring and ultimately even personal robotics will become quite prevalent.
And then finally, there’s a whole new concept of what are the applications. And they aren’t the things that are your IT interfaces to your instrumentation or your healthcare payment system; those will all be important, I’m not going to in this short time focus on that, because I think there’s something more important for us to think about in the IT area.
So now we have new technologies with which to reach the people. You have a whole new model of communications, all of them will be collapsed into one common, new network. How people access information is rapidly moving to be online. You have new items in the lexicon like “Googling” things, right, I mean, basically the ability to search, the ability to find information on a content-addressable basis is actually going to make the world of information technology operate a lot more like your brain does. You don’t have little yellow folders in your head and you had to remember where you put each fact you ever saw, you just think about something related to it and it pops into your mind. Computing will essentially and is beginning to work that way today. That will fundamentally change the facility by which people can discover things they want to know about. And we already see that phenomenon in the Internet itself.
Another thing that’s very important for you to understand is that information technologies, due to the Moore’s Law effect and others that have been commented on, it evolves through fundamental changes in about half of a human generation. There really have never been any technologies that have been this pervasive and this important that have evolved in fundamental ways at that speed, and that has a very profound effect on things that really has to be contemplated as you think about these problems.
So if you think that we change technology fundamentally the world of information technology about every 12 to 15 years — I’ll call that half a generation nominally — then what it really means is that we have a set of pipelined, generational user groups, pipelined in the sense that they’re overlapping in terms of their understanding of the basic technologies. Very few things — automobiles, television, things that are all popular — they didn’t evolve that fast. Televisions have only gone through two generations since the 1940s: black and white, and color. It’s just beginning to get the digital versions of the same old thing. There really haven’t been any other technologies that move at this pace.
So when we think of the marketplace and the customers, we think, oh, we’ve got six different groups of people, ranging from the seniors down through. I’ll call it preretirement, then the middle-aged people, the 20- and 30-somethings, the secondary school and university students, and the kids, ages up to 10 nominally. One of the things we’ve discovered is that kids, in fact, from age pretty much birth to 10, they just absorb this stuff, they use it in an incredibly facile way. At age 11 they become the youngest knowledge workers, and they can use all the tools you have at your desk, in your job, they have no problem; in fact, would arguably be better at it because their brains will have essentially been wired up to do these things, much as you see them play video games, they’re fundamentally wired up to do these things in different ways.
One of the things we’ve already observed is that seniors today, aged 65 and older, they are and will be the fastest-growing group of online users, and their incomes are a great predictor of their ability and interest in participating in these things. And so there’s no reason to believe, even though they were the oldest group, that they will be in any way resistant to this.
I think we also have three economic strata solutions that are required. I’m struck again as I listen to this conference talk how much focus there is again introspectively on the healthcare system itself, but similarly also largely on the use of it in the upper tier of our society. And yet some of the earlier comments in the conference pointed out that our real problem is going to be how do we get this to the 3 billion people who are I’ll say the bottom half of the economic capability on the planet?
So now that we look at bringing these technologies not to the top billion people in the world, the next business for companies like Microsoft and Intel will be the next 2 billion people, who are the emerging middle class of the planet, but we also recognize that we have to go out and figure out how to do something very specific for what I call the “poor” or “welfare class.” And that’s going to require a different notion of both the products, but more importantly the economic and business models. And there, some collaboration between business and government is going to be required, the infamous public-private partnership, and I personally think that the only way we can do anything for those 3 billion people is not to believe that we can ever find the economic base to give them all the things we’ve given to the richer people on the planet, we have to create a self-help environment, and that requires a new way of thinking about enabling these people to help themselves.
I then think that there are three classes of healthcare requirements. Most of the conference has focused on what I’ll call the chronic disease and eldercare part, or the infectious disease and acute care part. But I’m not going to talk about that at all, I’m going to talk about the role of IT in preventive medicine and wellness. Many people have mentioned that. What I don’t think you understand is the degree to which the technologies can be embedded in the life that people already lead in ways that essentially give them assistance in dealing with these problems. Eric mentioned some of the studies Intel has done on the interest people have in using technologies to help them solve their problems, whether it’s eliminating embarrassment because their memory is failing or forgetting to take their medications. We can do a tremendous number of things with very clever devices and software, and we will be infusing these things into just about everything that is electric or electronic. I think even there it gives us a new capability to deal with medical interventions with a lot more basis on predictive information and a great deal more notion of personalization.
I think information technologies create a tremendous opportunity here to focus on consumers and the rural poor. We have new models of interacting with the computers; handwriting recognition, vision and speech will be standard parts of these products so we don’t have to deal with a lack of education. The computer itself will become a vehicle for delivering basic education and fundamentally dealing with the literacy problem, which is a prerequisite to self-help. And the question is, How can we give access to these things to people?
The whole notion of how to monitor diet, exercise, et cetera, it certainly can be done by the use of the computing technology, but people have to know that it’s an issue. And so I think getting the technology in their hands is going to be really critical.
To do that, we have to solve a number of problems, one of them being communications, the other being the low-cost electronics. There’s not really any doubt in my mind we can solve the low-cost electronics question, but the rate at which the world’s cell phones are being adopted in even the, quote, poor countries or emerging economies is pretty staggering.
And so we are all expert at this question of how to use high volume to lower cost and then get the thing diffused into the environment; what we have to solve are some of the infrastructural problems that make it difficult. One of those is communications. For example, in the rural environment, we use some new technologies like ad hoc community broadband networks. These are organic networks that form themselves; nobody has to install them, nobody has to manage them. It allows essentially a community, like a rural village, to have a zero marginal cost broadband network. If you did that and then hooked it up, perhaps through a government connection to a national backbone, then the whole village essentially can participate in many community processes.
One of the things we’ve learned in IT-land is the power of communities. You’ve seen it in everything from how people want to develop software together to how people share information. And the IT capability to build and nurture communities is unlike anything we’ve ever known. It also doesn’t stop at any geographical boundary and increasingly won’t be limited by the knowledge of a particular language.
That then gives us the ability, for example, to have technology and sensing that would allow family members to know whether their elderly parents are actually ambulatory today, whether they met and talked to anybody today, and to do this without being completely invasive in the way you interact with them or to personally have to take all that stuff on for yourself. The ability to have devices in the home to be able to know whether somebody fell over: Why? There’s a camera. It sits there, it just watches you, and if you fall over, it can tell. So even if you don’t know enough, whether it was a heart attack or whatever, it can call 911 for you or call a family member. We are very, very close to being able to do these things at very, very low costs.
I think another thing that this community needs to think about is, in fact, the influence of the community at large. Today the Internet, the Web as we know it, electronic mail, these are all a big factor. The new communications technologies, instant messaging, buddies, et cetera, bring people together in an interesting way. Search allows people to find almost anything they want, any time they want. Even our panelists noted earlier today they went on the Internet to find out what the term was that they were looking for, even in a field of expertise.
I think one of the things that the community better get itself ready for is blogging. Maybe some of you know what it is; basically this is essentially Web activism where the people are coming together to essentially share information on a dynamic basis. When blogs essentially find your industry, you are laid bare for the entire world to watch what happens. And because of all the problems that we can sit here and talk about, we know what they are, well, the rest of the world is increasingly going to know what they are, and they’re going to be able to talk about how well they think their healthcare provider, their individual physician, or, in fact, their government is dealing with these things. And I think that’s going to create a level of pressure that largely has not been known.
Censors and instrumentation will essentially become vanishingly small. You all relish them in the form of these micro and nano electronics capabilities as part of the instrumentation world, but we will similarly be embedding them in all kinds of devices and linking them all together. That gives us the ability to collect staggering amounts of information and to mine things out of that information that people have never understood.
Another thing that I think is going to be a fascinating thing to watch, particularly in the eldercare area, is the field of personal robotics. An interesting statistic: In the United States, roughly we spend $110 billion annually on putting people in nursing homes. And it turns out that if we could find some way, by a combination of communications, monitoring, and perhaps robots, small, simple, even pretty inexpensive robots, where you could improve people’s quality of life and let them stay at home even six months longer, six months in a nursing home is about $35,000; if you spent half that amount of money to stay in your house for six more months and you liked it better, you’d have a net savings in some sense of half of the cost of that nursing home care, and the quality of life issue gets to be significantly better.
If you don’t think this can happen, you should look and realize in the last 24 months 1 million robot vacuums have already been sold in the United States. People are out there buying them; they pay hundreds of dollars just to have this thing run around and clean their floors. So that leads us to believe they might buy them to run around and pick up Grandma if she fell over.
Information technologies I think are at the core of solving the technical, operational and cost problems associated with healthcare. This will require a multidisciplinary approach however, to a degree that is certainly not representative in this audience, and I contend is not representative anywhere in academia, and largely not even in business. We both face similar challenges as industries, if you will, relative to complexity. And frankly, part of the time I spend with Lee Hartwell is he teaches me biology because I think there may be things that the computer people can learn from biology and billions of years of evolution. Similarly, I contend that the IT world is a super, high-speed experiment for many of the things that healthcare is now having to look at.
All the questions about networking; believe me, I mean, we look at networking at an incredible level of detail, and everything happens fast. If you think it’s fast to get a microbe to fly on an airplane across the Pacific Ocean, you should try watching a packet go from to the United States in a few milliseconds.
And so the speed with which the computer industry has to monitor and deal with and react to these things I think can actually be quite instructive to the healthcare industry about how they should think about their future challenges.
I think another area that people mention all the time but personally I think is going to become much easier to deal with is the question of privacy and security. Four years ago, at Microsoft, we embarked on a thing we call the Trustworthy Computing Initiative because we said if we don’t solve the privacy and security problem, people won’t keep using computers. And so we’ve made huge strides and many in the industry are doing it as well. And new mechanisms like role- and policy-based access, automatic audit trails, rigorous identity protection, all these things are going to be standard issue in the business environment, and I think can be brought to bear on these problems.
Another thing, just to give you a few thoughts to go into lunch on, I think one of the things you haven’t focused enough on in this conference is what I call consumer choice and competition at the level of the individual. We need to change the model from what I think of as entitlement to the opportunity to earn, the right to get healthcare. I think we could, for example, think about online auction of services. Today, there are literally tens of millions of people who get up every day, and they use eBay to buy things they want. One of the fascinating things about these online auction systems is that they can deal with locality, they allow you to find things you didn’t know were available, they can be integrated with the search environment, and a fascinating byproduct of their use is that it produces synthetic reputation.
The ability for people to know who is a good auctioneer and who isn’t a good auctioneer is an incredibly powerful thing. And I actually love to think the idea that says what if the healthcare industry basically just said, oh, by the way, I go to eBay, I offer my procedures, my empty beds, et cetera, they’re on eBay. And the individual consumer can say, “Hey, I just go to eBay, and I need a heart transplant.” And actually you’d find out all the people in your area who would be willing to give you a heart transplant tomorrow, and they would bid on the price. And, in fact, right there next to it would be five stars, four stars, two stars for how well the survival rates are for the people who provided those in the past. I contend it would have a dramatic effect on both cost and quality of service.
There are some other opportunities where I think both industries really can help each other as well, and this is more in the area of regulatory harmonization. The last two years I have gone to the APEC Ministerial, and I have basically been part of a cadre of technology people who have been advocating for harmonization among a number of things in the, quote, regulatory, unquote, environment. The one we’ve been focused on is intellectual property. This year at the APEC conference we said there is a process in APEC called the Pathfinder Process, and we’ve chosen to use the Pathfinder Process to start to harmonize even the earliest stages of the process of giving intellectual property patents out. And the thing we’re going to focus on is prior art. Prior art is the common understanding of what things people think they already know. And if that was done once in the world instead of “N” times in the world, we’d have better patents and we’d get them cheaper and quicker.
I think as I listen to this conference that in a whole range of areas we need to have a Pathfinder Process. I don’t know whether it should be in the ministerial environment of APEC, perhaps we should actually inaugurate the Pacific Health Summit Pathfinder Process, where voluntarily people come together and agree that we’re going to do some experiments.
Somebody I think talked about the experiments in Singapore. My notes are on this tablet, people talk about can we record this information, this thing records handwriting, it records pictures, it records anything you want. People just don’t use them within your discipline very much. We could get a huge amount of data from drug studies by just having the people who already have computers just inputting every day anything that they feel about it. We do this today at Microsoft for 650 million computers. Every day, and probably most of you have had this experience, when something funky happens on your machine, it asks, “Can I send this to Microsoft?” You click yes. We get millions and millions of these reports of weirdness happening. And we mine an incredible amount of information out of that. And there’s absolutely no difference in the tools we use to get data out of that and the incredible amount of information you could mine out of essentially the world’s largest continuous cohort study of almost any particular topic you want. But people are already engaged, they already have the computers, they’re already connected, except that this community doesn’t actually have any way to interact with that. So I think we could have a Pathfinder Process where we use things like we did in Singapore.
I mention the tablet because two years ago, I was meeting with the Singapore leaders, and we were talking about education. And I said to them, You know, these things can revolutionize education, except nobody in the education world thinks of it that way, because they, too, just think of their job as teaching reading, writing, and arithmetic, and all this computer stuff is extra work
I would tell you the only two industries in the world who have failed to really utilize information technologies to fundamentally transform their business are healthcare and education. They happen to be the two biggest line-item budgets of every government on the planet, and I think we could make fundamental change in each of these.
In Singapore I said that to them, and they said, “We agree with you. Let’s go and do an experiment together.” And two years ago we went and took these tablets and the Singapore government started to use them and develop curriculum and put them in high schools. And two years later now, I was there in the spring, I sat in a room with some 14-year-old girls who could basically talk to me just as fast as I am talking to you now, they could draw on these tablets, the whole class could do it, they could project their results, and they are now deploying them in about six high schools in Singapore.
The dream is to get the volume up, the cost down, and to fundamentally change the way people are educated. I think we can do all that in villages, I think we can do it in high schools, I think we can do it with people who today can’t read or write, and we really need to come together and do some of these experiments. So I do think places like Singapore have the advantage where small is beautiful, and we should as a community try to figure out how to leverage those places in these accelerated experiments, and then the rest of the world can figure out how to follow.
I’ve talked too long; I appreciate your time and interest. I wanted to give you some things to think about at lunch. Thank you very much.