Saturday, February 13

TED: Future of Mobile With Henry Tirri, Head of Nokia Research [INTERVIEW]

Disclosure: Nokia is a sponsor of Mashable%u2019s TED Channel

We had a chance to sit down at TED with Henry Tirri, Senior Vice President and Head of the Nokia Research Center, to talk about what the mobile landscape of the future holds. Read on to find out what we might expect from mobile technologies within the next five to ten years.

Q: Can you tell us a bit about what you do at Nokia?

A: I%u2019m heading Nokia%u2019s long-term research globally in our labs worldwide, from Santa Monica and Palo Alto to the easternmost lab in Beijing, and everything in between: Cambridge, UK, Los Angeles, Switzerland, and teams in Nairobi and Bangalore and so on.

Q: What emerging technologies do you see playing the biggest role in the next five to ten years: augmented reality, voice recognition, etc.?

A: Those two things are more user experience technologies, but you%u2019re correct. We also talk about %u201Cmixed reality%u201D %u2014 the terminology can be confusing, but there is a distinction between augmented reality, where I%u2019m looking at reality and add information to that from the digital world, and mixed reality which means you can do vice versa also, and put things into the virtual world from the real world. To me it%u2019s obvious that it%u2019s such a natural way of looking at the world and interacting with it.

The key question is how simple and how immersive it becomes. My prediction is it starts with rather isolated services like search and navigation but by the end of the day it becomes part of the interaction. You don%u2019t any more find it extraordinary that you can see the real picture and you get some digital information too or vice versa. And it might be visual digital information, or in audio, or even sometimes in sensing. If you%u2019re talking about a five or ten year spectrum, we%u2019re probably going to have some kind of haptic and sensing way of navigating and getting feedback.

All of this is a very Western view: The high end, cool things for those living in the %u201Cgeek world.%u201D But if you ask me then about growth economies and the emerging markets like Africa, India, greater China, Latin America and some parts of Russia, I would say that the experience and emerging technologies tend to have a different nature because of the constraints you have. You might not have the infrastructure to support data, for example.

So from an interface perspective, speech and gestures are very important there. But emerging technologies are not necessarily always related to the user experience, so things like energy-efficient networking are also a necessity in growth economies. Protocols like SMS are being used in these areas for things we wouldn%u2019t dream of doing with it here because we have access to broadband. There are the %u201Chundreds of millions%u201D who are doing all these very sophisticated and cutting edge things, and at the same time there is emerging technology for the %u201Cbillions%u201D which can take a different track.

Q: Do you think there will be an upcoming involvement with biology? Are we going to bring these devices into our bodies? Will I have a phone in my wrist?

A: Yeah, chip embedding is already an old idea in computer science so we%u2019re ready for that. I think there%u2019s a natural continuum from biosensors %u2014 we already have heartbeat sensors connected to a wireless device and measuring you for sports and wellness purposes. So again, if you talk about the five to ten years era, the questions there are more related to the sensors. In some areas, the sensor development is slower than one would think. Mechanical sensors are faster, but chemical sensors are much slower, so even in the five to ten year domain, certain things are not so easy to do.

When you talk about implantable electronics, you start having %u2026 challenges with your biological rejection mechanisms and other problems for medicine to solve. I would say in five years it doesn%u2019t become big, but in ten years I would be surprised if we%u2019re not seeing a lot more of it. Five years is surprisingly fast, because when you think about large scale deployment of something, there%u2019s a delay factor involved in getting the manufacturing process to be reliable and cheap enough.

I do believe health and wellness-related things will become part of our life, and may probably also merge with augmented reality too. Your body state will be communicated to somewhere, or you can start getting metadata and remote analysis on yourself.

Q: How important do you see cloud computing being for mobile, now that we have an increasing range of devices we cart around with us and are looking for a more seamless experience between them?

A: To me, the cloud has become, and will become, a much broader notion than a server farm sitting somewhere and doing something. So the cloud architecture will expand to more devices and the question is more of the seamlessness in actual usage. You may not even know occasionally what is computed close to you physically and what is computed far away.

There are two issues: One is energy. Sending information bits takes more energy than computing them, which means local computing consumes less energy. This is absolutely so fundamental that it will define the future of how our networks will be built. It implies that the cloud has to have a distributed architecture, because it will be too costly energy-wise for billions of people to be transmitting data. I%u2019m not talking about the bandwidth problem %u2014 this is much more fundamental. Regardless of how much bandwidth you have in the dynamic user spectrum, you will still face this problem.

The second problem is sociological, which is privacy. People are much more positive about something physically close to them and physically in their possession because they feel like they have more control over it. You believe that if your personal metadata sits in the device, it%u2019s better than to let it go away to some nameless server. So there will still be parts of metadata and bits of information sitting close to you for these sociological reasons.

But the cloud itself will expand, and I think the term will eventually disappear. It will just be our default network architecture.

Q: Do you think people%u2019s notions of privacy might change over time too? I%u2019m thinking of Facebook (Facebook

) pushing on people%u2019s privacy, Google (Google

) taking Gmail (Gmail

) more public with Buzz%u2026

A: Yes, and my views on this have evolved a lot over the past 20 years. One dimension is that privacy is culturally dependent, so privacy in growth economies looks a bit different from privacy in the Western world. And even in the Western world, there are different approaches to privacy in Europe and the U.S. In Europe for example it%u2019s very much regulatory %u2014 Germans don%u2019t like Google Street View so they banned it. In the EU there%u2019s a lot of regulatory resistance. In the U.S. it%u2019s more like a community movement, %u201Cwe%u2019re going to make it public that you%u2019re evil.%u201D So it%u2019s a different approach. Asia is somewhere in between.

There are also very contradictory arguments that have been presented to me on whether there%u2019s a generation gap or not. Some say young people put more things up on Facebook or publish things people in my generation would never publish. I%u2019m not totally sure if the generation gap is the right thing to ask. I think it%u2019s more of a question of how much the technology is a part of your life, and it doesn%u2019t as much matter what your age is, although there might be a correlation between the two.

I think it%u2019s complex to predict how people will react, and if there will be negative consequences. Privacy is always considered with respect to the tradeoff you get in terms of utility. If one or two people didn%u2019t get a job or get fired because of something embarrassing they posted on Facebook, but there were 100,000 people that were recruited because of their Facebook presence, how does the judgment come down regarding privacy? Privacy is always relative to the benefits you get, so if people see enough value in sharing and feel safe enough, privacy isn%u2019t the same question anymore. There%u2019s no simple answer %u2014 privacy is an evolving factor.

Q: What do you think of the renaissance of the tablet form factor, and will we see another range of devices occupying this middle ground between smartphone and laptop?

A: I%u2019m a computer scientist and have been hacking with computers for 40 years, so I%u2019ve seen the development from mainframes to mini-computers to PCs to laptops to PDAs. The sarcastic comment is that all of them are %u201Cfads%u201D to some degree, they come and go and the form factor changes. But each can be a decade or two decades or more in popularity. On the other hand, the only thing that has really disappeared is mini-computers. Mainframes still exist, PCs still exist, and so on.

I don%u2019t think the tablet will %u201Ckill%u201D anything %u2014 I don%u2019t think it%u2019s strong enough. I would almost think that tablets and netbooks might see convergence. I don%u2019t think the tablet will become so dominant that you will drop your laptop or netbook and use it as your only device.

Q: How will the advent of 4G change the computing landscape? Will we see new types of applications become possible?

A: This is the capacity question, and right now data-intensive applications cause bandwidth challenges. The interesting thing is we have tolerance thresholds for new features, where we want to keep doing things as long as it%u2019s fast enough, but if the performance is below that threshold, we%u2019ll just tinker with it for a bit and, and I think real-time online media streaming will become more prevalent.

Right now the latency time is not good enough. You can%u2019t have 20 million people streaming their personal video streams around the world in real-time right now %u2014 that is not possible yet, but will become so. There will definitely be new applications emerging %u2014 it won%u2019t just be the old ones getting faster.

Q: In terms of online media streaming, do you think that%u2019s going to change things on the content provider end of things? There%u2019s a user behavior issue to confront too, and I think about how hard things like mobile TV have struggled to take off. How many people really need to watch TV while they%u2019re walking to their car?

A: That%u2019s again extremely culturally-dependent too, looking at places like Korea that have had mobile TV for years. But for me, the real-time media streaming is more about the popularity of sharing your own personal experiences, like your kids playing soccer or when you%u2019re out with your buddies at the bar. That%u2019s a different thing from traditional content; for one thing it%u2019s snippets so it tends to be shorter, but it%u2019s also participatory and it%u2019s human nature to want to exhibit yourself. It becomes a form of expressing yourself, and that will always be popular. And there%u2019s always a long tail of people who are interested in you expressing yourself.

I think the most difficult thing is scale, so something like Twitter (Twitter

) is interesting when you have few followers, and it%u2019s great when you have 2 million followers, but if you have something like 10,000 followers it%u2019s more like, %u201Cwhat do I do?%u201D They are not my buddies anymore %u2014 I don%u2019t know 10,000 people, and on the other hand I%u2019m not famous like someone who has a million followers. I believe in this idea of federated local community: It%u2019s good when you have this small audience, and federated means you have a common platform and you can actually reach things globally. There%u2019s a certain community that is local enough in a network sense %u2014 not necessarily a geographic sense %u2014 to want to follow you.

Q: That makes a lot of sense, especially considering the landscape of user-generated content on the web %u2014 that%u2019s a lot of what people want to share.

A: Yeah, they just want to share and if there%u2019s an easy way of doing it and there%u2019s a general platform, they will do it. Because there%u2019s always some people who want to follow it.

Q: How far along are we in terms of bringing mobile and artificial intelligence together?

A: People talk a lot about intelligent agents, but I think in a computer form factor it doesn%u2019t make that much sense. Think of the annoying Microsoft Office clip guy that no one wanted. The devices we%u2019re talking about are much more personal, so if you can get help when doing real things and interacting in the world, it becomes more persuasive and appealing to have an intelligent agent or avatar type of thing.

The greatest intelligent agent%u2019s behavior can be specified by a good secretary, who can predict a lot of the things I do, can handle a lot of tasks and information flow, and only checks on the things which are important for me. People want to do this and there%u2019s a lot of development around it, but it faces the same problems that any AI activity does: Any time we introduce an automized way of doing something, our own cognition changes to a different abstract level to assume that.

When there%u2019s a more intelligent layer in a device or in software, we start using that in a different way. This is very fundamental and has nothing to do with mobile devices specifically. But I truly believe there%u2019s a good place for AI %u2014 we have elementary things in navigation assistants already that can provide intelligent traffic information. There%u2019s actually a lot of hidden intelligence already and machine learning is already used a lot.

Radio technology will be using AI techniques too, in a deep and unseen way. Dynamical allocation of the spectrum based on availability has deep machine learning components %u2014 it has to learn to predict when certain spectrum is available and so on. So there is a lot going on, but it isn%u2019t necessarily always as sexy as the intelligent assistant everybody is looking for.

Q: As location-based services become more and more popular, do you see any killer apps emerging?

A: The first things that come to mind are local search, really relevant search results based on your positioning. Social search is another no-brainer, because you want to start finding people based on physical proximity because it doesn%u2019t make any sense to go to the bar with someone far away. These are no-brainers and they will be very big.

The things people don%u2019t usually think about with location-based systems are aggregate things like traffic information, and collective information about air pollution and other environmental data. In growth economies there%u2019s a need for health-related and epidemic information collection. Mobile devices are key to monitoring things like this because they are globally prevalent and always where we are. They will enable us to aggregate data and get information that would otherwise be very difficult to get %u2014 I call these aggregate services.

The pollution example is a very good one. You can start to get real-time information about the environment %u2014 your exposure to pollution in LA for example. We did this in traffic already, so think about generalizing it to weather, pollution, and others. The platform allows people%u2019s position combined with something measured, and that gives us a new world.

Posted via web from sophie's posterous

No comments:

Post a Comment