Techonomy16 conference in Half Moon Bay, California, Friday, November 11, 2016. (Photo by Paul Sakuma Photography) www.paulsakuma.com
Techonomy16 conference in Half Moon Bay, California, Friday, November 11, 2016. (Photo by Paul Sakuma Photography) www.paulsakuma.com
President, Client and Internet of Things Businesses and Systems Architecture Group, Intel Corporation
Founder and CEO, Techonomy
The economic potential of IoT is unquestionable. But what priorities, policies, and implementation structures are required? Why does the US only have “recommendations” while the UK has a national IoT strategy and other countries are making it a priority? What are the implications for national competitiveness and business?
Kirkpatrick: Murthy Renduchintala is recently arrived at Intel, where he is President of the Client and Internet of Things Businesses and Systems Architecture Group, which is a big part of the company. He came from Qualcomm where he was, for 11 years.
What we’re really here to talk about, though, is very much at the heart of the Techonomy theme this year of “Man, Machines, and the Network” and that is, “What’s happening with policy and the Internet of Things.” With all things policy, this conversation is quite a bit different than it would have been before Tuesday. But I think before we get into some of that ‘edgy stuff,’ it would be good if you talked from your perspective about what’s possible with the internet of things and why it’s so important to you and to Intel.
Renduchintala: That’s a question with infinite proportions I think. As you heard from THE panel that just left the stage, you’re really talking about an environment now where you’re fusing together great computing capability, pervasive connectivity that makes everything both ambient and pervasive, and you’re also talking about the advent of really advanced thinking in machine learning and deep learning. You fuse all of that together and I think the possibilities are enormous.
Some typical examples that are getting a lot of visibility right now is the whole concept of self-driving cars, and robots, and drones. These are entities which not only will be great statements of how far compute capability has come, but they’re also going to be activities which, really more than ever before, tangibly integrate compute with computing with artificial intelligence. Therefore that fusion of technology and the requirements to actually fuel the imagination are just really compelling areas to work on for many of us that are in high technology.
Kirkpatrick: The way we sometimes talk about it at Techonomy is pushing intelligence to the edge and obviously, given that you’re Intel and a primarily processing company, not any other kind, you’re getting into a lot of other stuff, you would presumably want as much intelligence in all those devices as far out toward the edge as possible. Is that a given?
Renduchintala: Yes, but let’s be a bit more specific. Intelligence at the edge comes from the ability to learn and aggregate a base of knowledge in some centralized manner, so I think it’s a very symbiotic process. Intelligence isn’t a static concept. It’s growing, it’s evolving, and it’s being enhanced. Intelligence at the edge basically provides a degree of feedback towards an ability to concentrate that data for many endpoints. You generate a new level of intelligence which then gets passed back to the edge. It’s a symbiotic process where each edge device adds an increment of knowledge to a central repository. That central repository aggregates everything, we call it a data center. Then that aggregated intelligence, which basically is elevated by everybody’s contributions, is then pushed back out to the edge so you have that symbiotic, virtuous cycle of information flow.
Kirkpatrick: But there is another way of thinking too which came up very heavily in our dinner discussion about AI on Wednesday: as you have more and more intelligent devices dispersed in various kinds of digital ecosystems, a new form of AI can be effectively emergent from the almost synaptic dispersion of very large numbers of devices over time, and that could be something very promising or very scary—probably more promising than scary, under most circumstances.
Given all these things that we see as possible that you’re excited about, how is the US doing, as a country, in positioning itself for that world? And then we’ll go from there.
Renduchintala: From a thought leadership point of view and technology development point of view, we’re doing extremely well. There’s a lot of great research going on that’s basically being viewed and studied by many around the world. But in terms of taking that technology and actually driving forward programs that really are partnering with the legislative processes in various countries, and driving toward being able to create industry policy making frameworks that really takes this into the fore, there are many other countries that have been able to get a degree of a head start on the US.
For example, if you look at many countries in Europe and Asia, whether it be with respect to self-driving cars or the concept of smart cities, there’s been a tremendous amount of progress made because people have really understood the economic value that lies within being able to really unleash the amount of technology we’re developing. From a technology point of view, the US is doing a phenomenal job. In terms of actually harnessing that technology into activities that are really driving some of these ideas into the here and now, I think we’ve got a bit of catching up to do.
Kirkpatrick: In terms of what a country needs to do collectively to make those strategic moves, is it entirely about government getting smart? Is the government the real problem in this country vis-à-vis those other countries, or is it industry and government together? Where do you see the real need for shifting?
Renduchintala: First of all, I wouldn’t describe it as a problem; I think it’s seizing an opportunity. Like many things, it comes down to an appropriately structured partnership. Clearly, you can’t talk about concepts such as autonomous driving or drones or even to some degree robots without having a degree of legislative involvement. It’s really how you structure that engagement, and how you make sure that legislature is really trying to talk about policy frameworks and performance standards, as opposed to dictating a technical implementation and allowing the industry an innovative environment to really figure out what great ideas fulfill that policy framework and those performance requirements. So, how do you structure that in a really strong and partnership-oriented way?
Kirkpatrick: So what would you like to see happen in the United States, and particularly, what could the government do?
Renduchintala: What would be great is really to try and set an aspiration in time as to when we want to be able to really see these services be mainstream, and what would be the types of policy framework and the engagement of discussion in developing those policies to allow that technology to develop along a harmonized way. One of the things we’ve really got to really make sure we establish is all of this innovation moving along vectors where we can interoperate with a lot of that technology. We need to make sure that we have means and ways of how, for example, a vehicle developed by one manufacturer behaves consistently with a vehicle developed by another manufacturer. Therefore regimes of interoperability and consistency and normative behavior need to be established. That’s an area where a good discussion can be promoted.
Kirkpatrick: You’ve said three things that you’d like to see happen. Legislation was one. What do you think should happen legislatively?
Renduchintala: Again, it’s a case of the legislative bodies understanding a bit more deeply about what’s happening in the technology environment and figuring out how they could be capitalizing that technology towards a mainstream burgeoning. It’s really all about things like understanding what kind of security regimes would be required, what kinds of interoperability regimes would be required, the way data can be shared and aggregated to generate higher orders of value. All of those are areas which I think some degree of involvement from legislature is required in order for us to make sure that we’re moving along vectors which, ultimately, when we’ve concluded will have the ability to be deployed.
Kirkpatrick : Another big piece that concerns you is a national R&D strategy, driven, in large part, by Washington, but not only. Talk about that.
Renduchintala: It goes back to the fact that there are some things where you need to have a clear understanding set amongst everybody playing on a specific field so that their efforts are coordinated toward being able to provide added value to each other. Again, we talked about some examples such as autonomous driving and the regulation, for example, of things like drones. The role in which the country can set a national agenda in terms of being able to foster the growth of those areas is pretty important.
Kirkpatrick: Security, you mentioned it, that’s a key part of what you think the government needs to very methodically tackle in this regard, right?
Renduchintala: Again, I go back to, more in the roles of defining policy and defining performance expectations. Clearly security is a very topical subject in the area of IoT today. It’s not just about security of the end device. What we’re going to really see is data being transferred across very wide networks, and that data is probably going to be aggregated with other sources of data in order to create higher orders of information. Therefore, the processes by which secure data is exposed so that it can be aggregated with other types of data to create higher order information needs to be governed as well.
Kirkpatrick: Interesting, in terms of the aggregation of data, yesterday on this stage we had Marco Annunziata who is the chief economist at GE, and he’s Italian. One of the things that really concerns him is the attitude that’s setting in in many countries about data sovereignty, particularly in Europe. They really want to put borders around the data on national bases, and he said that could be disastrous for the ability to achieve the advances that we need in many of these instances, where free flow of data is going to allow us to optimize and to get knowledge across barriers that really are only artificial ones. Does that enter into your thinking? How big of an issue is it?
Renduchintala: Very much so. I go back to this example of self-driving cars, because it amplifies the concern so conspicuously to my reasoning. Today in most of Europe, you have open borders and you don’t even realize when you go from one country to another, and it amplifies the example. Let’s say for example, traffic management data or data related to the efficient trafficking of vehicles. If you were in a position where that could not be seamlessly transferred across boundaries, you can clearly see the kind of difficulties you’d get into in a place like Switzerland or in a place like Italy where they have so many borders. The ability to be able to share data in the right context is fundamental for some of these opportunities to really come to life.
Kirkpatrick: You’ve said things that you would like to see happen legislatively. We do have a new President. One of the things that unfortunately is something I’ve been repeating throughout this conference, and people who know me say I say it a lot. I say it about business but in this case I’m going to say it about government. We have a disturbing dearth of leaders who understand these issues, particularly in Congress. We have had, in Obama, somebody who gets it, but he’s going away. What’s the strategy to move forward given that we have, essentially, a clueless Congress? We have a new President coming in who doesn’t even use a PC, although just today it was announced that Peter Thiel is going to be on his transition team. I think we would admit he understands these issues. Help me understand what we can and should do, even politically, about all this.
Renduchintala: Well I think the burden continues to be very much on industry to chart a course for greater education. The economic opportunity here is profound. The more we can education everybody concerned, including Congress, about what can be created with the technology that we’re investing large amounts of R&D on, and being able to engage them in a discussion of how they can help, and being able to define the value proposition as to what would come as a result of their help, I think is our burden and we need to continue to drive that forward. That is something that very much needs to be taken as a proactive stance from the industry community and continue to pound that message of positive gain that can come from legislature and industry working harmoniously together.
Kirkpatrick: Who has a comment or question? Please identify yourself.
Lemmey: I’m Tara Lemmey from LENS. David, you and I have certainly talked a lot about this. One of the questions that comes up on the policy front, having spent a lot of time in the last year on this very issue with a lot of lawmakers, from a valid point of view. Is it creating a different kind of economy? Is it going to create new jobs? Does this kind of policy, increasing IoT (cars, drones, roads) really help in the economic shift that people are looking for? Two, where do we look at the cyber security issues? Because clearly it’s creating a lot of additional security concerns. As you know we already have some. I think for those of us in the industry here, unless we can come back and talk about how the economic growth is going to be more broad spread and also how we’re going to deal with the security issues and risk mitigation at the same time, are going to have a big challenge in having that conversation.
Renduchintala: Sure. Maybe I comment on a couple of items? On the economic benefit of IoT, just go back to the example of autonomous driving cars. If you look at the amount of lost productivity the US economy occurs because of the amount of time we spend in traffic jams in cars in this country, it’s phenomenal. From memory it’s the orders of hundreds of billions of dollars. Really, if you can harness that productive time back into productive efforts, that in and of itself is a very simple example of just how a policy can liberate a tremendous amount of lost productivity in our organizations.
If you look at, for example, the ability to rapidly disperse, whether it be information or goods, to be able to allow people to make sure that they are able to live in a situation where they have a crunch for time. Being able to distribute goods via means that aren’t hampered by congested roads. For example, in the experimentation some are doing with drones is an example of how you are liberating productivity.
In the area of security, I look at it like this, we’re looking for pursuits that ultimately are valued by the people who are consuming them, and then there is a responsibility to maintain their security. I think there is a tremendous amount of innovation in going forward in heightening our security measures. That will just provide these guys with an even higher order of task to be able to surmount to this. I think there’s a lot of work going on in addressing some of those concerns.
Kirkpatrick: We’re going to have to wrap but one thing that I would say, and if you want to briefly address it is, and I think it goes to what Tara is saying about the policy argument. If this cannot be cast in this environment we are in right now in terms of the creation of jobs, it will be a political non-starter. Are you going to disagree with me when I say that?
Renduchintala: I think that the challenges we have will generate jobs and prosperity going forward. Because the solutions are going to solve problems which are holding back either productivity or the ability to use efforts we have maximally.
Kirkpatrick: Well I know this is a huge issue for you at Intel, but I would urge you to formulate your arguments along the lines of job creation, because if anything has been said in this recent time, it’s that we need that. Not that it isn’t even true necessarily in some cases, but it has to be talked about that way.
So, Murthy, thank you so much for being here. You’re really doing fantastic work.
Renduchintala: Thank you, my pleasure.
Transcription by RA Fisher Ink