(From left) Jon Bruner, Frank Chen, Kerrie Holley, Dave Icke, Trae Vassallo
Jon Bruner (left) and Trae Vassallo
(From Left) Jon Bruner, Frank Chen, Kerrie Holley, Dave Icke, Trae Vassallo
Partner, Andreessen Horowitz
IBM Fellow, IBM Research, Scalable Business Services
Chief Executive Officer, MC10
General Partner, Kleiner Perkins Caufield & Byers
Director, Hardware and IoT, O’Reilly Media, Inc.
Vast potential awaits entrepreneurs and investors as everything and everyone gets interconnected. Where will we see the most exciting innovation? Read the full transcript below.
Bruner: Thanks very much, I’m joined by four terrific panelists this afternoon. Farthest to my left is Trae Vassallo who is a general partner at Kleiner Perkins. Next in toward me is Dave Icke, who is the CEO of a company called MC10. We’ll see a little bit more about what he’s doing in a few minutes. Then we have Kerrie Holley, who is an IBM fellow at IBM Research. And next to me here is Frank Chen, who is a partner at Andreessen Horowitz.
So I’ll start with Trae and just ask, why now?
Vassallo: I thought I got to go last.
Bruner: No, I’m sorry, you’re farthest from me.
Vassallo: All right. Now this is actually something I love talking about. I was explaining backstage, I am an embedded systems geek out of Stanford. I was kicking myself for the longest time because I studied mechanical engineering and I finally feel like my day has come, because building connected devices kind of seems like the latest and greatest new thing. The reason why I think it’s finally happening now really comes down to smart phones. So now with billions of these devices out there now, the cost to build a smart connected thermostat or a light fixture is orders of magnitude different than what it even was two years ago. So there’s the bits and pieces inside the phone that are mass produced in such high quantity that it’s now cost effective to put them in sort of smart phone like stacks and embed them everywhere. But also everyone now has an interface to the world in their hand. So we talk a lot about how our digital lives on the Internet are very full. It’s time to bring the rest of the physical ecosystem around us that is equally if not more important into that same wonderful digital paradigm. So I think that’s part of the reason why. The technology’s finally here and customers want that same kind of great digital interface.
Bruner: So that ligament is here, the connection to the network is here, through the phone.
Bruner: Fantastic. Frank, Mark Andreessen has famously said that software is eating the world.
Chen: [makes eating gesture with hand]
Bruner: That’s right, it’s replaced by a gesture now in fact, because it’s such a crucial thing that people say. But now we’re talking about devices a lot, connected devices, hardware interfaces, things that act in the physical world. Is that a counterargument to the idea that software is eating the world?
Chen: Well, that’s a great question and we actually think of it as the complete embodiment of software eating the world rather than as the counterargument. So fundamentally these devices need to run software. They need to send their data to clouds that process that data. And so it’s the whole ecosystem together, hardware, network, a cloud storage solution that analyzes and takes action on it. So you need all of those pieces together. And we couldn’t be more excited, we’re just like Trae on this, now is the time.
Bruner: And Kerrie, you’ve worked with a lot of big companies and governments at IBM. Big companies and governments, as it happens, have been the leaders in a lot of connected devices for some time. And the idea of an instrumented jet engine is not new, right? But this area feels like it’s accelerating, so what’s going on among these sorts of companies that’s bringing about a renewed interest in connected devices?
Holley: Well, I think the big interest in large companies in connected devices is wearable computers and this whole concept of Internet of things. Most large companies are trying to continue to grow. They are trying to have a very unique, differentiated customer experience, one that enables them to grow, attract more customers. So to a large extent, what we find is a lot of organizations are trying to engage differently. And most companies, which we agree with, which we are pushing as an agenda, we actually believe that looking at these trends as macro trends is interesting, but the actual convergence and confluence of those trends is actually more interesting in terms of when we look at mobile and social and analytics and cloud and context-aware computing and artificial intelligence. When we look at all of those trends collectively, we see a much more powerful experience. But at the heart of that are all these endpoints. And these endpoints are only limited by the imagination of you and I.
Bruner: Do bigger companies enjoy any advantages over startups in developing interesting connected devices, connected world type technologies.
Holley: I don’t know if they offer an advantage. I think we’re seeing a shift in the marketplace. What most large companies are experiencing is that these companies that we can describe as born on the Internet or born on the cloud have increasingly created these web-based platforms as strategic control points, they’ve increasingly, through media and through other venues, have shown new channels to market and they’ve shown an opportunity for innovation and collaboration that’s been fundamentally different than what big companies have done. But what we’re seeing is a shift, this whole democratization of IT, the consumerization of IT is causing Fortune 500 companies to recognize that they can no longer innovate within the four walls of their company, that they will actually have to not just use third parties like IBM or Google or whatever, they’re actually going to have to embrace an open community. So that’s the shift that we’re seeing in the marketplace.
Bruner: So they’re pushing openness?
Bruner: And Dave, I’d love you to explain to the audience a little bit about the technology that your company is developing and I understand we have some photos set up as slides that we can put up on the screen.
Icke: Sure, well, MC10 is a venture-backed company. We’re based on the East Coast, Cambridge, Mass. And what we’re trying to do is take conventional computer chip technology out of rigid, boxy packages with the premise that Moore’s Law is not enough, that if you want to interface with soft tissue on the skin or inside the human body, you need to take it to another level. So you can see, there are a couple of slides that show conventional chips being thinned and then deployed in new soft form factors. We’ve focused first really on interfacing with the human body and being able to deploy sensing, some computing, and then ultimately the ability to deliver a targeted therapy really anywhere on the skin, depending on what you’re trying to measure or eventually even inside the body. So we’re in the process of, I guess, capitalizing on this wave of distributed computing. We can certainly do it in industrial settings as well. But oftentimes, a rigid, boxy thing is okay to attach to a piece of industrial equipment. Most of us don’t like to tolerate wristbands or chest straps or things on our heads. So if you can lower the barrier to adoption of wearable computing, you have an opportunity to really accelerate deployment of what you can do in terms of connected healthcare and wellness optimizing activity and also participating in the Internet of everything.
Bruner: A lot of healthcare challenges come down to getting the patient to just follow the prescribed treatment, like applying something to your skin.
Icke: Sure, compliance is a big issue, whether it’s taking a medication or doing a follow-up. If you think of a cardiac patient discharged and what a Holter monitor looks like, or if you’ve ever done a stress test in an EKG lab, today you’re clipping huge belts with heavy pieces of equipment onto you, multiple electrodes with lots of wires hanging off. Same thing, when a child is in the NICU. So if you can break that paradigm, use the power of wireless, use the progress that microelectronics has made in giving the power into the smart phone, but instead remove that boxy form factor and interface with the body in a new way, I think you can unlock lots of applications.
Bruner: Do you think that big institutional buyers of healthcare devices, the hospitals, the insurance companies, see the promise in this? Are they willing to buy things like this or are they still to risk-averse at this point?
Icke: You know, I think that there’s no question that there’s that something has to change in the healthcare system and getting access to care in an episodic way every once in a while when we go to the hospital or the clinic or when we happen to end up in the emergency room, when something bad happens, is a pretty inefficient way to access care. So it rides along the same lines as some of the interconnected systems for optimizing performance or efficiency that we see in factories or in transportation. The same kind of thing happens if we can be connected all the time, we can catch things before they become a big problem. We can notify healthcare providers or notify you and you can get attention much more proactively.
Bruner: So interoperability becomes a crucial factor in that model then, right?
Icke: Absolutely, so no single company—especially not a startup—is going to figure the whole thing out but it really does require an ecosystem where big companies and motivated players that are attacking different parts of the problem can work together with some common standard and common vision of where this needs to go.
Chen: Can I piggyback on?
Bruner: Of course.
Chen: So one of the things we’re really excited about when we think about wearable computing is really breakthrough user interfaces because anybody who has typed anything for a long time on their phone knows that this is not going to be the last word in UI. And so if you think about wearable, you think about gesture recognition tied to natural language processing, tied to intuitive interfaces where you can just sort of wave your arms. The keyboard and mouse have dominated our world for longer than you would have thought. The mouse got invented right near this building. And the opportunity to meld vision plus natural language processing plus gesture recognition to do something that is very intuitive is a great opportunity.
Bruner: And we have Trae over here on the end who is wearing what is still at this point a very conspicuous device.
Vassallo: Yes, I am a walking wearables lab, actually. And I’m embracing the full geekiness of it. I’ve got every wristband known to man. I’ve got a Jawbone, a Nike, I currently have my Basis [ph] and then I’ve been test-driving Glass for the last three weeks. Just quickly, I think there are going to be a couple of wearable categories that emerge. We’re already hearing a lot about it. There’s a lot of speculation around the wrist and what the big platforms are going to do around the wrist. I am super excited about that for a number of reasons. First is that this is highly socially acceptable. It may be ugly, but it’s really easy for me to wear a watch. And the thing that I’ve learned, especially about wearing Glass for the last three weeks, is the hands-free, always-on nature of my access to the Internet through my glasses is truly transformative in a way that is hard to explain, you actually have to experience it. And I can be doing anything and with a command I can super easily send myself an email, send someone else an email, get directions. And that is really, really powerful. And I think what’s exciting is that that means there’s new platforms being built that have very different inputs and outputs like you were saying. What is different about Glass and what is going to be different about watch platforms? Well, they’re going to be very motion driven or audio driven and it changes things in a really fascinating way. So the reason I’m excited is, look, these are baby steps into a whole new area and they’re quite kludgy right now and I’m still amazed that these wristbands have sold as well as they have. There’s only so much you can do with an accelerometer. It’s not that interesting in the grand scheme of things and frankly, number of steps? You know, do I really care about the number of steps I take every day? Well, sometimes I do, but the mass population is not going to be engaged and motivated by that on a daily basis. But would I love something to tell me how to live better and smarter and then socially compare me to my peer group and give me advice? Absolutely. And so with new sensors and new configurations, that is in our very near future, and that I find incredibly exciting.
Bruner: And so do you think the value in things like Glass, the wristwatch pedometers, is going to go to the companies that are making these devices or that are sucking the data in and then learning a lot about you as the wearer?
Vassallo: So personally, I think it’s going to be a combination. I think there will be several parts of the market where—let’s just take Nest as a great example. Because that is one of our portfolio companies and that’s a perfect example of a really smart team that identified a mundane product that everyone has with forty-year-old technology and they said, “You know what? We can reinvent what a thermostat really means.” And so I think there’s stuff all around us that just hasn’t gone through that transformation yet and so I think there are opportunities for startups to really focus on something that’s transformational and be a big company. But the big giants have woken up. And I think what’s going to be so exciting is you know that Google’s doing something, there’s rumors Apple’s doing something, Microsoft’s probably doing something. So it’s going to get pretty exciting from a platform standpoint, I think, in the near future.
Holley: I was just going to add that I agree fully and I also think the ecosystem does have enough room for a lot of players. So in the scenario that was just mentioned, clearly the device manufacturers are going to win, the platform providers are going to win as well, but also the organizations that provide to consumers. So if we look at medical facilities, hospitals, it’s a big win for them in terms of lowering costs, in terms of better diagnostics, in terms of better healthcare. If we look at the cross-industry relationships that will grown in terms of healthcare and, let’s say, grocery chains, there will be a big win for those two in terms of perhaps more healthy eating. The scenario you described, in terms of wanting to get data, and not just data, insights is actually what you’re looking for, which is what Trae mentioned, insights that would actually improve your life. So the consumer wins, the organizations that provide those insights win, and those organizations are going to be multifaceted, they’re not going to be just one industry; it’s going to be multi-industries.
Bruner: And Frank, as an investor, are you interested in the platforms or in the companies that own the user experience?
Chen: Well, you hit upon probably the most active debate that we have internally about this right now, which is, is it horizontal? Is there going to be an Android of the Internet of everything? And there’s this base software and then lots of people innovate around it. Or is it going to be vertical? So companies like Nest who, you know, Tony Fadell pops out of Apple and builds exactly what you would expect an Apple alumni to build around thermostats: beautiful, engaging, completely hermetically sealed. It’s not a platform to run other people’s software. And that is really fundamentally the biggest question, which is where is the value going to emerge? And as an investor, you’re trying to pick the place where all the value gets created and I would say, look, it’s an open question right now. In the long sweep of our industry, openness and horizontal has won. Now, over the last five years we have this blip where the most valuable company on the planet got created on vertical, which is, “I’m really in charge of my ecosystem.” And so for the first time, Apple demonstrated that you can build a ton of value by tightly controlling everything from the operating system to the store to payments to the end user experience. So it’s an open question and the one that we are most actively debating internally right now.
Icke: I would just add that I think the user experience has to be mindlessly simple for this to be broadly adopted. So you’re not going to be able to neglect the user experience and even if there are collaborations and an open environment, a set of standards where people are collaborating, the user experience to move beyond the early adopters has to be super simple. And that is not an easy problem, especially when you start looking at things like healthcare.
Bruner: And on the continuum between a sort of point solution and a platform, where is MC10?
Icke: Well, it depends how broadly you define the problem. If you think about connected health, there are lots of companies that are doing wireless connectivity and smart phone apps and big data in the cloud and trying to derive the insight back to the user. We’re focused on seamless sensing, which is kind of the frontend input to that, and really lowering the hurdles for doing that. We’re trying to do that across a number of different problems and industries. So we think of ourselves as a platform technology but we have a very focused piece of that overall solution that we’re contributing to.
Holley: And I was just going to add that we as a company have a pretty long history of vertical success and seeing the downside of that. So I think actually from a market standpoint we’ll see the ebbs and flows. We’ll see companies with vertical solutions that will make a lot of money. But we actually believe that the industry is in a transformational stage right now. And you and I talked about this a little bit before, but we believe fundamentally that if we go back in time that we started with this sort of pre-electronic, mechanical tabulating machines, then we went quickly to programmable computers, which is where we are now, and we believe that we’ll quickly move to this era that may not stick in terms of what we call it, cognitive computing, but it is fundamentally different. It is the sense and response. It is the insight. It moves away from von Neumann’s model, where you organize a computer around memory and processors and CPUs and so forth, to a model that is software defined. It embraces visual analytics, embraces big data, embraces machine learning, embraces context-aware computing. And that’s the world that we actually see. And we believe that if you look at each of these eras, they last about forty to fifty years. We think we’re in the first one to five years, who knows? Maybe it’s year one, maybe it’s year five. But we believe we’re in an inflection point and the actual transformational opportunities are enormous. And we won’t be the single company that pushes that forward, it will be everyone on this panel and everyone in this audience collectively in their efforts and contributions. But the point I’m making is that if we look at the micro stage, I think we limit our understanding of the possibilities of the impact that it will actually have to our lives, to cities, and to societies.
Bruner: So you actually see this as changing the type of computing to something that’s better suited to interacting with the physical world?
Holley: Exactly. And I also see, you know, even though we talked about two worlds, digital and—I don’t know what was the second? Physical we called it? But we actually see three worlds. We call the world of physical all the stuff that’s got embedded devices, Internet of things, everything that’s sort of physical, that you can touch and feel. And then there’s us, as people, that’s a world as well. And then there’s the world of things that we built—software, systems, IT. But those are three distinct worlds that we see there that are actually blending today into a powerful new opportunity space.
Bruner: I would love to come back to the Nest, because this is something that we’ve brought up a handful of times and also we have a nice talk from Alex Hawkinson early today.
Vassallo: Yes, who’s running away, I see.
Bruner: And I always enjoy hearing Alex’s presentations because I think he lays out a very compelling argument for the value of platforms, the value of bringing the physical components into an abstract relationship with the software. You’re an investor in Nest though, which we often talk about as being on the opposite end of the spectrum from SmartThings.
Vassallo: I think that’s exactly right. They’re two very extreme ends of a spectrum where I think there’s a lot of opportunity. I’m a huge fan of what the SmartThings guys are doing and think they could be entirely complimentary to an ecosystem that emerges. Especially in emerging new areas, vertical solutions are kind of needed to pave the way and show people what’s possible and what can be done, which I think is oftentimes why you see these vertical opportunities arise. Customers want to buy something that works and they want to buy something that solves a problem. People don’t buy platforms, they buy “This things going to help me do X and oh I’m really happy because it’s going to do all these other things too.” And so I think the key for platforms is first of all getting some of those developers on the platform, creating momentum, but then ultimately, it’s about creating solutions that really resonate with groups of people who are excited by it. And the thing that I think is exciting about what Alex and SmartThings team are doing is that there are more things in a house that you could ever think about tweaking or designing for. And there could be three huge Nest companies out there and they’re probably not going to make the connected coffee pot or the automated dog feeder but there’s going to be a bunch of people that want that. And so to have that ecosystem that enables that I think is pretty darn cool.
Bruner: Silicon Valley has a way of designing the products that Silicon Valley wants to use itself and occasionally that’s—
Vassallo: True and so that is one of the questions. There’s something here, but the question is how broad is this?
Bruner: I know a lot of Google engineers and their preferences are not identical to the preferences of the American public as a whole.
Vassallo: Yes, and I will say I’ve wreaked some havoc on my family too. Deploying a lot of this technology in the house, I’ve been getting some flack.
Bruner: Mom, are you recording me now?
Chen: Look, we’re having a big sort of revenge of the nerds moment here, which is you’re right, Google engineers don’t represent mainstream, but for a generation Hollywood taught us what normal and mainstream was going to be and I think there’s an element of that that’s sort of coming back to Silicon Valley, where Silicon Valley’s going to teach what mainstream behavior is going to be. So a good example of that is that we have an investment in a company called If This Then That. And if that’s not a revenge of the nerds–titled company, I don’t know what is. But it’s a programming framework that allows you to connect any web service to any web service. If you see this post on Craigslist, then email me. If I get tagged on Facebook, then retweet it. And the whole Internet of things is dying for a platform like this. When I open the garage door, then turn on my stereo, turn on the thermostat. So there needs to be a programming framework that allows us to stitch all of these Internet-enabled coffee pots and garage door openers and washer/dryers together. And so I think it’s not farfetched to think that in a few years mainstream people will be doing rudimentary programming of exactly that sort. And then we’ll have reached our revenge of the nerds moment.
Bruner: So I’d love to turn from consumers to government for a second. And you’ve worked with governments with IBM on things like smart cities initiatives. What do you see governments wanting out of connected world? Are they willing to take risks on new technologies around the Internet of everything? What do you see there?
Holley: So multiple questions. The answer’s yes in terms of risk. The challenges governments have we see in the United States, and we see elsewhere, is they have limited funds and they have lots of both political challenges and social challenges. But what we see is that each city has a different set of problems. So in Singapore, we saw a city that didn’t believe that you could actually identify patterns to actually make a difference with traffic, but you could. You could actually use not just smart technology in terms of traffic lights and so forth, so intelligent traffic lights, but you could also use pattern and data analysis to actually make the city go faster. Who knew that—I don’t know the exact algorithm but most traffic problems in cities are caused by people looking for parking spots. So if you can actually direct a person to a parking spot faster, you can actually alleviate the traffic. In Brazil we did a project, we have a piece of technology called Deep Thunder. A lot of people have heard of Watson and Deep Blue. But Deep Thunder is a piece of technology about weather. And in Rio de Janeiro, where the Olympics are coming, they’re very concerned about many things. They’re concerned about crime, they’re concerned about weather conditions. So for example, maybe I don’t want that Olympic diver to jump off the board at this particular time if the wind is going to be a certain—
Bruner: Or if there’s lightning in the area.
Holley: No, not lightning in the area, actually using data based on a weather pattern that the national weather whatever has forecasted, to be able to take that, plus what’s actually happening through sensors on the ground at the moment, within a few minutes or hours to be able to pinpoint what the actual condition will be for that diver at that point in time.
Bruner: So system-wide intelligence applied in a very concrete way.
Holley: Yes, so those were a couple of examples, but we’re seeing cities embrace technology for making citizens’ lives better. I did a project in New Orleans for this very purpose. So we’re seeing cities very eager to invest, very eager to see the returns of being able to use technology in innovative ways.
Bruner: They have a lot of physical assets to control and measure and so forth.
Holley: They do.
Bruner: A lot of manhole covers to instrument.
Bruner: I’d love to have some questions from the audience. If we could just bring up the lights a little bit—I think there are microphones.
Q: My name’s [off mic], I work in [off mic] currency. I’m really curious how you see virtual currency and digital currency meshing into the Internet of things. We’re working a lot on machine-to-machine payments. I’m just wondering how you see that happening, especially with regard to OpenCoin, which you guys just invested in at Andreessen.
Chen: So we’re big fans of the digital currencies. We’re starting to place small bets in and around the Bitcoin community, like with OpenCoin, and we think that another revenge of the nerds type invention, which is instead of fiat currencies from governments that can keep on printing money, let’s invent a digital cryptography-based currency that basically has a fixed supply and can’t inflate, just sort of by the nature of the currency. Nobody can run the printing press on Bitcoin, the mathematics don’t allow it. So we’re big fans of that concept and the big question is how and when does that go mainstream? And I think the intersection between digital currencies and machine-to-machine is it’s tied to the whole, like, well, when will it become mainstream? It’s got to get consumer adoption.
Q: As you know, Bitcoin passed a billion under management and in ten days it was two billion. But there’s some real problems with Bitcoin in terms of the way the architecture is created and one of them, depending on if you’re a human optimist or not a human optimist, is the ability to control that. So our currency, Ven, runs through a central reserve and is 100% backed by commodities, carbon, and other currencies. So it’s fundamentally the opposite of Bitcoin. Because the idea that the coin is not actually backed by anything and that it’s anonymized means that corporations can’t really use it right now. So how do you see that adapting to become something that people could actually use in the light economy versus the dark economy?
Chen: I think evolution of any sort of payment system or currency is going to come down to adoption. So that’s what I was saying earlier, how do you get people to adopt it? And there’s always a chicken and the egg, right? Do you start with the merchants who will accept it? Do you start with the people who have it? And I think it’s just an open question. How you incent people to change their behaviors around how they pay.
Bruner: I’d love to make sure we have time for a couple more questions. Anyone else have a question for our panelists? In the back, there on the aisle.
Pell: Hi everyone, I’m Barney Pell with QuickPay Corporation. I’m really enjoying the panel. So my question is all of this Internet of everything requires data connectivity. What do you see as trends where everything will actually be connected? For example, is it city WiFi, low bandwidth, or low energy Bluetooth and stuff? So that everything can reliably be connected to everything else?
Holley: I will take a partial stab at it and let my colleagues here expand. Throughout the conversation of the day, there has been a lot of conversation about the physicalness and the physics of connecting everything and the speed at which that will occur based on spectrum and blah, blah, blah. But I think that we are seeing and we are witnessing this API economy that’s forming and I do think that this notion of using this web API, the web as a programmable platform actually provides a great opportunity to create greater connectivity in a way to allow us, as you say, to stitch, to assemble, versus having to custom develop everything. And a lot of that’s going to be able to be very dynamic, to be done on the fly.
Chen: Barney, I think this is going to play out like Obama’s energy policy, where he wants all of the above and I think that’s what’s going to happen with Internet connectivity. In the house, it’s going to be Bluetooth, low energy, plus ZigBee, plus Z-wave, plus WiFi, plus things that haven’t been invented yet. And similarly for metropolitan areas, there’s going to be WiFi, there’s going to Linus Light, there’s going to be microwave. And the challenge is that we put a software layer over that so the developer doesn’t have to care. And that’s sort of the thing that we’re looking forward to. Radio agnostic, maybe a little software defined radios, so we don’t have to ship new custom-designed radios into every device, and then a software layer that makes it so that you don’t care how things are connected.
Bruner: Dave, as you’re developing medical devices where it’s crucial that they are low-impact, do you find that the technology for radio communication that you need is available now?
Icke: Yes, I think it is. It comes down to a tradeoff between how much data you’re trying to communicate and upload to a smart phone or to the cloud and how much you try to compute locally. So you have to be thoughtful in architecting a system as to what the power budget looks like and how you want to deploy that. But I think there are good alternatives already. And technology trends will continue to drive lower power and higher bandwidth.
Bruner: Well, thank you all so much for joining me on the panel today and this has been a terrific conversation.