Intel may be focused exclusively on making and selling chips, but that doesn’t mean it has its eyes closed to the rest of the technology market. The company has to make sure the ecosystem around its chips and the markets for computers and other electronic gadgets stay healthy. The result is some serious market research into what technology consumers want.
Intel’s research delves into ethnography, or the social science of observing people and how they behave in their everyday lives. Two of the top Intel people doing this work are David Ginsberg (pictured first below), who runs the Insights and Market Research Group at Intel, and Tony Salvador (pictured second below), senior principal engineer at Intel’s Experience Insight Lab.
I caught up with Ginsberg and Salvador recently over lunch together with a few other reporters, and we asked them about the kind of research they did for Ultrabooks — a category of fast, tiny laptops that resemble the MacBook Air. Dozens of Ultrabooks will be on display at the upcoming Consumer Electronics Show, where Intel chief executive Paul Otellini is giving a keynote speech. Here’s what they had to say:
Question: Tell us about your Ultrabook research.
David Ginsberg: The Ultrabook story actually starts a little bit in the past, about a year and a half ago. My team was doing some work on how people experience this notion of performance. You ask people what matters most about their PC, and they will tell you performance. You ask them, “What does that mean,” and they will tell you speed. You ask them what brings speed and, if they have an opinion, they will tell you the CPU. Then you say, “What CPU companies do you know?” and, if they have an opinion, they will tell you Intel.
We’re the company they know about, the maker of the chip that provides the most value in what they’re buying. But we all know in the real world that when they go to buy, there are a lot of other dynamics going on. We really wanted to take a deeper dive. Maybe they’re telling us speed matters because that’s the only word they know and can verbalize. But maybe there’s something else going on.
So we tried something different from market research surveys or focus groups. We used approaches that are based on understanding the emotional side of things. These tools were based on psychology and neuroscience, to get at the underlying emotions that we can’t really get in traditional survey work; and from that, we did a big project worldwide. We learned some pretty startling things that made us start to look at things differently.
We have a culture at Intel of thinking about performance, meaning speed and gigahertz and that kind of stuff. But what really came out of the study is, when people think about their highest performing computer experience, they verbalize that with the word “speed” because that’s the only word they can come up with. But that’s not actually what most people are thinking about. When we ask people to describe their experience, it was an experience of “flow.” It wasn’t just one thing. It was about the entire experience of using a device. It turned out to be much more about this notion of being in the zone or the moment — being in the “flow.” (Flow is the mental state in which a person is fully immersed and energized in a state of concentration).
DG: It’s almost like the technology disappears and you’re just one with your content, with what you’re trying to do. You might have only had five minutes, but an hour goes by because you’re so engrossed in what’s happening. When they’re talking about performance, it’s that experience that they’re thinking about, which is a very different notion than just amping up the speed.
And so this notion arises. How do we enable flow? How do we enable this uninterrupted seamless experience? Because what happens is, as soon as the video gets jaggy or that hourglass comes up, and it pulls you out of that flow experience, you suddenly remember oh, I should be mowing the lawn, or I should be doing the dishes. It jolts you out of the experience you’re having with your device.
That was a big realization for us as a company, and not just from a marketing perspective but from a product perspective. We really want to enable great experiences. We have to enable this notion of flow — or, as we translated it to our engineering community, to make it a little bit easier for them to wrap their heads around, this notion of “responsiveness.” Responsiveness is a bigger part of performance than anything else. So we started talking to engineers and the PC group about how to enable some new computer experiences.
This notion of flow is really at the heart of the Ultrabook. The Ultrabook obviously is a physical manifestation of the newest thin-and-light devices. But when you look at the four vectors that define the Ultrabooks from our perspective, it was more than just that.
Q: What are the four vecyors that define the Ultrabooks?
DG: Responsiveness was one of the key vectors. That’s why you see things like instant boot-up and all the other responsiveness features built into it.
Mobility without compromise was the second piece of it. People love the mobility their tablets provide for them, but there’s a desire for the performance that they’re used to in a traditional computer.
The third vector for Ultrabooks was the design that reflects me. Obviously, thin and light plays into that. Style and the looks of these things really matter today.
The fourth was security. We’re seeing this desire for security and the willingness to value it and actually pay for it, which is a huge shift in the way people are thinking about computer devices.
That was how we approached Ultrabook. It was really grounding how we thought about this and consumers’ needs, as opposed to just going about designing the next CPU architecture.
Q: With the design of the Ultrabook, all Intel really had to do was look at Apple’s MacBook Air, right?
DG: Well, that’s why we say it’s more than just about the thin and light. Obviously that’s a key piece of it, and there’s no doubt that the MacBook Air has shown how important it is; we’re proud of Intel’s role in the MacBook Air — in helping power it. There’s no doubt that thinness in general is an important thing in today’s market. But things like enabling responsiveness, enabling the security — those are huge parts of what we are trying to do.
Q: This research is all going on during what time period, relative to the launch of the MacBook Air?
DG: This performance research was end of 2009, first half of 2010. When we did that research, which got us to the notion of flow, we weren’t really thinking about four vectors. Over time, we had this notion of how flow applies to all technology devices. Whether you’re talking about a smartphone or other devices, flow is important. People want the technology to fade into the background so that they can focus on what’s important for them. I would argue that’s a truism across technology today.
Q: Can you see this flow ever becoming formalized or standardized like a service-level agreement for how well your technology or device works or your computer works?
DG: It’s actually really interesting, and I don’t know that I’m the right guy to go into all of this part. But a lot of folks in both the labs and in our benchmarking groups within the company are looking at, “How do we quantify this?” You move to some base metrics that measure the experience. How can you create an experience of base metrics that would allow us to actually do that? They’re doing some interesting work on that.
Tony Salvador: I look after the social sciences. And it’s our job to try to figure out what kinds of experiences people might like and how to design for them. We have a team of people who actually do what’s called “psychophysics.” It has been around since the late 19th century, and what they do is actually match physical capabilities like frame rate on video to perceived quality. You can chart that and you can see, “Is the framework slow?” I can still see frame, frame, frame and at some point, the single still image turns into a moving image, based on how my eyes and brain perceive it.
They actually do quantify the experiences in terms of these physical to perceived relationships and think about applying that then to a service-level agreement, for example, of I’m watching a video on my TV, remote control thing, or I’m downloading some data or something like this. To quantify that in the context of a flow experience is a very interesting proposition.
DG: And that is what some of the benchmarking folks are looking at. How do you do that? We came out with this research and we were so excited about it and we went and talked to some folks in Tony’s team and they were like oh, that’s just perfect.
TS: So they’re doing it from like a physical point of view, right? We look at the flow experience from a psychological view. There’s been a lot of research and brain study on flow. The science they call “brain work” has been really coming along, and people are understanding more and more about how the experience really works. It turns out you can be interrupted for about a minute when you’re doing some sort of writing. That’s when you are deep into it. Somebody can bother you quickly, you can answer them and go back to your work. Your brain structure is still in place to get you right back into it again. If it’s more than a minute, then it takes about 20 minutes to get back into that flow.
DG: What was interesting about the concept of flow is that it wasn’t meant to be applied to technology.
TS: That’s right. To do the work that your team did sort of questioned what performance means. From that point of view, it was really interesting work for the company as a whole.
Q: You can apply that to video games as well. Jane McGonigal talks about that in her book, Reality is Broken.
TS: They do apply the theory of flow to games. It turns out that flow, when you look at it this way, is actually this edge between order and anxiety. It’s that knife-edge when you’re working on something that is right at the edge of your capabilities, or you’re still able to do it but it doesn’t go off into the really high anxiety mode of, “I don’t know what to do. I can’t do it.” That’s the edge where people are just sort of totally engaged in games and can’t stop playing.
On technology and flow, we heard this coming out of the work we did with real people. It’s like when you’re giving a great presentation. You’re giving like the hardest presentation of your life and people are shooting questions at you and you’re just answering them left or right without having to stop and think. It just all flows.
DG: For sure. An Ultrabook has features like the instant boot-up. That’s one of the things that people have been frustrated by with PCs for a long time because maybe I want that better PC experience, the full experience, but it’s going to take me a minute and a half or whatever to turn that thing on. You’re not going to wait for that. There’s no way you’re going to get a full experience with that kind of thing.
So some of it is the instant-on. Some of it is the push, sort of faster read-write memory caches in there so that it uploads faster. Things don’t slow down as much. So you get less hourglass experiences and there’s a whole lot of technical stuff going on underneath to make that happen. On an engineering level, it’s about responsiveness. It’s about [making sure] that the computer isn’t interrupting you, because when a computer interrupts you is when you get out of a flow experience.
TS: Going forward, if you think about flow and maintaining and managing and encouraging the flow experience across a variety of activities, then the Ultrabook has to do other things as well. You can start imagining that it would have touch interfaces and “non-touch” interfaces along with the traditional ones.
Let’s say you are drawing a picture and you don’t want to interrupt your train of thought. The technology should let you draw a picture and not have to worry about the mouse thing or the little touchpad thing because you can actually use a variety of input devices to do the thing you want to do without having to worry about how it works. That’s a simple example; but if you think about all the decisions you have to make to make a computer, if you’re actually helping make those decisions based on encouraging a compelling flow experience, then you make different decisions.
DG: We’re the guys who will tell the rest of the company what the user desires are. What we have done, really for the first time as a company, is put in place some expectations into the Ultrabook requirements. We have to say what an Ultrabook needs to deliver. And obviously right now, we’re starting with the core essence of it. That is built around the look and feel and the responsiveness and all those kinds of things that are part of the specs that go into what makes an Ultrabook. It’s all about how we enable a better flow experience.
Q: Not every computer maker is going to be able to hit this experience right out of the gate, especially if you don’t want the machine to cost too much, right?
DG: Price is a big part of making an Ultrabook into a mainstream desire. This desire for these ultra sleek, ultra responsive, ultra secure PCs is something that everybody has. The people who are going out and buying PCs have a desire for these features we’ve been talking about. Thin laptops have been around for a long time. But this is about trying to change what computing means to people. We want to enable a different experience on the PC.
TS: So to expand the conversation just a little bit, one of the trends we’re seeing is that the kinds of devices, the form factors of the devices, is actually diversifying and it has been diversifying since the 1990s and it’s going to continue to diversity. So this notion that we will converge on one mega device is kind of – it’s put to rest.
There is no one device that everybody will converge to. There’s going to be a diversity of devices and a lot of it is shown by what David was saying. There’s a style to suit me. It’s not only a style. There are capabilities to suit me. One of the reasons that can happen is because the base technologies, the base commodity technologies across the board, have become less expensive, small, very capable, very reliable. We’re seeing an industry in transition, an industry where you’re seeing a much wider diversity of devices, sort of a Cambrian explosion of devices. Prices are going down and down, and as they do, you enable things that people have always wanted. The idea of thin and light computers has gone back to the 1990s. Now we are delivering it.
Q: What prevents people from reaching a state of flow with a laptop?
TS: Ultrabooks have faster flash memory. The slow hard drive access on a laptop creates a barrier to that flow. If you have to keep going back and forth then it sort of pauses. Just as you’re doing something, you get an hourglass. Crashing really interrupts your flow.
DG: Even things like the graphics and how video plays matters. If your video experience becomes jerky, that interrupts you from being immersed in the world of the video. You’re lost in the experience you’re having, and when that happens, the interruption reminds you that you’re using a computer. It kind of jolts you out of that experience that you’re having. So it even has implications at the design level of the CPU as to how to think about what we’re doing and where we’re putting energy.
TS: It goes back to the specs and quantification question.
DG: We do a lot of work on this. People want highly mobile devices. Everybody aspires to have a device that you can take anywhere. But the reality is that, for most people, it’s used within the home. I mean even tablets. Something like 75 percent of the time, people are using tablets or laptops at home.
So people want the high mobility that these devices represent but they also want to have that performance. They want a “no compromise” experience. So when you talk to people beyond the early adopters, they love that tablet experience and the fun of it. They love the ease of use. But they don’t like the performance. Mobility without compromise is addressing that.
Q: In productivity or entertainment?
DG: Both. I say that because, when you start to think about an Ultrabook a little bit down the road, not these first instances, you see that it can have a screen that is convertible from a tablet to an Ultrabook. It has the processing power to be a full PC or be a tablet. You will see more of these hybrid devices.
Q: Do you worry about too much fragmentation or diversity?
DG: Yes. I mean I think that is one of the reasons why the Ultrabook is trademarked and has got specs. We’re going to say, hey, if you want to qualify for the Ultrabook name, you have to meet these specifications. We worked with the computer makers on that. It’s not something that comes out of thin air. If I’m buying an Ultrabook, I know I’m going to get a certain type of experience and so that’s the thinking behind that.
TS: I’m going to expand on this a little bit. I think there are two directions you can go. We did some research back in 2005 or so. We had a notion for a particular kind of electronic book back then and we were playing around with a business group. People had tried for years to make electronic books. People wanted them, but they never bought them.
There were a few key issues. There was clearly the business model issue that has been addressed with app stores, and there was the issue of digital rights management. But there were also, it turns out, some physical issues related to the form factor, not the least of which was just its weight. The first really popular eBook reader, the Kindle (pictured above), solved that. Earlier devices were too heavy to hold.
With the convertible computer, or these Asus Transformer (pictured right) style of machines, [you have to] start figuring out the design elements that are critical to a detachable. They start figuring out the design elements that are crucial to a slider thing. They will start figuring those things out and then we will get them right and then we’ll move on to something new.
For instance, we have examples of natural user interfaces. Companies like Intel and Microsoft have been thinking about them for a long time, but it just wasn’t practical yet from a technology standpoint to have it actually work in the market. Now it’s out there with things like Kinect. It’s not perfect, but it’s in the market now. I think in the long run, there will be a significant move to natural user interfaces, like voice and gestures.
DG: Some people are going to love every new gadget that comes out no matter whether they’re 75 or 25, and some people are not. I do think when you look at how younger people are growing up with technology today, it is fundamentally different. I mean I look at my five-year-old or even my two-year-old. He constantly walks up to the TV and gets his grimy hands all over the screen and tries to swipe it because that’s just what he’s used to. He has grown up in a world of touch-enabled devices. So I think over time, you see these things shift.
TS: I’m going to give you a slightly different twist on that. This the hypothesis that I hold and we’ll see how it holds: I think we’re going to see a diversity of input devices. Some things are going to be the new natural user interface for a computer. But some things have actually so far proven to be quite resilient. The keyboard is actually quite good for putting a bunch of words into text form. It’s a hard thing to replace.
Will we always use the keyboard for everything? No, probably not. We’re going to use different kinds but you see that diversity of input devices. If you’re a painter, you’ll want a paint brush on your electronic device. You will touch it to the canvas with a very light touch or a heavy touch and you will get a different effect.
At this point, we believe it’s actually limited by imagination. It’s imagination that is not just ours at Intel. We have a good imagination. There’s a lot of really, really great imaginative people but there’s a whole industry of folks and people who are going to think about all different things that you can’t possibly predict now.
Q: If Ultrabooks are just one of a diversity of devices, what else interests you guys now about where trends are going? What other devices are interesting to you?
DG: Well, from the devices, to me the big thing is that the lines will blur between devices. We used to be able to say, this is a desktop, this is a laptop, and this is a phone. Those were the lines, and it was very obvious what was what.
As operating systems start to stretch across the devices, it becomes a little bit less obvious. We have to define the line between traditional laptop, Ultrabook, hybrid, tablets, small tablet, large smartphone, and other devices. The lines between the devices will disappear and that will get really interesting over the next couple of years. We have to understand what that means from a compute perspective.
At some point, the only real line, the demarcation becomes what can I fit in my pocket and what can’t. And the second is the thing I think we started talking about earlier is the convergence around data. There is all kinds of data that is now accessible to you.
TS: Well, first I think I like that idea of how you can assemble a gadget or take one apart and change what it is. You can carry it around like a phone but then reassemble it in a way that does something else. We may be interested in devices that change their identities. Like the wearable devices.
DG: I mean one of the things that both Tony and my teams do is work with the CPU planners, who at times have to be thinking five, six, seven years out, and obviously it’s hard to predict. But we try to look at scenarios for the underlying technology. At some point, you see a world of intelligence embedded everywhere. What will that mean for computing?
People are wondering about the desktop now. Five years ago, six years ago, everyone, including people at Intel, were predicting desktops are over. They’re going to be gone. The reality is, the desktop market is thriving. I mean, in terms of certainly price and all those things, the desktop market is a very strong part of our business and part of the ecosystem, in part because of form factor renewal with the all-in-one and things like that.
I don’t want to say everyone feels this, but there is a security in the desktop. I don’t mean in a virus-protection sort of way but the emotional security of having all your stuff in one place where you know it is. That really matters to a lot of people, and the desktop emotionally represents that to a lot of folks. It also has a big screen that people want. It conforms to you. When you’re sitting at the desktop, you can put your foot up. You’ve got your keyboard on your lap. You can move the screen. There’s something that people really like about that flexibility the desktop provides them. It is resilient.
DG: I think that processing power will exist in variety of forms in all aspects of the distributed network. I mean you need intelligence at all points of this network and at least what we assume with the Chromebook (pictured right, a Google-inspired machine that is always connected) is that it doesn’t need local processing power. In fact, it’s the opposite. So for a company like Intel, obviously we work very hard on the cloud.
TS: I want to pick up on this because it goes to the thought I asked you all before. Before we talked about this notion, we talked as an industry about convergence, and we don’t think that that’s happening.
So that’s from just an ethnographic point of view. We don’t think people are going to settle on one thing. There’s too much idiosyncrasy that exists, and there’s too many different kinds of things people want to do. One device won’t do it.
But where the industry has converged is on the notion of data and data mobility. For the last 50 years, we’ve been making everything analog into digital form. Now you have books and forms and driver’s licenses and all this stuff in digital form. But once you get everything in digital from, then you can start doing a whole bunch of new things, and now you have everything in data all the time.
The fact that we’re sitting here doing this is all data. You have this thing recorded. It’s all data. Once you have everything in digital form, then you can start running all kinds of analytics.
The data can actually be finding their own hypotheses about patterns in the data, and that can start running and happening on a continuous basis. That means you need a lot of processing power.
That convergence we think has occurred and is occurring on a going forward basis is that everything is data. The face that everything is movable across devices and everything has information about itself suggests that you can start creating new patterns and finding new value. It’s like the company that allows you to make your driveway in Palo Alto into a parking spot that somebody pays for.
So you’re driving down the street and you want a parking spot and you find somebody’s driveway. You pay them through PayPal or whatever. Then you park. Or you can give up your space here in the streets in San Francisco and sell it to somebody who’s driving around. It’s not even your space. You just have the right to the space.
If you’re starting to do this now and you expand it, you can do it for all kinds of things. You can do it for knowledge. You can do it for data. What you have, who you are, what you do can become a sellable commodity. You can do something with it, and that creates processing power not only in the cloud but also on the devices. As you’re walking down the streets, you can have sensors in these devices. This is a form of natural input. They will know where you are. They will know what you’re doing. What the temperature is. Where you’re walking. All these kinds of things not only provide you with the traditional location-based services and stuff but can actually give you indications of things that are happening around you and tell other people about things that are happening around you.
So one example of that is something called an Asthmapolis. If somebody has asthma, they have a little inhaler (pictured right) for when they have an asthma attack. It has got a little global positioning system (GPS).
When you puff, it sends a signal to your phone that sends a signal to the GPS that tells you where you were, and then that uses all kinds of other data like the temperature in the area, the pollen counts in the area. It knows these kinds of things. It knows your location. It knows the kind of environment you’re in, city, urban, rural, whatever; and then you can build out this massive database for every time somebody has asthma.
It’s a puff of an inhaler and you can start running patterns across that and sort of say, oh, look, these are kinds of triggers we have never seen before. These kinds of triggers affect this kind of person. You can actually relate it then to the person’s age or height or weight or other medical issues the person has.
And if you have all the data, you’re not relying on the sample size anymore. You’re relying on all the data, and you can start figuring out these sorts of little triggers, these little patterns, and now you’re a part of a value network that extends far beyond you and far beyond your asthma that’s creating an entirely new ecosystem of data. You’ll be creating data all the time, and that’s going to require a lot of processing power.