For innovation to really explode, we may have to rethink traditional ways of protecting proprietary information. Is it time to leverage the latent opportunities hidden in open datasets?
In today’s hyper-connected world, access to information fuels innovation and the progression of ideas. And, as information omnivores, we expect access to an open web. Google’s mission statement “to organize the world’s information and make it universally accessible and useful” becomes more important than ever. With the Internet of Things — the idea of that objects can be interconnected through sensors and other embedded devices — moving from a mere blip on our radars to an idea spoken about with some familiarity, it’s becoming clear that our expectations around Internet openness should be no different.
Open source access to data can play a big part in giving developers the tools they need to build unique applications and services. Aiming for the seamless interoperability and fluid transmission of data will allow new systems and protocols to be built that will challenge traditional notions of ownership and privacy.
To learn more about this trend of open source access, PSFK met up with John Sherry, director of User Experience Design at Intel Corporation, and Brandon Barnett, director of Business Innovation at Intel Labs, and asked them a few questions.
First of all, how does allowing open source access benefit the people who create or own this data, and what are some of the current barriers that proponents of this idea face?
John: Well there is a huge opportunity to spur innovation off of data, but the way we’ve been thinking about it is almost as a search problem. How do you find those nuggets of value in all that data? It really requires a few things to be in place. One is the data has to be interoperable. If you are going to find nuggets of insight, it’s going to be because different datasets are being brought together or being juxtaposed. Except, without too much work cleaning and normalizing and making those datasets compatible, how do you bring them together?
Then there’s making that data be able to flow. Right now, so much data just goes into the silos of different organizations who, either with their devices or through other means, are helping that data to be generated. Those companies and organizations that generate that user data will have to take different attitudes about how they’re sharing it back with users and potentially with other developers.
Finally, we need better tools for analytics. Actually making sense of that data is one of the things we’ve been working on a lot in our group. How do you provide tools back to the people who are looking for insights in that data, questions they are trying to answer, but who don’t have advanced degrees in data science? How do you make it easier for them to play with, understand and start to explore the meaning in it?
Brandon: In addition to the technology, business models need to change. Our hypothesis is that there is much more value and actual business opportunities if that data is open and interoperable so that communities of developers can find ways to discover new value based on their own views of the world and expertise.
Certainly the technology needs to be in place, and the business models must adapt to allow all stakeholders in that new community to benefit.
Will a shift like that require rethinking the indexing of this information as well?
Brandon: I don’t know if indexing is the right word, but we’re going to need to invent new tools for creating an ad hoc interoperability of data. There have been efforts before, with things like Semantic Web and other standardization efforts. With the explosion of new devices and new sensors, we’re going to have new data types emerging and so many different ways of measuring things, different frequency of measurements and different units that we’ll need new tools.
Do you see this idea of interoperability between devices as a key consumer consideration?
John: I think we see this already starting to happen certainly at the bleeding edge of this with examples like the Quantified Self Movement, and I think it’s going to grow.
People who have a particular problem to solve or answer they’re trying to get to are going to great lengths to cobble together diverse data sources to help them figure out what’s going on. Whether it’s a health issue, or an athletic performance issue or something else like that. Ordinary people have those kinds of questions as well, but just may not be willing to go to the lengths that the Quantified Selfers have gone to. So making it easier for devices and data sets to talk to each other is a must.
And it’s not just about devices talking together. People have to be able to do this for their own reasons and purposes. One other thing we found early in our research is that people don’t just want a device that tells them if they’re doing good or not at some goal that has been arbitrarily set by some programmer of a device. People want to be able to identify their own goals, problems or areas of need, and to pull in all the right information to will help them understand that.
Can this extend into practical value with people building an application with this information once it’s available? Could it be used to spur markets and economies?
John: I think it will. I think that the interesting thing here is, “What’s next beyond the app store model?” How you provide a wide range of services and tools, widgets and different kind of analytics that are analyzing everything from your speech patterns, to your physical activities, to your spending habits? How do you make all that information interconnected and recompostable and weaved with new insights?
What we’ve been seeing over the last year is that line between consumer and producer and knowledge creator merge. We have people who want to create schemas or create these macros that they can then share with other people who will be able to take that example and tweak it and do something that’s appropriate to their own lives.
The app store model won’t work for something like that. If we’re going to see an evolution toward a new model that maybe it has to address that large space of connectable things. That’s something that has yet to be created in any great way. If done right, it could really unleash just a ton of creativity, and an explosion of insight. Not just of consumer behavior but of really creative use of data, and the creation of knowledge and information.
Brandon: I agree. For instance, If you’re monitoring temperature in your home, you know what kind of sensors and measurements and algorithms you need to do that. But having this vision of open and interoperable data allows others to come in and say, “I think the temperature in my house when combined with the weather patterns outside and the humidity and my own personal health can do something else.” Now, I have the capability to experiment with that combination of data and break across silos. That’s where it gets really interesting. The opportunity to create something truly novel opens up with so many possibilities.
Who is responsible for this analysis, and who really benefits? What potential applications could people build of to leverage it?
John: The benefits run up and down the chain of the human social graph. From their organizations, their personal individual lives, all the way up to much larger collective benefits.
We’re working with a team right now that has an analytic tool for looking at language-use patterns to help smaller organizations. By looking at language-use patterns, we can maybe gauge the mood of an organization of people, and see recurring issues that keep cropping up but never reach them on a conscious level. We’re trying to see how to help organizations tap into what’s collectively on their minds even though nobody, individually, was aware that there was something bothering everybody.
There are so many applications for data with open source access both at the individual and collective levels. Personal health, and all the way up to cities with understanding traffic flow, understanding where economic development is needed, understanding where latent opportunities might be hidden.
Brandon: Looking at it from a top-down view, we see small uses that could perhaps be aggregated for large applications. But we’re also seeing that in the open data space, the opposite as well is true, where datasets that were collected for government and federal purposes are being opened up and used by individuals. You can see this with transportation or education data, with human services or even with FEMA. All of these are datasets which weren’t necessarily gathered through IoT, but were sensed somehow and contributed and stored, and now we see that IoT‑type of capabilities only augment that.
The federal agencies have their own purposes for using those datasets. When opened up we’ve seen, through our work with incubators and accelerators, people can combine those big open datasets with personal-, local- and community-level data to get new values. Something like looking at EPA data, combining it with my own personal wellness data and using that to decide if I want to go downtown today, because I’m seeing that there’s some potential health issue that could be going on. Data from up and down the chain can complement completely different datasets to create solutions from large problems to very local community and personal problems.
What are the main challenges around privacy, and security, in a world of open data?
John: Trust. For all of these ideas to work, people are going to have to trust that their data is not going to be misused or used in ways that they don’t warrant and authorize. That’s actually a big part of what’s wrong with how data collection and use happens today.
We’ve obviously had big news headlines around the NSA and some other violations of the social contract. But I think it happens more generally. We’re starting to see backlash against this notion that large institutions are somehow entitled to gather a lot of data about us and use it in ways that only suit their needs — backlash against an evident lack of regard for what individuals want and think should happen to that data.
Another piece of the puzzle that has to be in place to make this all work are both technical and policy measures that help people feel secure in sharing their data. We need to make sure people understand what the quid pro quo of sharing that data will lead to, and what the obligations are on both sides of any exchange that involves data. All of this will require more transparency and recourse into abuses of that as well.
A lot of those tools don’t quite look finished yet. Those are things luckily that Intel is working on, among others. I think that’s a real contribution we can make as a company.
Any final thoughts about the direction of this? What does the landscape of open source access look like in a couple of years.
John: Good question. The landscape I would hope to see in a couple years is one I described earlier where you just have lots and lots of contributions from people who might be specializing a little bit in analytics tools, people or entities or companies that are specializing in helping users negotiate the value back that they deserve from their data, or helping users reaggregate their own data and mine it for personal insights.
Above all tools to enable that sort of open sharing and interoperability — and kind of collective search for insight — that’s not quite happening today, but I think we’re right on the cusp of it.
Brandon: How near the cusp, I think, is yet to be seen. What the data suggests is that certainly there are entities using sensors and analytics to provide very specific services in many verticals and niche markets.
What I agree with, is that we’re starting to see what is a disorganized network of companies and products that are probing this new consumer-facing distributed value from things on the Internet.
But it’s not there yet. We don’t yet have an Internet of Things.
What we’re interested in is how do we look for that phase change, that change from a disordered network of value creators to an ordered value network. What role can Intel play to help move that forward, and solve problems for people who really would benefit from the existence of such an ordered network?