As we move into the next era of mobile media development, situational awareness has got to be one of core areas for thinking and innovation. Part of this is a natural evolution of personal devices. They just need to get more personal. Part of it is driven by the growing dependence we all have on a small screen that is increasingly cluttered with icons. And of course the app structure itself is woefully siloed. We need another layer of intelligence to the OS that cuts across the clutter more effectively and better understands what we need now.
The role of context is one of the things we intend to explore at the annual Mobile Insider Summit in Tahoe. I am actively recruiting the brands and agencies that are exploring how mobility is changing their notion of customer context with mobile experiences that activate moments, change states, and more deeply register individual habits.
Context is not just about a place or a situation. It is also about me -- what I need at a given time and place.
Nokia just introduced an interesting step in that direction with its Z Launcher home page replacement for Android. It is in pre-beta, they say, but it is already fairly well formed. The launcher surfaces to the home screen only a select group of links -- upcoming appointments, and about six apps that it learns from your behaviors are most likely useful to you at different points in the day. I have only had the app installed for less than a day so it is difficult to attest to its learning skills. But the concept is quite good. I have been able to see selections change over time.
The other cool thing about the interface is its so-called “one-second access” to everything that isn’t on the launcher. All you have to do is draw letters on the screen and it starts running a search through apps and contacts or search terms. The Scribble feature is easy to get used to, especially compared to a keyboard.
This is the kind of adaptive interface Apple and Android should be baking into the OS itself. Personalization needs to be serious, passively rendered, and aimed at understanding the user in context. But that context is also intimate -- a product of how the user interacts with circumstances, times of day, places.
This kind of machine learning also puts the persona in closer touch with the algorithm itself. We get to see at least the results of a robotic entity trying to understand us. Much like activity trackers and the emerging platforms for the quantified self, we are at an interesting point in industrial history, where people are working in partnership with the machines at a more intimate level than ever before. You could see in this encounter two very different outcomes. On the one hand, the closer scrutiny of algorithms and how they work might surface their inherent and persistent weaknesses -- their inevitable non-humanness. On the other hand, one could argue that this kind of proximity and intimacy with AI diffuses a valuable wariness we will need to maintain about the rising role of machine learning, serving and control in our everyday lives.