Instagram and Facebook customize news feeds based on what you like. Pandora creates playlists for you based on just one song you like. Amazon knows when you run out of what and suggests fast and simple buys. Uber can take you somewhere and back without you having to make a call, all based on the location history and the probability of returning time. Internet of Things is the concept that was being actively explored just a couple years ago is now a reality. Google’s Nest regulates your house temperature according to your preference analysis. Smart everything is around us.
The Idea Behind Anticipatory Design
What makes these things work is the concept of predicting user’s action and helping them get rid of the abundance of choice they make every day. This concept is called Anticipatory Design. It emerged as a natural progression of technology and psychology synergy. How can psychological problem of decision fatigue be solved by means of newest technology? Is there a way to teach AI learn from human behavior, predict choices, and suggest actions? What are the chances of anticipatory design hitting the bull’s eye and what happens if it fails?
The basic elements of anticipatory design are the following:
- Machine Learning
The power of probability math in human-machine interaction gives it a universal tool to fascinate, distract, inspire, terrify, and what not your life. What it gives you is the abundance of options to choose from, even when you don’t require that choice! Let’s look back in time a little bit: in Soviet Union where I was born, there was not a lot of options to choose from in terms of commodities available and things to choose from. There was no Amazon or Ebay, or anything remotely close. As the result everyone looked pretty much the same, owned similar things, and when finding stuff was not a national sport, the people spent time doing other things.
One of the favourite New Year’s Eve movies in Russia to this day, called The Irony of Fate and its comedic brilliance is based on the idea that people throughout the country live on identically named streets, in identical apartments, packed with identical stuff. Truth is back then, people did not have much of a choice and were naturally attracted to the idea of great amount of things to choose from. It was a matter of time that technologies developed to the point where we started producing way more things than we can consume. This is here it became evident that in order to sell your product or service, you need to advertise it and make it noticeable. The media and later, the Internet made the choice profusion so obvious that it became the hallmark of our generation and the prerequisite at the turn of the century.
The sheer knowledge of all things available and being there for you any second is what has complicated our life counterintuitively to what it was originally intended for. Such thing as decision fatigue is a real problem considering the importance of choices you have to make through the day. The ultimate goal of anticipatory design is to technologically kill the decision fatigue and save us the time to live life.
Our life is frittered away by detail. …Simply, simplify.
― Henry David Thoreau
This resulted in a number of natural filters that we began to apply on the information flow pouring on us every day. One of them being the design filter – we intuitively gravitate towards aesthetic things. And the industry replied with a profound UX design scope. We began to appreciate our time more. The industry reacted harmoniously. But you can’t filter everything instinctively. At some point the technology has to take over it. And it did with the concept of Anticipatory Design. As defined by Aaron Shapiro from HUGE, Anticipatory design is a condition where decisions are made and executed on behalf of the user. This means the IA has enough information about you to know how you act and rid of a great amount of choices you have to make on daily basis. This involves the concept of Machine Learning (ML) and unlike a library of possible outcomes, programmed into the Ai, this is the ability of an AI to break human actions down and see how the effect of an action can determine the action in advance. This leads to the perception of future affecting the present.
Even though the idea behind this sort of integration is futuristic and surreal, it is closer than we might think. I believe whenever you hit the sweet spot in UX when everything falls into place and works the just the way it was supposed to work, it is partly anticipatory design, as the original user intention was predicted by a UX designer and executed with high level of accuracy, leaving no jointing edges.
Real-life Examples of Anticipatory Design
As soon as the concept of anticipatory design was established, and spread throughout different media, it was a matter of time that UX designers of different caliber started playing with it. Among the first aficionados were Facebook with its customized target audience feeds of suggested news, Instagram that analyzed users’ like to generate similar flow of content to spread involvement, and Amazon with an algorithm that by going through your purchase and interest history, time periods of search, query depth etc., provided a list custom catalogue of items, unique for every user.
Anticipatory Entertainment and Convenience
Media giants like Youtube, Netflix, and Pandora all analyze user behavior and use patterns, tags and categories to create personas and distribute content between them. Later on, the suggested feed became so sensitive and accurate that it started changing under the influence of user interaction as it happens.
In terms of device capabilities, anticipatory design has also been around for quite some time, to the point where we started taking it for granted. Apple iPhone’s iOS uses geolocation analysis to figure out where your home and work is, or other constantly visited places to understand which places are to be perceived as what when setting things up on calendar, alert notifications, and modes. Uber picks up your location and commute history to analyze your trips and book an Uber car back whenever you are going somewhere. Voice assistants like Siri, Alexa, Cortana have already penetrated into our lives to the point where we take them or granted. This takes our focus to the purely utilitarian approach in anticipatory design. Let’s look into it.
Anticipatory Internet of Things
Utility, convenience, and time-saving are the motives behind the concept of empowering everything with internet. This involves anticipatory design as the internet in things is just the means of executing utility. Basically IoT is the context for anticipatory design to be unleashed in. However, this becomes a matter of marketing and sales that things you don’t really need re presented as the new essentials. Take Nike HyperAdapt 1.0, the ‘self-lacing” shoe… really, Nike? Is this all you could come up with to charge us $720 for? Technology for the sake of technology is never a good idea and it strays us away from what really matters and that is taking a decision load off of us, allowing to concentrate on more important things in life.
Curiosity drives progress and always seeks the new edge the new mountain to climb, the new dragon to slay, and by doing that, occasionally we stumble upon true greatness, like Raspberry Pi 3 computer, medical monitors, smart light bulbs, Ericsson’s Smart Metering software, Google Nest, or LG Signature Refrigerator that allows you see inside the fridge without opening it, and so on.
Humans are the reproductive organs of technology.
― Kevin Kelly
Google’s Project Soli is what seems to be the beginning of a touchless device interaction. Can anticipation be transferred by gestures? Absolutely. High speed and accuracy motion tracking radar makes the universal conversational UI possible regardless of the language, culture, and other graphic expression media.
This technology has limitless application opportunities once the UX industry discovers and embraces gesture controls. Everything, from phones to cars and heavy machinery can integrate this technology to anticipate user’s intentions and react responsively. What’s even more impressive, Soli is not a box of sensors, cards, boards, and wires, it’s a chip that does not depend on environment conditions and materials. So sky is the limit for this technology. This, definitely is not a Utopia.
With all the opportunities, ideas, and expectations that anticipatory design concept brings to the table, a lot of assumptions about it have to be taken with a grain of salt. So let’s talk about the dark side of it.
The Hazards of Anticipatory Design
Internet of Things does look like a real-world facility for anticipatory design implementation, but the muddy waters of sales technologies, fashion, and viral commodities make it irrelevant in terms of anticipatory design purity and also presents some potential dangers among which are the following:
- Compromised privacy. The more duties you are willing to delegate to the AI, the deeper the level of personal integration is required. From a bank account to your dental record, everything automated runs on data, which today is the most valuable asset. Anticipatory design is the next level of protection. There can be no adaptation unless security is a guarantee, but truth is along with the light, grows the dark and scam artists will surely tackle this new way of living.
- Algorithm misinterpretations. Anticipatory design operation runs on processing algorithms that machine learning utilizes as part of its performance. By handling data, AI relies on numbers, as empathy is yet to be taught to it. As a passionate being, humans tend to imply different things in similar scenarios or perpetuate ambiguity and use context as a foundation. Will anticipatory design be real if a user will have to change their behavior in order to be decoded correctly? Perhaps, they may as a social norm. But never in real life.
- Cyclism. With every pathway of UX being anticipated and defined by algorithms, where do the new experiences come from? Is it an ever-expanding process? If it is, then how is resource distribution set up? Is there a risk of being caught in an experience bubble with limited number of options that rely not on the anticipation of the actual action, but the anticipation of the redefined action which deprives the idea of its essence.
- Legal issues. If information and knowledge are the main sources of incoming data, what are the legal stamps to be put on such information, who is responsible for it and how do you keep track on what is being generated throughout the media designed in an anticipatory logic. Is there any censorship and validation from the standpoint of intellectual rights protection and morality? These are the issues to be addressed.
Perhaps the most challenging and difficult part about anticipatory design is the actual technical fulfilment of it. For an idea as appealing and promising as anticipatory design, there seems to be no leading card to follow and more importantly, it breaches the essential principles of user interaction and object structuring.
The constituents of anticipatory mechanism are:
- Recurrence. In other words, the consequence of an action determines the action in advance. So the future state of things can affect the current state.
- Cause-and-effect relationships. It’s the natural life logic of constructing models to define the possible outcomes of events.
- Deviation recognition. This field operates beyond the existing models and allows the anticipation process to be continued even with new variables introduced to the game.
Modern technologies allow huge data to be processed, analyzed, and manipulated in any way, unlike the biologically-based concepts of anticipation that are yet to be discovered. What adds to the muddiness of the waters here is the fact that all three constituents can easily be perceived as UX parts. Automating and bringing them to life via technology is a hard task that requires deep consideration of Machine Learning (ML) capabilities of today. ML operates with data labels, features, and models via complex algorithms, concisely anticipatory development can use the same algorithms in conjunction with Deep Learning neural networks for data identification, recognition, and prediction.
As the researchers continue perpetuating the ideas of Cognitive Computing, the latest findings in humanizing machine choices can be instrumental in our movement towards genuine anticipatory design. Keeping an eye on these trends gives a good insight to what’s coming for us in future and whether the true anticipatory design is a reality within reach.