A look at the convergence of data and design and where machine learning is headed.
Gmail, the world’s most popular email service, will now respond to your emails for you. Using a form of artificial intelligence called deep learning, the new system, Smart Reply, analyzes emails and suggests a few short responses tailored to the message’s content and tone. This type of deep learning system is based on what’s called a neural network, an extensive network of machines inspired by the web of neurons in a human brain. Smart Reply “learns” to respond to email by analyzing a vast amount of correspondence across Gmail.
The trend represents what may be the next crucial shift in UX design, what Aaron Shapiro, CEO of digital agency Huge, calls anticipatory design. Shapiro points to the fact with the incredible conveniences technology has introduced come an equal number of decisions. We are inundated with choice. His example: buying a Monopoly board game for his son means choosing between the 2,767 versions available online.
Essentially a convergence of design and data science, the goal of anticipatory design is to minimize decision-making using data, prior behaviors, and business logic to create highly tailored and seamless user experiences. Based on the assumption that the more decisions we have to make, the lesser ability we have to make effective ones, anticipatory design empathizes with burnt out users struggling with an overabundance of choice, who crave less freedom and more efficiency.
Shapiro sees UX professionals working side by side with data scientists to help brands transform products into services.
Take Amazon’s Tide button that lets consumers order more detergent simply by pressing a button on their washing machine. Within an anticipatory approach, a scale in the machine could sense the weight of remaining detergent and automatically order more, eliminating the need to push a button or even consider your dwindling supply.
In this emerging model, Shapiro sees UX professionals working side by side with data scientists to help brands transform products into services. Of course, creating this “ecosystem where a decision is never made” means relinquishing personal data and connecting disparate systems, and even if initially the user may be asked for feedback the goal is eventually the system will complete jobs without input.
Fewer Choices, More Questions
It raises some interesting questions in relation to privacy, security, dependence on technology, and the role that design should or should not play in our lives and in society. “With finite amounts of data, you can create a rudimentary understanding of the world,” says Andrew Ng, chief scientist at Baidu, the Chinese Internet giant leading the deep learning movement. As artificial intelligence and machine learning become ubiquitous, we’ll have to consider what it means as more and more of our daily decisions are made by entities with no moral obligations and a limited understanding of the world. When these technologies progress, that conversation gets further complicated.
It raises some interesting questions in relation to privacy, security, dependence on technology, and the role that design should or should not play in our lives and in society.
Google Now claims its services are offered so you can “focus on what matters”. Decision fatigue is of course only a problem suffered by affluent communities, and as our daily lives become administered by machine, how should we spend this luxury of time? What is it that matters? With such extreme disparity of privilege in the world, perhaps we might want to consider how we can turn our attention to solving issues beyond how efficiently our groceries can get to our front door.
Got you hooked? Here’s more:
A NYTimes look at what affects decision-making
The Paradox of Choice
Psychologist Barry Schwartz takes aim at freedom of choice
How Twitter’s new service Moments aims to simplify news