The era of human-centered design is coming to an end. That’s because the next user you design for won’t be human at all. It will be a human-machine hybrid built to do more than any person or computer could accomplish alone. Welcome to the age of Centaur Design.
The centaur refers to the Greek myth of a half horse and half man. More recently, it has been used to describe human and artificial intelligence working together toward a common goal. This is the inevitable march of technology. Computers are faster and smarter than ever, and they will only improve, but they lack key cognitive skills like common sense and the ability to draw on a diverse set of experiences–things people do well. People and computers can be more effective working in tandem.
Chess offers a compelling example. In 1997, IBMs Deep Blue supercomputer defeated world chess champion Garry Kasparov. It was a stunning upset. If a computer could defeat one of the best players on earth, what was the point of people playing anymore? But Kasparov saw an opportunity. The following year, he competed in the first public “advanced chess” game in which both players used computers to augment their abilities. Instead of computers taking over the game, they could enhance it.
Today, advanced chess games are common at both the grandmaster and amateur level, and human-computer hybrids are often more effective than either people or computers alone. Computers are fast. Human players are creative. Togther, these networked “centaurs” can even beat supercomputers. The real advantage is not the power of the individual computer; it’s the diversity and variety in the network.
Toward A New Design Paradigm
Such a network requires a new design approach. Today, most tech products are designed to optimize for attention, which requires elevating the individual user above all else. Human-centered design has been great at streamlining users’ everyday activities, like navigating a city (Google Maps), socializing (Snap, Facebook), and shopping (Amazon). It hasn’t been great at understanding how users interact with, and are influenced by, larger political, economic, cultural, and technological systems–sometimes to disastrous effect. Design’s next big paradigm will have to account for these systems. Put another way: What happens when computers stop working for us, and start working with us? What does that look like?
There aren’t many examples yet. But there is one. And it’s probably not from the company you expect.
Waze: A Case Study
Turn-by-turn GPS navigation has been available since 1995. Today, nearly everyone has access to it through their smartphone thanks to Google Maps. That app has undoubtedly changed the way people get around by placing users at the center of maps and allowing them to quickly find the most efficient route from one place to the next. But it’s still just an optimization–a better, faster, more efficient version of what people were already perfectly capable of with paper maps.
Waze is different. When you first start using it, it feels like any other navigation app. You punch in your destination, and it tells you how to get there from where you are. But if you keep using it, you notice that there’s something more going on under the hood.
Unlike Google Maps, Waze doesn’t give you much control unless you ask for it. It just starts telling you where to go. As people have begun to notice, the routes it uses can be a little strange, often taking drivers down side streets or surface roads instead of using thoroughfares or highways.
That’s because Waze has built a machine learning system that’s monitoring every driver connected to the network in real time. Tell Waze where you want to go, and it sends you on a route optimized not just for you, but for the entire network of drivers using Waze. So sometimes Waze will send you on a route that’s a few minutes slower than the fastest possible one available, because sending you on the fastest route would have a negative impact on the network as a whole. Waze makes a decision that’s bad for the individual but good for the network.
How does it get away with this? Users expect technology to shave time off a task, not add to it. They also expect to have agency over their digital experiences (or at least the appearance of it). Waze’s ingenious design move is to give users a new job. Waze automatically tracks drivers when they install the app, but it also lets them report accidents, road blocks, and other conditions that inform real-time route recommendations. The computer takes care of directions, the human drivers provide context. That gives users a sense of control–they are active participants in making the whole system better. In this sense, Waze is one of the first companies to build a centaur for everyday use.
You can imagine how difficult it will be for companies steeped in the tradition of human-centered design to think in terms of centaurs. Building a system that users trust, despite knowing that they don’t have much agency within it, is a tremendous design challenge. Done poorly, it can make people feel powerless and disconnected from the very system that is meant to serve them. Done well, it can enable new user experiences and benefit the greater good.
More than 50 years ago, the legendary computer scientist Douglas C. Engelbart (creator of the mouse, the graphical user interface, hypertext, and much more) wrote a landmark paper called “Augmenting Human Intellect.” In it, he argued that because computers can represent ideas symbolically in the same way that our brains do, they are the first machines capable of truly enhancing human intelligence, and maybe even taking on a life of their own.
Engelbart was right, but he was a few decades ahead of his time. The technologies he and his contemporaries dreamed up proved too demanding for the relatively slow computers of that era. Not so anymore. We have the computing power. Now, it’s time for the design and engineering to catch up. We have to remove ourselves from the center of our tech products and build for our creations, too.
Matthew Milan is the CEO and cofounder of Normative, an innovation and design firm headquartered in Toronto.