Last week at Microsoft's Build conference, CEO Satya Nadella said that the future of the company was "conversation as platform." In other words, less Windows and Office, and more Cortana and Tay—conversational interfaces that can understand the natural language of human users.
If Nadella thought he was expressing some unique vision of the future, though, he was fooling himself. The idea of conversational UI has quickly colonized nearly every corner of Silicon Valley over the past year. Now seems like a good time to ask: What is a conversational interface?
What's a conversational interface?
A conversational interface is any UI that mimics chatting with a real human. The idea here is that instead of communicating with a computer on its own inhuman terms—by clicking on icons and entering syntax-specific commands—you interact with it on yours, by just telling it what to do.
What kind of conversational interfaces are there?
Right now, there are two basic types of conversational interfaces. There are voice assistants, which you talk to, and there are chatbots, which you type to. I'd also probably distinguish a third "fake" kind of conversational interface: the pseudo-chatbot, which mimics a chatbot in appearance but is really a traditional point-and-click GUI. Microsoft Clippy and Quartz's weird text-messaging news app are good examples of pseudo-chatbots—they borrow the visuals of a chatbot but don't actually allow you to converse beyond their canned responses.
What are some good examples of conversational interfaces?
On the voice assistant front, pretty much every major tech company in the mobile space has its own at this point. Apple has Siri, Google has OK Google, Amazon has Echo, Microsoft has Cortana, and so on. All of these voice assistants allow you to do things like play music, do a Wikipedia search, call someone, set a timer, and more—just by speaking. There are even more chatbot examples, though. Facebook has M, a human-assisted chatbot who lives within Messenger and can do anything for you from book a dinner reservation to buy you a car. Slack's Slackbot is another great example: It's a chatbot that poses as a real user in any Slack team, and is used for everything from onboarding new users to working as a notepad. There are countless other chatbots for Slack, too, including Howdy, a Slackbot, that does everything from schedule meetings to take lunch orders for your office.
Why are companies so interested in conversational interfaces?
Loads of reasons! For one, conversational interfaces are truly cross-platform. They work well everywhere, including smartphones, desktops, smartwatches, and even devices without screens at all, like the Amazon Echo. They can integrate with services like Twitter, Facebook, or Snapchat, or run just in a text message window. As Amazon, Google, and Apple have all shown, you can make a conversational interfaces better over time without pushing out regular updates. Conversational interfaces also mean that every single function in an app or service no longer needs to be buried in a menu, or represented by an icon.
But the biggest reason everyone is excited about conversational interfaces is because they have the potential to eliminate the underlying friction that makes it hard for every person to get things done on a computer.
What do you mean by "friction?"
Historically, computers and humans have essentially spoken different languages, with graphical interfaces as the translator. Computers are real sticklers for syntax—machine-assembly pedants who basically fall over if you tell them "hi" when they're expecting "hello." So for most of computer history, we've communicated with computers essentially through Rosetta stones: We point at a symbol representing what we want a computer to do, and then it does it. For example, clicking an icon to open an app. With conversational interfaces, computers and humans can finally speak the same language without a Rosetta stone in between.
Are conversational interfaces new?
Nope, they're pretty old, at least conceptually. HAL-9000 in Stanley Kubrick's 2001: A Space Odyssey and W.O.P.R. in WarGames are both examples of conversational interfaces in science-fiction movies.
If the concept is so old, why are we suddenly hearing about conversational interfaces now?
Because technology has finally gotten good enough to make them practical. Over the last few years, advances in voice processing have made it easier than ever before for computers to understand natural language, while the rise of smartphones has put an Internet-connected microphone in every pocket. Meanwhile, AI projects like Google's Knowledge Graph and Wolfram Alpha have made computers better than ever at understanding more than just syntax, but what we actually mean.
Why is the rise of conversational interfaces important?
The dream of conversational interfaces is that they will finally allow humans to talk to computers in a way that puts the onus on the software, not the user, to figure out how to get things done. That's not only the way things should be; it has the potential to totally change the way we use computers going forward.