The most memorable scene in the 1986 movie, Star Trek 4: The Voyage Home involved a time-traveling Scotty trying to use a computer from the 1980s. He walks up to a Macintosh Plus and says: "Computer!" When the computer doesn't respond, it occurs to Dr. McCoy that, because this is a primitive computer from the past, perhaps it needs a close-up microphone. So he hands the mouse to Scotty, who tries his voice command into the mouse.
The scene is as prescient as it is funny.
What Star Trek always got right was that the user interface of the future was conversational.
A conversational UI can take place as a back-and-forth text chat, email or spoken conversation. The difference between text bots and the virtual assistants you talk to is slight. By simply adding off-the-shelf speech recognition on one end and text-to-speech on the other, you can turn any text bot into a speech assistant.
In fact, bots are highly portable, and the companies that make them don't care where they show up.
Dennis Mortensen, the CEO and founder of a New York startup called x.ai, which makes the Amy virtual assistant for scheduling meetings, told me that the Amy virtual assistant could in the future be made available via Amazon's Echo, Apple's Siri or through some other channel. It doesn't matter. While today the Amy is an email-based virtual assistant that lives in the cloud, it could in the future be conversed with by phone, text, in a social network or any other space where a conversation could take place.
Mortensen also said recently that he believes bots would soon replace apps.
Rise of the conversational UI
The idea that chat bots and virtual assistants that we can talk to would replace apps is as alien to us now as today's computing scene would have appeared 10 years ago.
For example, if I told you in 2006 that in 10 years the mobile web would be faster and more feature rich than the version of the web you get on the desktop, it would have made no sense. (Refresher: This is what phones were like in 2006.) If I told you in 2006 that Apple would be the world's leading phone maker, most valuable company and was working on a car; that Google would be delivering Internet access via balloon; that Facebook's CEO was the world's fourth richest person; that stringing together little cartoon icons would become a major form of social interaction; or that everyone can stream live video globally but it's too banal for most people to bother with -- you would have thought I was nuts.
Likewise, a few years from now, we'll use computers and the Internet in ways that make no sense today. The dominance of conversational UIs sounds less appealing than how we use computers today. But that's because we can't picture what that will be like.
Lucas Ives, who works as head of conversation engineering at ToyTalk (which makes the conversation engine for Hello Barbie), told me that "in five or 10 years you'll be walking through your kitchen, and your refrigerator will say: 'Your milk is going to go bad in three days, do you want me to order some more for you?' "
Ives' example is a perfect illustration of three ways the conversational UI will change our lives. First, the interface is a conversation. Second, the conversation is with the refrigerator -- the Internet of Things will turn everything into an Internet-connected computer. And third, the refrigerator can start the conversation. Pre-emptive interaction today is a novelty, found mainly in Google Now. In the future, many objects, devices and apps will initiate conversations with us.
In fact, the conversational UI trend has already begun.
The conversational UI has got everybody talking
Quartz this month introduced an app that gives you the news in a conversational UI. It simply chats with you, as if you were getting the news from a friend via text. It tells you a little bit about a new story, then if you ask to hear more, it will go into more detail, complete with photos, links and, eventually, ads.
While the possible user input is narrow (ultimately you tell it that you want to hear more about the current story or you want to move on to the next story), the experience is just like texting with a friend, where the subject happens to be the news and the friend happens to be a fast-typing journalist who's banging out news stories just for you.
At Mobile World Congress last week, Sony unveiled a range of products, including something called Xperia Ear, which is an "intelligent earbud," and Xperia Agent, an Amazon Echo-like virtual assistant appliance. In both cases, it shows that Sony is preparing for the conversational UI future.
Sony's Xperia Ear video is a perfect illustration of the subtle shift to conversational interfaces. In the video, the users are doing normal things like texting, making calls and getting directions. But instead of doing these things directly, they're asking a virtual assistant to do it. And the assistant responds with the information.
One of the surprise darlings of the Los Angeles Auto Show in November was a Silicon Valley startup called Capio, which makes a conversational UI on a chip for cars (and other appliances). If you visit their website and watch the short video, you'll get a sense of what the future conversational UI of the future will be like inside a car.
These new products and services are part of a much larger trend toward the conversational UI.
What parents got wrong about Barbie
Hello Barbie works like Siri. You talk to the doll, the doll talks back. And like Siri, children are actually talking to software in a remote data center. Their voices are recorded. The recording is sent and processed. A Barbie-like response is constructed and sent back down through the Internet and home Wi-Fi network to the doll, which replies.
The initial press on this product, which shipped in November, skewed negative. As an Internet-of-Things appliance connected to a home Wi-Fi network, Hello Barbie was called insecure. Others said that a product that records the voices of children is creepy.
The critics are completely wrong on both counts. Hello Barbie should serve as a model for how IoT devices should handle security.
Late last year, a security company called Blue Box discovered potential vulnerabilities in Hello Barbie and the companion app, but then later admitted that due to ToyTalk's "fast response time, a number of the issues have already been resolved." No doll has been hacked. And the company launched a security bug bounty program, paying security researchers for finding any future problems.
Few IoT device makers are acting this responsibly on security.
Regarding the discomfort people feel about children interacting with a virtual assistant bot housed on a remote server, I'll just come right out and say it: Get used to it, people. This is the future.
Talking to a virtual assistant and the virtual assistant talking back is what using a computer will be in just a few years.
It's also worth pointing out that actively engaging a virtual personality in child-directed conversation is probably better for kids than passively watching TV for hours, or using any screen-related technology, for that matter.
So if you want to prepare your daughter for the future of technology, get her a Hello Barbie. Because Barbie works today like everything will work tomorrow.