Chatbots Will Change Your Life But They Won’t Be Your Friend

Welcome to the post-appworld.

Chatbots Will Change Your Life But They Won’t Be Your Friend

Natalie Holmes

Antonio Manaligod/Dose

Welcome to the post-app world.

If you’ve noticed Siri getting quietly smarter, that’s just the start. Chatbots look set to replace apps, infiltrate our homes and predict our every desire. They might even free us from the trappings of the web.

A chatbot is an AI-driven computer interface that converses via text message or spoken language. They’ve been around for years now, but thanks to big developments in AI and machine learning, the latest models are a whole different breed.

Whether you’re excited or scared (and you should probably be a bit of both), here’s what to expect from the rise of the chatbots.

1. Chatbots will replace all those apps

Ever feel like we’ve reached peak app?

It’s not uncommon to use hundreds of apps and rack up endless login credentials — to the point at which most of us have lost track. “The Web is fragmented,” says Amandine Le Pape, who co-founded Matrix, a nonprofit that uses open source technology to build bridges between apps. “If one person prefers to use WhatsApp, they should be able to talk to someone who uses Skype.”

Matrix provides the foundation for Riot, a messaging service that does just that. Connecting an array of apps, platforms and online tools, it glues the shattered Web back together, shard by tiny shard.

So what does all this have to do with chatbots? In April 2016, Facebook opened up its chatbot development platform, effectively allowing companies to start building bots and communicate with their customers directly through Messenger. Other messaging platforms like Slack are also big on bots, and these days it’s possible to book a flight and order a taco via chatbot.

Many in the industry predict that one day we’ll conduct the bulk of our business this way. In that case, instead of using Messenger to talk to one person or brand, and Slack to speak with another, a service like Riot could become the only one we ever need to open.

In another, less rosy scenario, one major messenger service will emerge dominant and hold everyone at its mercy.

2. They’ll know you better than you know yourself

China’s WeChat is a messaging app that doubles as a platform and operating system. For its one billion registered users, it’s a one-stop portal for almost every online transaction, from ordering taxis to booking tickets to sending and receiving money. Chatbots are an essential part of WeChat, letting users interact with businesses and services at unprecedented scale. And, thanks to AI, people’s experience gets more relevant the longer they use it.

When you converse with a brand’s chatbot, every detail of that exchange is saved for future reference. With the amount of data available, not just from our conversations with chatbots but from practically everything we do online, brands will be able to make scarily accurate predictions about us. The AI will probably come to know you better than you know yourself. It will know when you’re running low on shampoo, and when you’re falling in love.

It’s convenient yet terrifying. And as objects other than your phone become smart, don’t be surprised to find yourself chatting away to your dishwasher or car. The chatbot is only the interface. Yet the means by which we interact with so-called “always on” technology is important, as it scurries into every crevice of our lives.

3. They’ll help companies creep deeper into your life

As we outsource more of our decision-making to chatbots, it’s easy to be distracted by novelty and forget about the corporate AI lurking behind them. As Will Oremus points out in Slate, when we ask a chatbot to order us pizza or search the web, we delegate out a crucial part of the process. It’s the bot that decides where we spend our money—and where we get our knowledge.

But, of course, it’s not the bot making that decision. It’s the AI — whose code was written by a company, with its own agenda and biases.

“The AI that allows companies to predict what we want is both good and bad,” says Joanna Bryson, an expert in AI ethics, and fellow and affiliate at Princeton’s Center for Information Technology Policy. “We’re getting better at knowing what people want — which means we can help them faster, but also that we can essentially control them. If you can predict what someone’s going to do, you can maneuver them.”

“The fact that there’s a bot involved perhaps makes people trust it more,” says Bryson. “People think, ‘Oh, it’s like a dog.’ It’s not like a dog; it’s like a microphone and a camera. If you knew there was a mic and a camera in the house, you’d probably freak out. But many people don’t — they happily buy their kids a doll that’s uploading conversations and then talking back to them.”

4. They’ll be friendly but not your friend

The jury’s still out on exactly what our conversations with chatbots will look like. Having built himself an experimental home AI named Jarvis, Mark Zuckerberg was surprised to find that, given the choice, he preferred chatting to it by text rather than voice. He concludes that though text will likely be the predominant mode of communication, voice will also be important.

Whether we’re chatting to them vocally or via text, the question remains: Will we ever confuse a chatbot for a human? In some cases, like when you’re chatting to your fridge, that seems unlikely; less so when, say, seeking customer support via messenger.

Bryson suspects we’ll develop an intuition to tell the difference: “When ‘King Kong’ first came out, people were screaming and fainting in the cinema. Over time, as a culture, we got used to it. Maybe that will also happen with chatbots: We’ll recognize the subtle indicators that tell us we’re not talking to a human.”

But maybe it won’t come to it. Bryson thinks we should be wary of creating humanlike bots. “We say we want to be friends with robots, but that means we want a friend whom we own,” she says. In other words, it’s precisely because we want to own and use robots that we should never think of them as human.

For Le Pape and her work with Riot, ethics are also central. “Our chatbots begin by asking the user if they are happy for their information to be gathered and used,” she explains. Such a conversation might represent the chatbot version of smalltalk, or at least an icebreaker.

5. They might even free you from your internet addiction

Social media feels powerful and all-consuming. “Information overload” is an overused term, and for good reason. Yet on a global scale, messaging apps are set to overtake social networks in terms of user numbers. With this shift comes a huge opportunity to overhaul some of our bad habits.

“We want to find a way to get people out of the echo chamber created by social media,” says Le Pape. At Riot, one aim is to provide relevant recommendations alongside windows into other perspectives. “It’s a challenge that needs to be addressed now. In many cases, we are locked in a world that’s virtually unrecognizable from our neighbors’.”

Instead of binding us to the web, Le Pape sees AI as our liberator. “We want to apply it to helping people disconnect,” she says. “The technology should be providing what you need, rather than bringing in more and more irrelevant information.” The next step is finding new business models that don’t focus solely on hogging our attention.

As an interface, chatbots hold the potential to help us break our internet addiction by learning not only what we want but when we want it. By replacing apps with chatbots, we can be constantly connected in a way that helps us, paradoxically, to make the most of our time in the real world. In that sense, we’re approaching what you might call the ambient internet, an omnipresent force we interact with in ever more subtle ways.

In the meantime, as the struggle for our attention and data continues, ethical and practical challenges remain. “There’s not one single answer, but a lot is happening in the right direction,” concludes Le Pape. “It’s the beginning of an interesting era.”