Even If AI Can Cure Loneliness — Should It?

MIT Sloan Management Review

Many experts believe augmentation and automation are the shining stars of business uses of AI; their promise of greater productivity has lit up the executive imagination.

In their shadow, however, a growing number of AI applications and devices are helping humans satisfy a basic need to connect with others. In particular, markets are slowly forming around artificially intelligent, emotionally attuned, responsive robots that people can relate to as companions. Certainly, if army veterans can form bonds with drones, many people can form emotional connections with their bots, either in lieu of human alternatives or in addition to them.

Where there is an unfilled human need, there is a business opportunity. Large-scale social problems, like the global loneliness epidemic, are driving demand for robot companions. The AARP estimates that one out of three U.S. adults older than age 45 suffer from chronic loneliness. In Britain, researchers estimate that 9 million adults are often or always lonely; one out of three adults over 75 say their feelings of loneliness are out of control. In January 2018, Britain named a loneliness minister after recognizing its serious, multibillion-dollar toll on the U.K. economy. Loneliness is associated with premature death, productivity loss, and various health costs. More than a dozen startups are developing robot home companions. While some are struggling, there is little question that demand for these products exists, and is likely to grow stronger with an aging population.

Eldercare “is rapidly becoming one of the most daunting health care challenges of our day,” says former Harvard Medical School professor William A. Haseltine. An NIH-funded Census Bureau report estimates that by 2050, nearly 17% of the world’s population, or 1.6 billion people, will be at least age 65, double the percentage of today. Many companies are developing robots to provide services to that growing cohort, such as making schedule recommendations, offering medicine reminders, and coordinating care. Although most of these aren’t designed specifically with loneliness in mind, they do provide companionship, which many of the elderly desperately need.

So technology isn’t just a cause for the loneliness epidemic, as many suggest. It’s also possibly a solution. The early popularity of social robots suggests that there’s quite a bit of pent-up demand for nonpharmaceutical alternatives — not only to address loneliness once it sets in but also to stave it off in the first place.

Of course, the human need for connection is physical as well as emotional. Market demand for social robots that satisfy sexual appetites is also on the rise. You thought Tinder was bad: Sex with your own robot is now an option, and thousands of people are taking advantage. Entrepreneurs are combining advances in materials science, robotics, sensor technology, and natural language processing to create anatomical simulacra that provide physical pleasure. Granted, this approach to satisfying a basic need may not be long-lasting for a given person. Some may doubt whether it constitutes relating at all. Even so, with no regulatory structures around sexbots — and none in the making — the effects of this growth industry will inevitably rejigger social norms. Noel Sharkey, from the Foundation for Responsible Robotics, says: “We’re just doing all this stuff with machines because we can, and not really thinking how this could change humanity completely.” The makers of some robotic sex dolls are looking to expand sales to eldercare facilities; sex doll brothels already exist in South Korea, Japan, and Spain.

The market opportunities for social AI extend beyond caring for the lonely and the elderly: Collaborative robots (aka “cobots”) provide a substitute for traditional social connection in the workplace (for example, SoftBank’s Pepper is used as a customer service aide in hotels). AI-driven online games like Fortnite or dating sites like Match.com rely on algorithms to help like-minded people find and connect with one another so that they don’t have to do the work themselves, which can be excruciating for the introverted.

Understanding social AI as a market maker is critical for company strategists, as well as product developers. But it’s equally important to recognize that, for better and worse, the social arrangements we take for granted today are also at stake: The roles people play in their own and others’ lives are increasingly mediated by technological third-party actors with greater emotional, linguistic, and social sophistication. The implications are both daunting and clear: If part of what makes us human is to connect emotionally with others, and technology increasingly plays the role of emotional connector, what it means to be distinctively “human” becomes a much more complicated question.

The societal ramifications of social AI products are not entirely for regulators to address. Makers of social AI products and services need to tread carefully and really consider the difference between “Can we make and sell this?” and “Should we make and sell this?” The question is “Will executives looking to address real and pressing social ills with social AI work effectively with regulators to address unintended consequences of their business efforts?”

No comments