For the first time we will cohabit with an intelligent alien species, writes Illah Nourbakhsh
Industrial automation’s relentless march is the subject of scrutiny by economists eyeing its long-term ramifications on wealth and employment. But robots are also poised to march out of factories and into our everyday world, sharing our homes as well as our workplaces.
They will offer solace to the lonely, care for ailing elders, offer personalised tutoring for our children and even merge with our bodies as smart prostheses. We call these new machines social robots because they will interact with us using the same social norms and interaction patterns as our neighbours and friends. Should we celebrate or worry?
That depends on your degree of technology optimism — on whether you believe robots will free us from the toil of work and enable an age of abundance; or force a shift of capital ownership, concentrating information resources and power in the hands of a robot-owning elite?
Our future companions will possess four idiosyncratic social qualities that will offer a sense of where we are headed.
First, they are likely to change rapidly and without warning. We are growing used to cloud-connected devices that frequently shift functionality. Tesla has blazed this path with remote uploads to its intelligent electric vehicles; owners wake up to find new console graphics and that their car has new driving skills.
The forever-upgrading car may be exciting, but it is also unsettling because it threatens the control dynamic between owner and vehicle. Social robots will all be cloud-connected devices, just like Tesla cars. They will have skills, personalities and idiosyncrasies. And every aspect of these functions will be subject to change at a moment’s notice. Will we have a life of unease, always wondering just what our home care robot will be like tomorrow? Or will we learn to enjoy the daily surprises, as robots evolve on a timescale unlike that of any human companion?
Second, authenticity is overrated. In Black Mirror, the British television series about the imagined near-future, the episode Be Right Back imagines a widow who orders a robot patterned after her late husband to help soften the pain of her loss. The robot uses his digital footprints, from tweets to videos, to develop a model of his language patterns, memories and opinions.
At first, the grieving woman adores the robot because it triggers positive memories, but she gradually recognises it is a pale shadow, with scant connection to true human feeling and agency. Desperate to be rid of it, she takes it to the coast and commands it to jump off a cliff. As the robot prepares to jump, the widow takes offence that it does not even cry about its fate; at which point, taking her cue, it cries, theatrically emoting anguish and fear. While she knows it is acting to align with human expectation she can no longer bring herself to destroy it. Why? Because millions of years of evolution have fine-tuned human reactions to facial expressions. Robots with enough verisimilitude to imitate our emotions will be able to manipulate us, even when we are cognisant of the fiction we face.
Third, like it or not, you will become, in effect, a celebrity. You sit at a café table and have a friendly conversation with the stranger next to you. The situation is non-threatening because the two of you have common ground: you are strangers to one another. But what if you were Beyoncé; or Justin Bieber? When they sit down and fans approach, there is no such parity. Everyone has read all about their private lives — it is no wonder they ensconce themselves in the privacy of five-star hotels to escape constant awkward exposure to attention.
The coming social robots will be consummate tabloid gawkers, able to scan your face, identify you by reputation and search online for every last bit of information about you, all in a millisecond. In this scenario, we all become celebrities. If you prefer selective anonymity, prepare to duck away from unbalanced interactions by social creatures eager to make your acquaintance.
Finally, the social robots will be backed up by humans. This year Facebook announced the release of chatbots: intelligent agents that converse with customers, and designed so that we all make purchases, not by clicking, but by typing and speaking. Mark Zuckerberg, Facebook chief executive, also hinted at how these chatbots, imperfect as they may be, will be ready for primetime right away: whenever they have trouble conversing with us, they will draft the help of an army of standby humans.
Social robots will be modern centaurs: half-robot, half-human. It will be virtually impossible to distinguish when we are “talking” to a robot or to a living, breathing human in a call-centre. Our future robot sisters and brothers will always have human proxies ready to lend a hand. For us, this poses a deep question of identity: do we treat robots as sentient machines, subhumans or corporate agents? Will the way we behave with authentic humans change?
With the advent of social robots, for the first time we will share our planet with an intelligent alien species. Rather than descending from Alpha Centauri, the aliens will be the products of our own wealthiest corporations, all destined to monetise us with the same vigour as today’s internet. The social robots are coming, and they will unleash both celebration and anxiety.
The writer is professor of robotics at Carnegie Mellon University and author of ‘Robot Futures’