We talk to our phones (thank you, Siri), so why can’t our tissue boxes respond appropriately when we sneeze?
People who need people may be the luckiest people in the world, but most of us may not need that much human contact after all. As smart technology and sensors expand our ability to communicate, not just with each other but with inanimate objects as well, human-object interactions are becoming far more common — which naturally makes them a ripe subject for research.
How comfortable are we with non-human devices that ask, probe and provide us with all kinds of information? They’re far from insensate, and that’s the point. Would responsive, but inanimate objects like talking tissue boxes be too creepy for us to accept? To find out, researchers from Penn State tested peoples’ reactions to smart tissue boxes that reacted to sneezes and learned that most people viewed the chatty boxes favorably. The investigators videotaped 63 participants who thought they were involved in a study of their cognitive abilities. They were split into three groups — one stayed in a room with a laboratory worker and a talking box of tissues. When the lab worker launched a computer-operated sneeze, the tissue box either said, “Bless you,” “Here, take a tissue,” or “Take care.” Another group went to a room with a life-like robot that reacted in the same way to the sneeze. The final group only had a researcher who responded to the sneeze.
Afterward, the researchers watched the videos and asked the participants fill out questionnaires about the lab environment and the presence of the smart objects. The participants reported that they found the talking tissue box just as human and autonomous as the human-like robot and responded strongly to the voice.
“We were pleasantly surprised. We were expecting reactions that might be somewhat negative. We thought people would be somewhat quizzical or thinking it’s weird. Mostly we got, ‘this is cool’ kind of comments,” says study author S. Shyam Sundar, a distinguished professor of communications and the co-director of the Media Effects Research Laboratory.
According to Sundar, the mandate in the design field is generally to make objects more human-like. However, the results showed that people might build positive social interactions with objects as long as certain well-known social cues and ways of interacting are used. People expect, for example, to hear a ‘Bless you’ or be offered a tissue after sneezing. “Just using natural language goes a long way in building social human-object interaction. You don’t really need to work very hard to make it look human,” says Sundar.
His findings are an increasingly important part of what’s known as the “Internet of Things,” or the idea that little objects are always talking to the Internet. Shipping and tracking the location and activity of a package is a good example. “There are automated systems that track movement of objects, and these automated systems are directly connected to the Internet, and that is why it is called Internet of Things,” says Sundar. “There are several applications that technologists forsee, including embedding sensors in animals to figure out where they go throughout the day to feed and prey upon other animals. A lot of information can be obtained.”
Understanding how people interact with objects could lead to interesting applications of such sensing devices, he says. These technologies could answer very specific questions, like understanding the movement and activity of animals, for example. Already, researchers at the University of Waterloo have equipped cows with RFID tags that track their activity and alert a robotic milking pen when the animals are ready to be milked. Such efficiencies could also spill over into daily life — in the form of a refrigerator that tells you when you’re almost out of orange juice, for instance.
“There are going to be more and more objects like this around us, as part of this era of ubiquitous computing. As a communications scholar, I see there is going to be a real need for interaction between humans and these objects. So we are looking ahead, and trying to come up with some of the ways we can outfit these objects to start communicating with humans,” says Sundar.
Of course, some people may be more open to smart objects than others, and these individuals tend to be more parasocial than their peers. They’re more likely to feel connections to people they don’t really know, such as an evening news anchor. Studies showed that those who scored high on parasocial skills were more welcoming toward talking objects. “People who are already into mediated technological representations are more likely to accept it compared to people who are less technologically savvy,” Sundar says.
As the researchers continue to investigate ways of making smart objects more engaging, keeping people’s interest piqued over the long term becomes an issue. Sundar says the the novelty of talking objects may wear off, and become annoying instead. Or, if the objects are meant to alert or warn people of potential dangers, they may be ignored if they are repeated too many times when hazards don’t occur. “On the one hand, we do like machines to be predictable. In fact, we like humans to be predictable. We have very perfunctory things to say every day to our colleagues. Only when they don’t say the expected response do we know that something is wrong,” says Sundar. “On the other hand we may need to design these objects with a more personalized repertoire so it can be more contextually aware.”
There’s no doubt that objects will become smarter and sensors that talk to us more ubiquitous, but as with the friends we have, it’s up to us to decide how much we can tolerate being in their company.
The researchers presented their findings at the 2013 Annual Conference on Human Factors in Computing Systems in Paris.