Dean's Notes: Interaction

I am not a gadget geek and Thanksgiving is one my favorite holidays. Except for the one frosty morning when I stood in line with my daughter outside a Target, hoping to score a Nintendo Wii, I have avoided Black Fridays. Not this year. For some inexplicable reason, I found myself at Macy’s a little after their doors opened on Thanksgiving Day.

I was in the market for a robot, a Shark Ion that would vacuum floors, and was enticed by the $50 discount. I brought it home, powered it and downloaded the app. Before long, I had named it Sharkey. Then I was talking to it and rooting for it to learn and successfully negotiate the layout of my living room. To my puzzlement, I was treating my household appliance as I would a new pet poodle.

The instinct to anthropomorphize inanimate moving objects and develop social bonds with computers is natural. In fact, I have studied it and written papers on it in the past, but to observe its grip on me gave me fresh insight.

Two professors of communication at Stanford, Cliff Nass and Byron Reeves, were among the first to identify our propensity to treat our interactions with computers as real. From their research on Clippy (remember, the overeager animated paperclip help agent that came with Windows), they found that we are naturally inclined to respond to computers as social actors.

In their book, Media Equation, Reeves and Nass attribute this propensity to evolutionary biology. Our brains are old and a product of thousands of years of evolution, whereas our media are new. Even the oldest medium, say stone tablets and papyrus scrolls, date back only a few thousand years.

It is no surprise then that our old brains are not hardwired to process new media. When our neurons encounter a new media experience, such as human-robot interaction, they translate that experience through the default template of human-human interaction. Studies confirm that our response to computers and robots is remarkably similar to our response to humans.

This holiday season may be the inflection point for human-robot interaction. New options have flooded the market and if you have been nice this year, you may find an Alexa or Google Home device under your tree. Much like Siri, the new chatbots are chatty and rely on natural language, which easily tricks our neurons into mistaking synthetic interactions for real. Further, the new generation of robots are at your beck and call, eager to turn on the lights, set the thermostat and close the garage door. It is magic.

My favorite is asking Alexa to find movies on Netflix or Google to find TED talks. I also request Google to give me a rundown of my day in the morning and set my alarm at night. I have to confess that I am eagerly awaiting the introduction of robotic lawnmowers and snow throwers. I will have to come up with more creative names for them than Grasshopper or Snowman.

That our neurons are easily tricked is a blessing and a curse. It is a blessing when we benefit from the advances in AI and natural language processing that make human-machine communication natural. It comes with a catch. We can be lulled into believing that these interactions can supplant the human-human interaction.

Recently, my friend’s wife complained that he spends too much time talking to Google, while ignoring her conversations. Therein lies the caution. I could be temporarily duped into believing that Sharkey is my new bestie, though no robot can ever take the place of Timmy or Mally, mutts whose memories have served me well over the years.

This holiday season, I wish you joyful interactions powered by AI that fill your life with information, entertainment and assistance. At the same time, I wish you interactions with family and friends that bring peace, joy, love and harmony. From all of us at ComArtSci, Happy Holidays!