Robots Imitating People: Funny? Scary?

(Last Updated On: January 22, 2017)
robots imitating people

KOBIAN robot, Takanishi Laboratory, Waseda University

Science Fact: Robots imitating people is a respected line of research in Japan, and the Japanese lead the world in this area.

Getting a robot to do anything is challenging enough, which helps explain why there are many university programs dealing with the design, construction and programming of robots (sample lists are given here and here). It’s even possible to get a PhD in Robotics at Carnegic Mellon and other universities.

It’s even more difficult to build useful functions into a package approximating the human shape as with a humanoid robot. However, perhaps it’s most difficult of all to create a robot with recognizable human-like behavior.

A previous blog (Robot Life & the Uncanny Valley) addressed how robots imitating people encounter an “uncanny valley” in which they appear creepy if they are almost life-like. This week brought two news stories that describe quite different approaches to robots imitiating people.

The robot illustrated above is KOBIAN, created by Prof Atsuo Takanishi and his colleagues at Waseda University in Tokyo. Prof Takanishi has been programming KOBIAN to be a comedian! This seems amazing on several counts: Why would you do such a thing? And what might you learn from this work?

I think it’s more interesting to leave those questions for your speculation. Instead of addressing them, I’d rather encourage you to watch the fascinating video in the news article. Take a look, and see whether you find this machine to be funny, in any sense! What’s also interesting is an article with detailed photos showing how the robot can move and tilt its eyebrows, change its mouth shape and even form asymmetric expressions on its face.

Another current instance of robots imitating people, pointed out to me by Charles South, comes from Prof Hiroshi Ishiguro at Osaka University. He has just put two human-like robots on display at Miraikan, the National Museum of Emerging Science and Innovation, in Tokyo. He plans to study how visitors to the museum interact with the robots, to “harmonize humans with [the] information-environment”, studying “what is the essence of human beings?

Prof Ishiguro has been prolific in robotic design and engineering. He’s also the creator of the Hugvie, a pillow that looks vaguely like a child, into which you insert your cellphone. Although some journalists sniff that Hugvie is “creepy”, Ishiguro and his colleagues have published research that shows a reduction in cortisol (a stress indicator) when hugging this device compared with having a normal phone conversation.

Science Speculation: It’s interesting to wonder why Japan has chosen such a difficult phase of robotics, robots imitating people. Several reasons come to mind:
– Commercial versus Military. It’s generally accepted that Japan leads the world in companion robots. Similarly, the US leads in military robots (example: the BigDog and its later versions, intended to serve as a pack mule for soldiers). Given the roles of our countries in the modern world, this division of labor is not surprising. Since World War II Japan has been anti-militaristic but very successful commercially, while the US has taken the burden of world’s peacekeeper.
– Aging Population. Japan has a higher proportion of older people than any other country. For this reason, there is social concern with the problems of caring for the elderly. Robotic devices to serve as companions and to assist with Activities of Daily Living are seen as a vehicle for science and technology to serve useful social purposes.
– Business Opportunity. Healthcare is a large and growing market, and robotics companies see a chance to share in a high-margin business. A market study of “service robotics” forecasts a huge demand for servant robots, especially to help the elderly.
– Cultural Acceptance. Japanese people are much more accepting of robotics than Americans. Journalist Christopher Mims traces this cultural difference to religion. Japan’s Shinto has a strong component of Animism, in which inanimate objects may have helpful or supportive spirits. America’s Judeo-Christian monotheism, in contrast, tends to view animated objects as uncontrollable if not outright evil, as in the Golem of folklore and Mary Shelley’s Frankenstein. For a recent take on this, see The Week’s article on “Hollywood technophobia.” However, despite the cultural divide, there’s evidence that our feelings about robots become more positive when we experience them as being helpful.

A measure of progress in robots imitating people might be their ability to make us laugh. Do you find the video of KOBIAN funny, in any sense of the word?

Image Credit: Photo from Takanishi Laboratory, Waseda University, Tokyo. Image reproduction claimed as Fair Use of copyrighted material because:
1. Proper credit is given, with links to original sources
2. It is a low resolution reproduction of an image that is publicly available in media and international conference proceedings
3. The image cannot be used to replace the original photograph or the work depicted, nor can it be used to reverse engineer the work
4. The image is necessary to clearly identify the research that is being discussed
5. The usage is noncommercial, for the purpose of educating the public on scientific advancements
6. No uncopyrighted image is known to be available that would serve the same purposes

Comments

Robots Imitating People: Funny? Scary? — 2 Comments

  1. hmmm … my response isn’t among the top six … though neither is my disposition. Before hand, I was open, ready-minded. By the end I was deep into considering the potential effectiveness of what was strikingly and obviously absent as to raise no question in my mind over why it didn’t elicit even a cracked smile. There i no registration of a human grievance within the content or medium … though it is not difficult to see how one could be evoked by subjective exposure to the behaviors of the would-be Kotomatomedian [to coin a term off the cuff, by just mashing up Komedian and Automaton actually presents as something more funny in imagining, than the attempt at realization produces].

    And that is the missing point – the intersection of a shared experience, creating an affirmative sense response which show by smiling and laughter. We within english-speaking cultures are quite apt to make reference to the human “sense of humor,” though without really giving consideration to it being in fact what those words describe. One reason for this is that it is far afield of the mundanely (and very small-mindedly) recognized mere five senses … especially since the short-list omits the mind from mention and in consequence occludes any senses which sensibly should be attributed to it. So, being seemingly senseless to go on any further in describing what should be a no-brainer in terms of what is critically absent from the KOBIAN behavioral routine, what emerged (in mind) as a portent to hilarity was very simple – to include a human audience, of even just a single person, reacting, or interacting with KOBIAN, improvisation ally or scripted, and you will have comedy.

    The first concept presented in contemporary introduction to linguistics studies is that ‘WORDS ARE SIGNS.’ In the most basic sense, a word is a sign that can substitute for something having to actually be present in order for two or more people to have a shared experience of what is being spoken or written of/about. For all its imitative qualities, KOBIAN has no experience that the human mind does.

    And just by confidence, the greatest wealth of humor derives from the difficulties, misfortunes, mistakes, follies and tragedies of human experience. Humor is the sense that enables humans to cope with such copious senses of depression, dejection, stress, anxiety, anger, hostility and confusion, most effectively by means of abstraction from those very sense and what gives rise to them.

    It could be food for thought, or I could be full of shite, but whether I’m the fool or the fooled, there’s someone who can get a good laugh out of it. The same can not be said for KOBIAN, as it is capable of being neither, though seeing someone trying to explain why that is to KOBIAN could be a falling-down rolling-on-the-floor crack-up.

    KOBIAN “researchers” apparently did not think to take into account “The World’s Funniest Joke” (For Real)

    [from wikipedia]

    “This article is about research on the relative humour in different jokes and cultures. For the Monty Python sketch about jokes as military weapons, see ‘The Funniest Joke in the World.’
    The “world’s funniest joke” is a term used by Richard Wiseman of the University of Hertfordshire in 2002 to summarize one of the results of his research. For his experiment, named LaughLab, he created a website where people could rate and submit jokes. Purposes of the research included discovering the joke that had the widest appeal and understanding among different cultures, demographics and countries.
    The History Channel eventually hosted a special on the subject.

    The winning joke:

    Wiseman said the joke contained all three elements of what makes a good gag: anxiety, a feeling of superiority and an element of surprise.

    The winning joke, which was later found to be based on a 1951 Goon Show sketch by Spike Milligan, was submitted by Gurpal Gosal of Manchester:

    Two hunters are out in the woods when one of them collapses. He doesn’t seem to be breathing and his eyes are glazed. The other guy whips out his phone and calls the emergency services. He gasps, “My friend is dead! What can I do?” The operator says “Calm down. I can help. First, let’s make sure he’s dead.” There is a silence, then a gun shot is heard. Back on the phone, the guy says “OK, now what?”

    • David, thanks for your thoughtful comments. It seems to me that if part of Prof Takanishi’s aim was to get people considering what constitutes being human, he has succeeded by stimulating analysis such as yours.
      I agree that if the goal is to create true comedy, including even a single human being interacting with the robot would offer the opportunity for real humor. Which makes me think of Craig Ferguson’s Late Late Show (http://en.wikipedia.org/wiki/The_Late_Late_Show_with_Craig_Ferguson) with Geoff Peterson, his “robot skeleton sidekick.” Of course, Peterson is a puppet in the tradition of Edgar Bergen and Charlie McCarthy but without the requirement of ventriloquism. Assuming that we find Peterson funny, is it because we know that he is actually a puppet? (The show credits reveal the designer/builder, writer and voice actor.) Or does he behave sufficiently humanly to touch us?
      Which brings me to your other point: about shared experiences. Your conclusions about KOBIAN sound completely valid to me. But your discussion causes me to wonder about what happens ten or more years from now. In the future we will have shared experiences with service robots, devices which eventually will be equipped with elaborate personalities that are shaped by their experience with us. It seems possible to me that in that future world there might be enough shared experience that robots could be funny in the sense you describe — if not funny to us, at least funny to other robots. But robots making other robots laugh is a line of research limited at present to fiction.
      Thanks for sharing the World’s Funniest Joke with us and for the reminder of a very funny Monty Python sketch.