EMMA: New Ground

The relationship between robots and humans

Current culture is fascinated by the ever-growing presence of robots in and among us. Vivian Sobchack refers to Ballard’s narrative about fetishes with robotics. This fetish indeed has been a common story theme that has been written about over and over again. The movie Millenium Man (based one of Isaac Asimov’s stories), He, She and It (by Marge Piercy), Blade Runner, and other similar stories discuss the relationship between robots and humans. What is the robot’s place in our society and how intimate do we become with them? There are many aspects of this question to explore.

One of the most important questions to ask is whether or not it is even beneficial to encourage closer relationships between people and robots. For example, with the world full of lonely people, are robots safe companions? Multiple research studies (Beck & Katcher, 1996; Katcher & Wilkins, 1993) have concluded that pets are good for the well being of elderly. Due to logistics of animal ownership in some apartments or homes and the possibility of accidental neglect of live pets, some researchers have been looking for pet alternatives. Kahn, Friedman, and Hagman (2002) investigated the effects of giving a robotic dog like Sony’s AIBO (http://www.sony.net/Products/aibo/index.html) to elderly people living alone as well as to children. Interestingly enough, the effects on the two age groups were different. Although there were positive effects on the elderly (decreased loneliness), the same research discussed the negative implications of youth forming relationships with robots since the lesson of responsibility and obligation (moral value) is lost with mechanical pets. Other types of relationships include romantic ones, such as the one Asimov covered in the movie Millenium Man, but I will leave it to the books and movies to do a more entertaining job of discussing these interesting and complicated issues.

If we accept that we are going to have relationships with robots, how do we facilitate this bond? One of the barriers to full acceptance of the robot is its physical appearance. The more mechanical the robot appears/sounds/acts, the more difficult it is for people to form an attachment. So do we just make them more life-like or humanoid to take away this obstacle to full acceptance? Japanese roboticist Masahiro Mori addresses this issue in his theory known as the ‘Uncanny Valley.’ (a.k.a the Creepy Factor) This theory attempts to describe how people respond to the level of realism in a robot’s appearance. If it looks like a robot, then we are okay with that, but as it looks more human, we begin to get “creeped out” as we approach the “dead man” look. Then, as the model moves past that, looking more and more lifelike, we begin to again feel more comfortable with the appearance. Here is the famous graph:


http://upload.wikimedia.org/wikipedia/en/c/c5/Wpdms_fh_uncanny_valley_3.jpg

So let us say we are successful in creating a robot whose appearance promotes comfortable interaction (the peaks of the valley): now what? One aspect of humanity that has always set us apart from machines is the creative process. What happens with robots and the body and performance…our triangular focus! What role do robots play in creative expression and performance? Please view this performance by robots:

Qrio: http://www.plyojump.com/movies/qrio/qrio_fandance.wmv

(To see other videos of Qrio: http://www.sony.net/SonyInfo/QRIO/videoclip/index_nf.html)

One of the inspiring things about performance is the agility and strength of the live dancer. Robotic dancers, on the other hand, are amazing right now because of the intelligence of the scientist who made it. Unfortunately, unlike our persistent appreciation for human agility, we become easily desensitized to the robot’s ability over time. As we see in the movies, cutting-edge realistic computer effects rapidly become the standard, and anything less is disappointing. Unlike the movies, live performance is just beginning to integrate with robotics, and we are still more comfortable with the left-hand side of the Uncanny Valley.

For example, the “ballroom robot” is very obviously plastic and shiny:
http://www.cnn.com/2005/TECH/06/07/robots.ballroom/

The Uncanny Valley concept does not just apply to visual appearance, but also to how the robot moves and communicates. The effects on movement and performance have been graphed below, and as you can see (and imagine), a moving human-like dead person walking towards you would indeed be pretty scary.


http://upload.wikimedia.org/wikipedia/en/a/a2/Moriuncannyvalley.gif

To see the progress on life-like physical robots, see Professor Hiroshi Ishiguro Repliee Q1Expo. http://www.ed.ams.eng.osaka-u.ac.jp/development/Humanoid/ReplieeQ2/ReplieeQ2_eng.htm
What does this mean for the future of performance art? When we get bored with the advances of robotic motion and life-like appearance, are we going to use them or reject them? Will robots be a standard prop in the performances of tomorrow? Will they replace us? Will dancers become puppeteers of these stronger and more agile robotic performers, or will humans still yearn to see the human form in action? Will the audience be happy with a synthetic score created and performed by robots?

Some other interesting thoughts for discussion:


http://upload.wikimedia.org/wikipedia/en/c/c5/Final_Fantasy_TSW_DVD.jpg

Final Fantasy (a 3D movie with life-like characters) had many reviews that basically stated that we are not able to fool the viewer with computer graphics yet. If this is the case, it will take longer before we can do it physically. Nevertheless, robots are looking more real and will be able to fool us one day. Do we allow them to be part of our existence, or will robots be banned to going in circles vacuuming our floors?

On the more technological innovations front:
See Darpa 2005 – Stanford’s Stanley is one of the first cars to be able to navigate a desert with human intervention. http://www.grandchallenge.org/

Artificial Intelligence “AS” performance:
The DARPA challenge was not just a test of AI, but also a performance where an audience was observing the story. Year one (2004) was a comedy of errors (none of the cars went more than 8 miles on the 142 mile path), and year two (2005) was a feel-good story. In addition, the $2 million dollar prize made it very similar to many game shows people spend hours watching.

Issues from the 2004 “performance” that will help us avoid common mistakes when we create our performance:

Palos Verdes High School reason for 2004 failure: Software contained decimal point error. Slammed into 2-foot-tall cement barrier at 22 mph. Knocked barrier over.

Team CajunBot reason for 2004 failure: On-off switch located on side of vehicle. Bumped into a wall on way out of start area. Turned self off.
Read about other issues from the 2004 challenge:
http://www.wired.com/wired/archive/12.05/start.html?pg=15

(Read the full article about the 2004 challenge. http://www.wired.com/wired/archive/12.03/robot_pr.html)

My only wish was that this was advertised a little bit more. Instead of watching people kick each other off islands, here is an opportunity to get the world (especially kids) excited about the sciences. But we will leave this for another blog space to discuss this wish…

posted by Steven G. @ 1:22 PM