The Cuteness Factor

In interaction design, one of the real questions seems to be how to get people to interact with something repeatedly so that the designer can understand what to change. New objects are fun to play with once or twice, but if they don't capture your attention and don't work very well on top of that, interest dies out pretty quickly. Over the past few months I've come across three different examples of interaction design in test mode: U of BC's "Cat", Carnegie Mellon's "Snackbot", and the "Blubberbot" created by a participant on the Maker website. These three bots look nothing alike, and were built with completely different goals in mind. However, they have one thing in common: all of them are pretty darn cute.
The Cat was built to test out a participant's reaction through touch to changes in the Cat. The device is supposed to act like a cat, and use your emotions to create different reactions. The Snackbot delivered snacks and some minor conversation in a university building. The Blubberbot is a toy you build yourself that bumbles around and reacts to light sources, like your cellphone. All of these guys sound like they would be fun to play with (after reading the Blubberbot description, I so want one. It's a cute pet, and it can be left alone!), but is that really the point? The first two bots were built to test haptics and push the autonomy of robots. Is cuteness something you usually associate with those fields?
In psychology, there have been several studies done on people with a babyface, or a round face with large eyes. Compared to others, these people are perceived as more naive and trustworthy, and are more likely to be forgiven for a negative action (aka, some type of crime) if it's perceived as malicious. If you apply this cuteness thinking towards objects, maybe we're more likely to forgive mistakes in function if the object is cute. Instead of getting frustrated, and feeling that the device is not working just to annoy you, the user would be more forgiving ("Awww it ran into the wall again!"). Users of all three designs described the bot acting inappropriately, breaking down, and requiring heavy amounts of work to change one aspect. But everyone seemed to like working with them. It's great to try new things in design; I never thought that making them cute would make them easier to test...