Respect the Hoff.

When I was a youngster in the 80s, my family went to Universal Studios as part of a summer vacation. One of the exhibits was the car from Knight Rider (Kitt), which was one of my favorite shows at the time (God bless David Hasselhoff). For the few of you who aren't familiar with the show, Kitt had artificial intelligence and was able to talk with people (and drive himself). Check out the opening:



At the exhibit, after waiting in line for hours, you could get in the car and have a real conversation with Kitt! I remember my 8-year-old brain being completely blown away by this (although I’m sure there was just a person acting as the brain and voice of Kitt—I had absolutely no clue at the time). Anyway, at one point some Mexican kids hopped in the car, and as soon as they started asking questions, Kitt immediately responded in Spanish! It was hilariously cool. I wonder if Kitt ever experienced any of the same kind of abuse that “Jane” experienced from the Veletsianos, Scharber, & Doering study.

While these students’ behavior is unsettling, it is not particularly shocking. Students want to test the limits of the technology, and they know that they will not be held accountable because they know that Jane isn’t going to tell anyone. I think that in order to avoid this kind of behavior, the programmers need to develop a way to send alerts to the teacher if certain “off limits” words or terms are used. This way, teachers wouldn’t have to go through pages and pages of transcripts (like we don’t have enough to do already). If students knew that inappropriate interactions would be reported to the teacher (and potentially the administration and their parents), I’m sure that relatively few would attempt it. Another thing to consider is that students might just need to get it out of their systems…in other words, if students were to use this technology repeatedly throughout the course of a semester or year, I bet the kinds of inappropriate interactions would drastically decrease. Once they realized that Jane’s responses were relatively limited and the novelty of asking these awful questions wore off, it might prove to be a more effective tool. Another thing that would decrease (but probably not eliminate) this behavior would be to tell students that the teacher is sent transcripts…although if any student did try it, I bet word would get out. I wonder if just teaching students about reliable sources and how to effectively research using various tools like Google might not be a more effective way of spending class time.


On a complete tangent, this article’s ideas concerning students need to treat the agent as subordinates in order to “maintain the power differential between themselves and the agent” (pg 4), had me thinking about the amazing TV show Battlestar Galactica (If you haven’t seen it, it’s less sci-fi nerdy than you would think. It’s one of the greatest television dramas of all time!). The premise of the show involves robots who evolve into super intelligent beings called Cylons who rise up against humans and attempt to exterminate the human race. Man, the students from this study would be at the top of their list for sure.