Things are moving along. I’m trying a new style of writing where I’m handwriting all notes and then begin typing. This seems to be helping tremendously.
Things are moving along. I’m trying a new style of writing where I’m handwriting all notes and then begin typing. This seems to be helping tremendously.
One important point that Sherry Turkle raises in Life on Screen is that bots and other Artificial Intelligences (AI) can be thought of as interfaces (Turkle, 1995: In particular 102-124). Any bot can be programmed to interface with and communicate to and from backend databases. TinyMUD’s famed Julia bot is a prime example (Turkle, 1995: 89-90). Julia would, among other function, continually update a map of the MUD space, providing directions to different locations. She also served as a sort of e-mail system, storing and delivering user to user messages.
Yesterday, I encountered another bot acting as a front end interface. In this case, the nameless bot was part of Chase Bank’s automated phone service system. This nameless bot, while given a female voice, had little to no personality and few trappings of ‘humanness.’ Unlike Julia, the Chase bot was given no name. “She”1 also didn’t appear have any colloquial responses, as opposed to Julia who can be downright sassy.
The most thought-provoking part of my interaction with the Chase bot, came about a minute into the interaction. As part of her interaction script, the bot stated:
By the way, if you know the option you want, you don’t have to wait until I finish speaking to say it. I don’t mind if you interrupt me.
I was more than a little taken aback by this. I already knew that this was the case, and in fact have known to talk over the bot for a while. But this was the first time a bot had given me permission to do this. And that started to make me think about how interacting with a service bot changed the cultural rules of the interaction. First and foremost, the notion of civility is more or less removed. It’s easy to see how, in these types of interactions, manners become inefficient. Interacting with an emotionless entity that doesn’t understand rudeness removes the possibility of abuse from the encounter.
Ironically, this eerily mirrors how robots were treated in R.U.R (Rossum’s Universal Robots), the play that introduced the term “robot.” In Karl Capek’s play, the heroine, Helena, is first confronted by a robot secretary who appears to both her and the audience as completely human. To prove the secretary’s true nature the factory supervisor orders her to be disassembled and her parts brought to Helena for inspection. While Sulla, the secretary has no reaction to these orders, Helena is horrified at this notion:
Helena: You would have here killed?
Domain: You can’t kill machines.(Capek and Selver, 1923: 16)
The robot has no care about its own existence and no concept of fair or unfair treatment. And as Capek notes through the voice of the factory supervisor Domain, this is key to their efficiency:
Domain: Young Rossum invented a worker with the minimum amount of requirements. He had to simplify him. He rejected everything that did not contribute directly to the progress of work. In this way he rejected everything that makes a man more expensive. In fact he rejected man and made the Robot.(Capek and Selver, 1923: 13)
Approximately 85 years after R.U.R. was first published we see how prescient the work was. Capek’s Robots were a thinly veiled allegory for the ultimately alienated and expendable proletariat. The Chase Bot fits this description perfectly. No personality, not even a name, and it gives you the permission to be rude to it, informing you that like Sulla, it doesn’t mind abuse.
Until this point my focus on Bots as interface has been focused on situations where they might maintain the illusion of humanness. However, my encounter with the Chase bot and in particular how identifying it as an AI allows it to be mistreated under typical conversational practices has led to question that theory. Perhaps, the promise of an efficient interaction is enough to allow the removal of human interaction. Perhaps.
1 – I begrudgingly apply female gender designations because this bot was given a neutral female voice and I have found that constantly referring to bots as it tends to become difficult to read/parse.
Bibliography
Capek, Karel, and Selver, Paul. 1923. R.U.R. (Rossum’s universal robots) A play in three acts and an epilogue. London: H. Milford Oxford University Press.
Turkle, Sherry. 1995. Life on the screen: identity in the age of the Internet. New York: Simon & Schuster.
Today, while mowing the lawn at Dre’s house I had an interesting experience. As I stopped to take a break, I noticed Lila, one of Dre’s cats, sitting in the kitchen window. When I walked to the window to say hi, Lila appeared to get very skittish1. She shifted position and began to furiously smell the window. At first this behavior puzzled me. As far as I could tell, I wasn’t making any motion that could be interpreted as threatening. So what was causing this reaction?
Then it occurred to me: Lila was sniffing so furiously because she couldn’t smell me through the window. So, for her, a key sensory identifier was missing. Perhaps she was experiencing cognitive dissonance or some feline version of Freud’s Uncanny. Either way, I suspect that she didn’t know how to resolve the fact that the entity in front of her looked like me but had no smell. I had essentially become a “bot” to her: something that appear to simulate myself without manifesting all of the necessary components to be me.
1 – Admittedly Lila is often very skittish, so I could be completely misinterpreting her behavior.
oops… had an issue with my hosting company… all resolved now!
Today I made the mistake of updating Yahoo!Messenger. While v7 seems to work quite nicely, it removes the “follow this chatter” feature that has been so useful to me. And with that my research capabilities ground to a halt. The reason is that without that feature, I can’t track a bots movement through different rooms.
Thankfully, after a bit of googling, I came up with:
http://www.oldversion.com/
Old Version collects and archives previous versions of popular public software products. And within minutes I was able to download the version of Messenger that I’ve been using.
So should you, in the course of your research, come up with need to track down a previous software version, I’d start your search at Old Version.