Robots & Agency:

Scraps & Conjectures page.

Note on 'bots, cats and manners:

Since I've only got one Bug so far, and since my interest is not in robotics per se but in social interaction and ethics, my main focus thus far has been on how the Bug interacts with our house's two cats [pictures here soon].

Unfortunately, the Bug's sensors & reaction times are too slow for them to react in a socially acceptable manner to the cats. i.e., when they bump into a cat, their first reaction is to treat the cat like any other obstacle, and to try first to bull-doze through it, and then, only after that fails, to attempt to go around. This, unfortunately, is beyond the pale as far as cat manners go.

I had anticipated (and, frankly, hoped for) a bit of a fight between the Bug and the cat. But to my surprise, the cat was smart enough to recognize the Bug as insufficiently reactive. In other words, in failing to respond (quickly enough) to the cat's presence, the bug showed the cat no signs of responsiveness (or, at least, no responsiveness within the cat's tolerance range). So, the cat treated the Bug in a parametric, rather than strategic, manner. This suggests that the difference between an agent that poses parametric decisions and an agent that poses strategic decisions is a fluid one. Bugs are reactive -- they do respond to their environment -- but not reactive enough to count as reactive to a cat (though cats are, after all, the paradigm example of fast reaction times).

My new Insecto-Bot is 'tame' enough to be socially acceptable to the cats, but mostly just get ignored.

Links:

Robots Birddog Pooping Pigeons (Wired, Nov. 28, 2001)

High-wire robot tackles messy birds (New Scientist, 20  September  2001)

SlugBot: Enemy of Slugs (Wired, Oct. 8, 2001)

Scarecrow Robot Built for the Birds (Discovery News, December 4, 2001)