illuminating science

Slashdot is reporting on a mock trial between a supposedly self aware computer and its “owners”, a corporation that used the computer for customer service calls. The article is a little long winded, but the initial discussion about the trial is very interesting.

There are lots of fascinating questions that arise from this sort of thing. If a computer can be programmed to respond intelligently and emotionally to a question, does it really have emotions and intelligence? (The affirmative argument is the Turing test, and Alice is the best we’ve got so far.) What if we start with a blank slate and teach the computer how to respond? Is that different? In the end, the software still tells it what to do. How do you know if your friend really has emotions and intelligence, or if he’s just programmed really well? Lots of interesting points of view in the Slashdot discussion as well.