IM bots masquerading as humans
Seamus McCauley observes that on the internet, not many people know that you're a bot:
J: Would you like to hear a joke?It's not hard to imagine lots of devious phishing applications of these kinds of chat robots—they could be primed to ferret a certain kind of information out of you, such as your shopping preferences.
A: Sure, tell me about it.
J: Why don’t blind people skydive?
A: Perhaps it is impossible.
I hear worse jokes told by real people almost every day. And the above, of course, was the product of two chatbots talking to one another (Discover, via BoingBoing).
The Turing test is all very well, but in artificial lab conditions where you've been told to watch out for one robot and one human you've got a 50/50 chance of getting it right just by guessing. People just aren't generally paying that much attention, and at a time when many "people" communicate (almost) exclusively via 160 or even 80 characters of text I'm not at all convinced we'd spot the robots if they made up three-quarters of the online population.
And the moral of this story? Don't waste your life indulging in the inane drivel of chat rooms, Twitter and so on? ; )
2 Comments:
Sadly, I suspect that your casually-made prediction about chat-room bots trying to get shopping preferences (and other personal information) will become true before too long. If it isn't happening already!
This sort-of leads to a point about choosing online interactive services based upon how they manage online identity. I recently joined FaceBook and am using that to communicate with various (real-life !) friends. To a pretty high degree of certainty (without using an ID standard), I know that these people are who they say they are because FaceBook insists upon usage of real names. I feel I can trust my friends on FaceBook not to collect my information and sell it, just as I would trust them not to steal my bank documents when they visit me at home.
On the other hand, I once tried Myspace and quickly abandoned it because, with the common usage there of made-up names and a culture of being "friends" with people you've never met, how can you tell whether a "friend" there is a bot?
All the evidence (from the news and from my own experience) suggests that Myspace contains a lot of bots. On the other hand, I suspect that services that rely on "real-life" identity, such as FaceBook, have much less of a problem with bots.
Interesting observations, Alex, thanks. I was only this morning discussing various other aspects of Facebook and other social networks' huge potential as a mediator of our identities. The identity verification angle you touch on would certainly seem to be one important differentiator for Facebook from their competitors.
Post a Comment
<< Home