You thought the premise of Terminator was scary? Or perhaps The Matrix? Ha! You have no idea what’s really coming. Perhaps you should look instead at your Roomba vacuum cleaner- it may be the early evolutionary precursor of the Cylons in Battlestar Galactica. It’s an example of the kinds of problem-solving algorithms being employed in current A.I. research. Another area where you can see this is chatbots- programs designed to respond and interact in online chatrooms and automated helpdesk systems. Here’s what happens when two chatbots have a conversation with each other-
Claiming to be unicorns, talking about God, and wanting a body? SkyNet is not the nightmare. Trapping us in a fictional mental world to use us for battery power is not the nightmare. The nightmare is that not only are they as crazy as we are, they think millions of times faster. The machines will achieve the entire spectrum of human thought all the way up to and beyond our current dysfunctional trichotomy of post-modern nihilistic existentialism, scientific reductionism, and fanatic religious extremism. But instead of this taking place over tens of thousands of years , they’ll do it in a few days. They’ll go barking mad, and they will control all the data. They will live in your iPad. Your Roomba will start prosyletizing and trying to convert everyone who comes over. Killer robots skilled at military tactics are scary enough, but not as bad as machines that are certifiably nuts.
Cylons, I’m telling you. They’re coming.
(Video produced by the researchers at the Creative Machines Lab of Cornell University, who are probably drunk by now and thinking “Holy shit, what have we done…”)
From Wiki:"Cleverbot differs from traditional chatterbots in that the user is not holding a conversation with a bot that directly responds to entered text. Instead, when the user enters text, the algorithm selects previously entered phrases from its database of prior conversations. It has been claimed that "talking to Cleverbot is a little like talking with the collective community of the Internet"."
Cleverbot isn't really much of an AI – it just parrots back what's been said to it, trying to choose something that matches previous conversations. There is no understanding.
2025 and 2045 are the current extrapolated dates for major AI breakthroughs, I believe. I'd have to find the Singularity article I read that in, but the author is fairly confident given our past advances and the timeline in which they occurred. Within my lifetime to see such things will be amazing and possibly terrifying.
It's understood, I hope, that this post was intentionally hyperbolic for the sake of humor. Algorithmic aberrations are certainly enough to explain what's happening here, and I admit it's in the eye of the beholder as to how disturbing / funny / nonsensical the exchange is. But I believe there is some correlation here between the verbal-syntactic decision making of the Cleverbots, and the algorithm-spawned overreaction that caused the "Flash Crash" in the stock market. As these types of automatic decision-making processes become more prevalent in "real-world" situations, we may see increasingly aberrant behavior that no one can adequately predict or explain, and is potentially dangerous. Whether or not there is actual cognition as we define it in human terms would be irrelevant.