In a horrifying turn of events ripped straight from a sci-fi novel about the technopocalypse, researchers at Rensselaer Polytech have created a virtual four year old that holds its own beliefs. Based on these beliefs, the virtual toddler is able to reason and draw logical conclusions in the manner of a child its “age”. In short, they’ve created the precursor to an artificial intelligence. They’ve named it Eddie.
Eddie’s thought processes are driven by a massive array of supercomputers consisting of POWER-based Linux clusters, parallel Blue Gene supercomputers, and AMD Opteron processor-based clusters (and if there is a just and loving God in this universe at least one PS3). A combination of some staggering computing power and a partially developed “Theory of Mind” allows Eddie to interact with, understand, predict, and subsequently manipulate the behavior of other Second Life inhabitants. But don’t comfort yourself in the knowledge that Eddie’s actions are limited to Second Life; according to one of the developers on the project, “our technologies can be applied to any digital environment, and indeed we are specifically aiming, with IBM, at environments in which the physical and the virtual directly interact.”
Having some experience with the clientele of <em>Second Life</em>, this is all more than a little bit disconcerting. What we have here is a virtual pre-pubescent capable of interacting, predicting and manipulating the behavior of <em>Second Life</em>’s vast constabulary of perverts. They should have named it L.O.L.I.T.A: it’s only a matter of time before a brainwashed army of Humbert Humberts marches on the Capital. The war is coming, my friends, mark my words.
[Via ITnews — Thanks, Adam!]
