IBM has passed another major mile-marker in the pursuit of practical artificial intelligence. They're newest brainchild, dubbed Watson (after the founder of IBM) has beat the two most prominent Jeopardy players in the world - Ken Jennings and Brad Rutter. And he beat them soundly.
At several points in these episodes the two men seemed about ready to give up since the computer kept beating them to the punch... this is kind of sad, but at the same somewhat exciting for what the future might hold for Artificial intelligence.
So how is this related to Spinbot?
In much the same way that Watson is designed to recognize language patterns, so also Spinbot is continually improving in this area for the purpose of rewriting already existing text. The better that the computer or program "understands" the language passed, the better it will be able to handle that language.
It is also worth noting that IBM made a splash before by building a maching (Deep Blue) that eventually beat Gary Kasparov in chess, even though the man was considered unbeatable by any human (let alone a computer). Jeopardy posed a much more difficult challenge, however, in light of breadth of knowledge that the computer would have to assimilate and the vagueness of the Jeopardy questions, which the computer would have to decode into something it could understand.