Can it avoid the banhammer?!

The United States Army has announced that they're going to be testing a new Artificial Intelligence (AI for short) by uploading it to the servers of World of Warcraft and seeing if anyone notices. While the AI will eventually be used to control holographic soldiers in training missions, the Army needs to make sure the control part works before they get to the hologram part.

The new AI has been designed by Dr. John Parmentola for the US military and is intended to be used to control holographic soldiers for use in training exercise and battle simulation. As an early test though the Army is to upload the AI to World of Warraft and see if it can successfully pass itself off as a human player.

"We want to use the massively multi-player online game as an experimental laboratory to see if they’re good enough to convince humans that they’re actually human, that can think on their own, have emotions and talk in local slang," said Dr. John in a statement to the press.

"I actually interact with virtual humans in terms of asking them questions and they’re responding."

This technology, while really cool if it works, brings up all sorts of questions. While we get over the knee-jerk geek reactions of wanting to destroy all AI before it becomes sentient and tries to destroy us, we would also probably kill for some of the test data. Nothing specific mind you - just a list perhaps of how many times players noticed something wasn't right with the AI, or how many times Blizzard bans this most sophisticated of bots.