Read the websites for IBM, Google or other companies working on artificial intelligence, and you’ll see glowing reports of all sorts of things electronic brains can do, ranging from diagnosing X-Rays to composing music, to tutoring students, to driving a car. Google, Alexa and Siri already handle many daily tasks for us, operated by vocal commands, just like in Star Trek. The science fiction dream of wholly electronic people — computers that are indistinguishable from organic, human minds, it seems, must be just around the corner. And yet… spend a few minutes interacting with an artificial intelligence on anything other than the most mundane level, and you start to notice something’s not quite right. Sometimes, it can get downright disturbing. Janelle Shane’s hobby is to train a particular kind of artificial intelligence called a “neural net,” and has certainly seen her share of strange AI behavior. So much so, in fact, that she publishes a blog, “aiweirdness.com,” filled with examples of AIs going wildly astray. Janelle Shane joined Monday 8 O’Clock Buzz host Brian Standing by phone from the Arapahoe Nation in Colorado.