Now that Artificial Intelligence (AI) has left the movie theater and started to arrive on our desktops and mobile devices, we need to create new methods to control information so that AI can really start to be smart. IBM's latest moves in the AI space reflect the need to put more fine-grained, granular, context-specific and data-type aware controls into our computing. IBM will of course tell us that it has created another a cloud-based software service designed to offer customers a route to "previously unobtainable insights from their data", when what we really want is a deeper explanation of what IBM wants to do with our data, where it thinks it needs to apply pressure to our business systems, when these functions will be exerted and how it will put this whole new more complex game play into motion. In other words, don't just tell us you're smart – please'show your work'. Diving into the IBM AI machine for answers then, general manager for IBM analytics Rob Thomas has spoken of what he calls a new'information architecture' for collecting, managing and analyzing data.