For autonomous and artificial intelligence programs, machines need to be able to understand the physical world. Currently, this requires a human to essentially teach the computer what things are, because while a computer can process images and identify objects, it still requires a human to tag the object and tell the machine what the thing is. However, the ability for machines to learn independently has passed a major milestone recently. Both Google and Microsoft have announced that their image recognition programs have surpassed the image recognition capability of the average human. This means that an AI could be left to its own devices and trawl an image database, matching names to things. This is huge news. Without diving too deep…