Machine learning or artificial intelligence is when a machine becomes able to think like a human. However, machine learning happens when thousands of choices are fed into a database, and the computer learns what do via trial and error. This isn’t actually how humans learn to think. While babies are not born knowing how to talk, walk or think critically, they do have some innate abilities. They develop and learn because of these natural abilities, and that is what AI engineers are trying to teach machines to do, to rely on knowledge that can be “born” into them.
It’s a form of common sense, more so than coded rules programmed into the software. Making AI better involves neural nets which are computer elements that are linked together the way neurons in the brain are. Neural nets might be more human-like, but they are lots more work, needing thousands of data points. Even then, they are not always correct, though some success has been found in a program that can play poker better than a human. Machines are slowly learning how to not only recognize objects, but also the relationships between them.