A European car company uses artificial intelligence to develop autonomous vehicles. These driverless cars need to be able to recognise objects around them, such as people, other cars, buildings, and lines on the pavement to guide them. To develop the ability to recognise other cars, the company uses a type of machine learning technique in the following process:

A very large database is developed, including $10$ million photos posted on the Internet, and then split into two sections, one a development database and the other a test database. Humans select a target $-$ a car image in any of the photos. Humans feed a large collection of verified pictures, some of which contain a car image, into a neural network that proceeds iteratively through the development database in millions of cycles, until eventually the system can identify photos with a car at an acceptable rate. The machine learning system is then tested using the test database to ensure the algorithms can achieve the same results with a different set of photos. Over time, with tweaking by programmers, and by making the database even bigger, using ever larger computing systems, the system will improve its performance, and in that sense, can learn.