Artificial Intelligence (AI)

Some of the most striking technology in your daily life works quietly in the background, making decisions that once only a human could make. Consider two familiar situations. When you open the map application on your smartphone, it analyses real-time data such as traffic congestion and guides you along the fastest route to your destination. And when you upload a photograph to a social networking site, the site often recognises the friends in the picture and tags them automatically. Both are everyday examples of the application of Artificial Intelligence (AI). The intelligent digital personal assistants you may have spoken to — Siri, Google Now, Cortana and Alexa — are all powered by AI as well.

So what exactly is Artificial Intelligence? AI endeavours to simulate the natural intelligence of human beings in machines, thereby making the machines behave intelligently. The benchmark is human thinking itself: an intelligent machine is supposed to imitate some of the cognitive functions that humans perform, namely

• learning — acquiring knowledge from information and experience,
• decision-making — choosing a suitable course of action among alternatives, and
• problem solving — working out how to reach a goal from a given situation.

How is such behaviour achieved in practice? The aim is for machines to carry out tasks with minimum human intervention. To make that possible, the machines are programmed to build a knowledge base and to take decisions on the basis of what that knowledge base contains. Rather than a human specifying every action in advance, the system consults what it has stored and acts accordingly.

There is a second, equally important ability: an AI system can also learn from past experiences or outcomes and use them to make new decisions. Each result the system produces becomes feedback for the future, so its behaviour is not frozen at the moment of programming — it keeps adapting. …