When people hear the words 'artificial intelligence', many still picture the humanoid robots of science-fiction films. In reality, AI has crept into our daily lives far more quietly. If you have ever unlocked a phone with your face, asked a voice assistant about the weather or seen a music app suggest a song you might like, you have already used it, probably without giving it a second thought.
So what exactly is it? Put simply, AI describes computer systems that can carry out tasks which normally require human thinking, such as recognising speech, translating languages or spotting patterns. Most modern AI is not programmed step by step. Instead, it learns from huge amounts of data. A system shown millions of photos of cats, for example, gradually works out which shapes and features tend to mean 'cat'.
The benefits can be impressive. In hospitals, AI tools help doctors to examine medical scans, where a computer may notice tiny details that a tired human eye could miss. 'It's like having a very careful second pair of eyes,' says Dr Priya Shah, a hospital doctor in Bristol. 'But the final decision is always made by a person.' Translation apps, meanwhile, allow tourists to read menus and signs in languages they have never studied, and for people with visual impairments, apps that describe photos or read text aloud can make everyday life considerably more independent.
However, not everyone is convinced that the changes are entirely positive. One concern is bias. Because AI learns from information created by people, it can pick up their unfairness too. If a system is trained mostly on examples from one group, it may work less well for others. Another worry is that AI can sound extremely confident while being completely wrong, so its answers should never be trusted blindly.
Then there is the question of work. Some people fear that machines will take over their jobs. Experts disagree about how far this will go, but many believe AI is more likely to change jobs than to remove them altogether, dealing with repetitive tasks so that people can concentrate on the parts that need creativity, kindness and judgement.
Teachers face their own dilemma. When students can ask a chatbot to write an essay in seconds, how can schools be sure the work is genuinely theirs? Some schools have banned such tools; others are teaching pupils to use them responsibly. At Fernhill Academy in Norwich, history teacher Mark Ellis lets his class ask a chatbot for ideas, then asks them to check every fact in books. 'They're often surprised by how many mistakes they find,' he says. 'It's a shortcut to ideas, not to understanding.'
What seems clear is that AI is here to stay. The key, most people agree, is neither to fear it nor to worship it, but to understand what it can and cannot do. After all, a tool is only as useful as the person using it.