AI Artificial Intelligence

Machine intelligence or Artificial intelligence

When computer programs and algorithms mimic the human ability to think, they are said to be using artificial intelligence. The 'intelligence' part of the phrase refers to the ability to build knowledge derived from acquired data. The 'artificial' part refers to the machine that is using that intelligence, usually a computer or, increasingly, a series of computers that can communicate with one another over the internet. The idea is that artificial intelligence will be able to replicate or at least offer something similar to human thought processes in the near future.

 Typically, artificial intelligence is used for computer applications where massive amounts of data need to be processed and sharing the learned results helps to speed up the process. For example, speech recognition is a typical way that AI is used. This is because devices that learn to correct themselves for certain words they 'mishear' can pass these results on to other speech recognition programs, and thereby helping the entire system to improve its accuracy over time.

For some people, artificial intelligence is very similar to the thought processes people display. The neural networks we form when we think or learn something new are, indeed, similar to the way AI works. That said, AI is probably more akin to a learning system the whole of society uses together rather than the truly imaginative thinking an individual is capable of.

Despite the dystopic visions some have for artificial intelligence, it shows no signs of replacing humans. What it may do is replace the sort of computational thinking people currently need to do. This because it is such an accurate way of number crunching large datasets. If you think about the way that weather forecasts were formulated in the past, it would have taken some complex mathematical models basing current observations on other meteorologists' prior observations to predict whether or not it might rain in the next 24 hours. With artificial intelligence, computers can not only take their own observations but also do the necessary computing to come up with a forecast, passing their data to other computers doing the same work. Crucially, of course, AI learns from its mistakes – when its predictions don't come to pass – and adjusts its own modelling algorithms for ever-improving results.

Artificial intelligence is already making better systems for humans allowing us to understand our environment and our own behaviour better. It will continue to be used in new and interesting ways so long as humans continue to think of novel applications for it. Of course, for every medical application and improvement in public services that can be delivered by artificial intelligence, there is also a military or a potentially repressive one, so there are downsides to take into account, too.