Artificial Intelligence: state of the art — Part I: Introduction to AI

In the first part of our blog series on Artificial Intelligence, our Content Analyst Lara Armas introduces AI to us.

You must have heard the term AI quite a lot these last years. Artificial Intelligence (AI) is indeed one of the hottest technological topics of the century. Nowadays, AI is able to drive cars, recognize our faces, understand what we write and say, and soon it will be able to identify diseases and establish difficult diagnoses. We also know how good AI is at games, as Locaria’s founder Hannes Ben explained in his article about AI and linguistics:

  • in 1997, IBM’s supercomputer Deep Blue beat the World Chess Champion
  • in 2011, the victory of IBM’s Watson over two human contestants in Jeopardy showed again the improvements of AI in the games field
  • lastly, in 2016 AI managed to beat the world’s best Go player, which was long supposed to be a game of intuition rather than cold thinking.

The latest news inform us that, within a few hours only, Google’s AI DeepMind learnt how to play chess and beat Stockfish, a programme considered as the most powerful chess software in the world! DeepMind’s programme won 28 times and lost none out of the hundred games they played. What’s more, it needed no supervision to do so: it was only taught the chess rules, and then improved on its own, playing 44 million times to master this sport (yes, it is a sport!), without having seen any games played by humans.


But what exactly is Artificial Intelligence, and what is it used for?

AI is intelligence displayed by machines and is the term applied when a machine mimics the human mind and its cognitive functions, such as learning and problem solving. AI started in 1956 on the claim that human intelligence “can be so precisely described that a machine can be made to simulate it” (source: Research touched on computer chess, robotics and natural-language communication, but nowadays its scope impacts many different fields, from neuroscience to linguistics, including optical character recognition, as it is now possible to scan and convert handwritten or printed texts into machine-encoded text. It can also understand human speech, produce autonomous cars, interpret complex data like images and videos, and compete in strategic games at a high level.


Machine learning, “the specific scientific methods currently in vogue for building AI”, is such a big deal now because inventing software able to learn over time and become smarter through the accumulation of huge amounts of data is getting easier and easier. The idea of Artificial Intelligence becoming smarter than humans at most tasks is “the single biggest thing that drives debates about effects on employment, creativity and even human existence”. (source: