What is AI ?
For a few years now, but even more so since the end of 2017, various actors of civil society, whether they're entrepreneurs, scientists or journalists, talk about AI as a problem for society. It is described as dangerous, as a threat for employment. But a curious fact is that there doesn't seem to be one agreed upon definition of artificial intelligence. How can it be defined ? What actually is artificial intelligence ?
It seems we should first choose a postulate in order to discuss it properly. John McCarthy, one of the founders of the field who started getting attention in 1956, defined artificial intelligence as follows: “Intelligence is the computational part of the ability to achieve goals in the world. Varying kinds and degrees of intelligence occur in people, many animals and some machines”. This is an open, and yet inclusive definition: human intelligence isn't the only type of intelligence mentioned. This definition is the one he uses as a preamble to his other, more elaborate definition of AI: “artificial intelligence is the science and engineering of making intelligent machines, especially intelligent computer programs. It is related to the similar task of using computers to understand human intelligence, but AI does not have to confine itself to methods that are biologically observable”. This second, more precise definition, remains open on how to make intelligent machines. The how is very important, and we will use it as our starting postulate.
On artificial Intelligence in the media
Let's go back to the reasons social actors see artificial intelligence as dangerous. According to whistle blowers, artificial intelligence is putting many jobs at risk. It could become autonomous, and to such an extent that it could spread everywhere and turn against its creators. This seems a rather extreme way of seeing things. Such concerns are indeed valid, but the way to address them is very important too, as it will shape the kind of technologies that will prevail in our society.
First, it seems important to point out that artificial intelligence is already everywhere, but it's definitely not intelligent. Still, we do need some ethical body responsible for AI. To be fair, such an organisation should already exist, as our economic model is already invested in AI. As for every new major scientific discovery, AI has the potential to cause a lot of harm if unsupervised or maliciously. But history did teach us not to repeat our mistakes, and we should be more sensible and not put aside all the good that it could also bring.
Mickaël Camus (PhD) Author
