Repository | Journal | Volume | Articles

Artificial intelligence (AI) has dominated recent headlines, with its promises, challenges, risks, successes, and failures. What is its foreseeable future? Of course, the most accurate forecasts are made with hindsight. But if some cheating is not acceptable, then smart people bet on the uncontroversial or the untestable. On the uncontroversial side, one may mention the increased pressure that will come from law-makers to ensure that AI applications align with socially acceptable expectations. For example, everybody expects some regulatory move from the EU, sooner or later. On the untestable side, some people will keep selling catastrophic forecasts, with dystopian scenarios taking place in some future that is sufficiently distant to ensure that the Jeremiahs will not be around to be proven wrong. Fear always sells well, like vampire or zombie movies. Expect more. What is difficult, and may be quite embarrassing later on, is to try to “look into the seeds of time, and say which grain will grow and which will not” (Macbeth, Act I, Scene III), that is, to try to understand where AI is more likely to go and hence where it may not be going. This is what I will attempt to do in the following pages, where I shall be cautious in identifying the paths of least resistance, but not so cautious as to avoid any risk of being proven wrong.

DOI: 10.1007/s13347-019-00345-y

Full citation: