When we were kids, we believed in magic, imagined superpowers, and a fantasy where robots would one day follow our commands, undertake our meanest tasks, and even help with no effort with just a simple push-button.

The general public and intellectuals still overestimate artificial intelligence. While some sound the alarm at the future emergence of a superintelligence that will dominate us all, the greatest fear is that we put out integrity in the hands of machines they are not as smart as we want to think.

Artificial intelligence is a branch of computer science that studies the design and construction of machines capable of replicating human reasoning and, therefore, of solving tasks for which they have not been previously programmed. On the other hand, machine learning is an artificial intelligence technique with which machines are trained to recognize patterns: defining the characteristics of an object (morphology, colours, proportions), giving a lot of photos of that object until, at a certain moment, it will be able to search and find those characteristics and determine the probability that in an image there’s that object.

The gap between expectations and reality has been an inherent part of the field since its foundation and has been permeating the collective imagination, blurring already diffuse concepts – what intelligence is – and feeding false conceptions about the functioning and capacities of these
systems, so we will need to lift the veil of vagueness that surrounds artificial intelligence.

In the resulting essay, the first stone of this problem that is now more than half a century old is the meaning of the term “intelligence”. Right now, there is no accepted formal definition nor are there tests that can be used to reliably identify it.

The machine can discover paths that humans had not imagined, that is, the machine can create,
but… how far will we handle it?