Humans, AI, and Heuristics: Toward a New Philosophy of Education

The advent of artificial intelligence requires a reflection on the very meaning of learning.

Machines are being entrusted with increasingly complex tasks . Precisely for this reason, it becomes crucial to cultivate that uniquely human ability to seek, imagine, and discover new paths.

A new term emerges from the past and is now appropriate to explain what skill we need to hone: “heuristics.”

That is, the ability to generate questions before answers. This will allow us to use AI to support our potential, without sacrificing autonomy of thought and freedom of research.

by Vincenzo Di Prisco

Faced with current changes, education must be profoundly renewed . Not only in its content, but also and above all in its foundations, starting with the methodological principle that defines it.

The term training has its etymological root in the term "form ." Something that leads the mind to associate it with the activity of shaping, modeling, and caring for one's external appearance . Since it also addresses the internal dimension, to the point of sometimes being confused with the term education, which in my opinion is more appropriate, it is ultimately characterized as the process through which an individual's personality, culture, or professional skills are shaped.

Here's the point: training, in my view, no longer addresses a subject conceived as a tabula rasa on which to draw the banks and direction in which the river of knowledge and experience should flow. Training is addressed to an empowered subject and, therefore, must fulfill a different function.

Let me explain: the technical aspects of a profession, the manipulation of objects and the use of tools, are, and will increasingly be, replaced by machines; just as the mind that governs their manipulation and use has already given way to artificial intelligence. And this is without wishing to delve into a dystopian scenario in which the above represents only the first step toward a future entirely governed by AI, which already has the flavor of the present.

We must assign another task to training, one that is far more complex and profoundly human:

a) no longer modeling, but revealing;

b) not to teach how to manage, but to encourage people to invent ways of managing;

c) do not prepare to answer, but guide to knowing how to ask.

I like to sum it all up with a single adjective associated with skill: heuristics. That research method oriented toward discovery and creative problem solving. From the Greek heurískein, "to find" or "to discover."

And here, in defiance of modernity, that giant of modern philosophy who goes by the name of Immanuel Kant comes to our aid : regulative principles (which I call heuristics) do not describe objects, but guide the activity of reason in the pursuit of knowledge.

It is with this in mind that we will master artificial intelligence. We will be enhanced by it, but not deprived of our specific neural activity. We will use it as a prosthesis, capable of enhancing us, not annihilating us.

And what about training? It retains a fundamental value. Try teaching people how to use heuristics with AI and how to avoid the temptation to delegate. It's not easy.

Related insights: