Marvin Minsky: il visionario dell’intelligenza artificiale e il dialogo con l’Europa

Marvin Minsky (1927–2016) è stato uno dei padri fondatori dell’intelligenza artificiale. Matematico, scienziato cognitivo e filosofo della mente, ha dedicato tutta la sua vita a un’unica, radicale convinzione: l’intelligenza umana può essere spiegata – e replicata – come un processo meccanico.

Dopo gli studi a Harvard e Princeton, approdò al MIT, dove fondò, insieme a John McCarthy, il laboratorio che sarebbe diventato il cuore pulsante della ricerca mondiale sull’AI: il MIT Artificial Intelligence Lab, poi confluito nel celebre CSAIL. Negli anni ’50, Minsky costruì uno dei primi modelli di rete neurale, SNARC, un tentativo pionieristico di simulare l’apprendimento tramite circuiti elettronici.

Negli anni successivi si affermò come uno degli esponenti principali dell’AI simbolica, convinto che per costruire intelligenze artificiali servisse dotarle di rappresentazioni esplicite della conoscenza. Questa visione trovò la sua massima espressione in due opere fondamentali:

• Perceptrons (1969, con Seymour Papert), un’analisi critica delle prime reti neurali,
• The Society of Mind (1986), una teoria secondo cui la mente non è un’entità unitaria, ma una “società” di agenti semplici che collaborano tra loro per generare il pensiero.

Minsky non fu solo uno scienziato, ma anche un pensatore provocatorio. Diceva:

“You don’t understand anything until you learn it more than one way.”

“No computer has ever been designed that is ever aware of what it’s doing; but most of the time, we aren’t either.”

Nel corso della sua carriera, intrattenne relazioni di scambio e confronto con la comunità scientifica europea, in particolare con studiosi che, come lui, cercavano di unire informatica, apprendimento e cognizione. Tra questi c’era Stefano A. Cerri, pioniere italiano dell’intelligenza artificiale e della formazione a distanza.

Cerri e Minsky si conobbero nell’ambito dei congressi internazionali sull’AI a cavallo tra gli anni ’80 e ’90, durante i quali il mondo accademico europeo cercava un dialogo paritario con i colossi americani. Cerri, con il suo approccio centrato sui sistemi intelligenti per l’educazione (i cosiddetti intelligent tutoring systems), proponeva una visione dell’intelligenza non solo come capacità di calcolo o ragionamento, ma come relazione educativa tra uomo e macchina.

Anche se con linguaggi diversi, i due condivisero l’idea che l’intelligenza dovesse essere scomposta, compresa, modellata. E se Minsky guardava alle “società interne” della mente, Cerri guardava alle “società di apprendimento”, in cui umani e sistemi intelligenti potessero costruire conoscenza insieme.

Pur non pubblicando articoli congiunti, il loro dialogo fu citato in diversi contesti scientifici, e Minsky riconobbe in Cerri un pensatore capace di portare l’AI oltre la pura teoria, verso applicazioni etiche, educative e collaborative.

Un’eredità condivisa

Oggi, il contributo di Minsky vive nel dibattito sull’AI ibrida, che unisce reti neurali e modelli simbolici. Il lavoro di Cerri continua a ispirare chi sviluppa tecnologie educative basate sull’intelligenza artificiale.

Due visioni diverse ma convergenti: quella americana, radicale e meccanicista, e quella europea, più umanista e relazionale. Due menti che, pur da continenti diversi, si incontrarono nel cuore dell’intelligenza artificiale, quando tutto era ancora da immaginare.

Marvin Minsky: The Visionary of Artificial Intelligence and His Dialogue with Europe

Marvin Minsky (1927–2016) was one of the founding fathers of artificial intelligence. A mathematician, cognitive scientist, and philosopher of mind, he dedicated his entire life to a single, radical conviction: that human intelligence can be explained—and replicated—as a mechanical process.

After studying at Harvard and Princeton, he joined MIT, where, together with John McCarthy, he founded the laboratory that would become the beating heart of global AI research: the MIT Artificial Intelligence Lab, which later became part of the renowned CSAIL. In the 1950s, Minsky built one of the first neural network models, SNARC, a pioneering attempt to simulate learning through electronic circuits.

In the years that followed, he emerged as one of the leading figures of symbolic AI, convinced that in order to build artificial intelligences, one needed to equip them with explicit representations of knowledge. This vision found its fullest expression in two seminal works:

• Perceptrons (1969, with Seymour Papert), a critical analysis of early neural networks,
• The Society of Mind (1986), a theory proposing that the mind is not a unified entity, but a “society” of simple agents working together to produce thought.

Minsky was not only a scientist but also a provocative thinker. He famously said:

“You don’t understand anything until you learn it more than one way.”
“No computer has ever been designed that is ever aware of what it’s doing; but most of the time, we aren’t either.”

Throughout his career, he maintained an ongoing exchange with the European scientific community, particularly with scholars who, like him, sought to bridge computer science, learning, and cognition. Among them was Stefano A. Cerri, an Italian pioneer of artificial intelligence and distance learning.

Cerri and Minsky met during international AI conferences in the 1980s and 1990s, a time when the European academic world was striving to establish an equal dialogue with American research powerhouses. Cerri, with his focus on intelligent systems for education (so-called intelligent tutoring systems), proposed a view of intelligence not merely as computational or reasoning ability, but as an educational relationship between humans and machines.

Though they spoke different scientific languages, both men shared the belief that intelligence must be broken down, understood, and modeled. While Minsky looked inward to the “internal societies” of the mind, Cerri envisioned “learning societies,” in which humans and intelligent systems could construct knowledge together.

Although they never co-authored scientific papers, their intellectual exchange was cited in various academic contexts, and Minsky recognized in Cerri a thinker capable of pushing AI beyond pure theory—towards ethical, educational, and collaborative applications.

A Shared Legacy

Today, Minsky’s contribution lives on in the debate on hybrid AI, which combines neural networks and symbolic models. Cerri’s work continues to inspire the development of AI-based educational technologies.

Two different but converging visions: the American one—radical and mechanistic—and the European one—more humanistic and relational. Two minds who, from different continents, met at the heart of artificial intelligence, when everything still had to be imagined.