Illustration: Pedro Nekoi for The Verge

Norbert Wiener, the father of cybernetics, once said that the ideas of every era are reflected in the technology of that time.

Over the past century, computers have borne this reflection, and the AI industry has applied this logic to explaining how the human brain works.

In a conversation published in Daedalus, the journal of the American Academy of Arts and Sciences, Demis Hassabis, head of Google DeepMind, called the brain a biological approximation of a Turing machine. In a public interview in February 2026, Elon Musk put it even more bluntly: people can absolutely think of the brain as a biological computer. On a side note, I’ve always found Musk’s own way of thinking pretty puzzling.

This straightforward comparison has now become an unspoken consensus across the industry. It is simple enough, easy to understand, allows model builders to construct neural networks along this line of thinking, lets product teams tell stories about cognitive advancement, and helps science communicators lower the barrier to understanding for total beginners.

Rarely does anyone question the inherent bias of this analogy.

The complexity of human beings cannot be summed up with a simple computer analogy, and this fact even exceeds the common understanding of most AI industry practitioners. When AI products enter educational spaces following this framework, the entire set of criteria for evaluating learning and intelligence shifts along with it.

This is exactly the core starting point of The Verge’s report. The junk food printed with logos of major AI companies in the cover image perfectly mirrors the current state of this analogy: it is standardized, easy to spread, and has a low barrier to acceptance, but few have seriously verified whether it actually matches the growth pattern of the human mind.

Read the full report here: [AI Education and the Computational Model of Thought](https://www.theverge.com/ai-artificial-intelligence/1003794/ai-education-computational-model-thought)