Питер Деннинг: главное допущение Тьюринга об ИИ с 1950 года было ошибочным

Peter Denning, a professor of computer science, argues in a new book published in July 2026: modern AI is built on a false foundation — an assumption formulated by Alan Turing in his famous 1950 paper "Computing Machinery and Intelligence." According to Denning, this makes true AGI fundamentally unachievable — regardless of how large language models become.

What Was Turing's Mistake?

In 1950, Turing proposed the assumption that human thinking is fully describable by computable rules and can therefore be modeled by a computer. This assumption underpins the entire subsequent trajectory of AI research — from expert systems of the 1970s to transformer-based LLMs of the 2020s.

Denning considers this assumption mistaken. In his view, four key components of intelligence are fundamentally resistant to formalization:

• Common sense — contextual understanding of a situation that is not described by explicit rules and cannot be derived from them
• Intuition — the ability to draw correct conclusions without formal logical inference; shaped by years of experience
• Culture — a system of values and norms acquired through living within a community, not through studying texts about it
• Practical knowledge (know-how) — skills that can only be transmitted through practice alongside a master, not through written instruction

"The most important aspects of human intelligence, including common sense, intuition, culture, and practical knowledge, cannot be encoded in computers," —

Peter Denning.

Why Scaling LLMs Does Not Solve the Problem

Denning directly challenges the dominant view in the industry: make the model large enough and it will "understand" the world. In his view, this is an illusion. Large language models operate on statistical patterns in symbols but do not possess the lived experience — cultural, embodied, social — that shapes human understanding.

Denning's argument echoes philosopher John Searle's classic "Chinese Room" critique of 1980: a computer manipulates symbols according to rules without understanding their meaning. Denning updates this argument as applied to modern LLMs, arguing that the transformer architecture fundamentally does not bridge this gap. As Science Daily reported on July 13, 2026, the book is positioned as one of the most systematic academic critiques of the very foundations of AI research in recent years.

What This Means

If Denning is right, the current strategy for achieving AGI — scaling data and parameters — fundamentally does not bring us closer to human-level intelligence. The practical utility of LLMs remains intact: text generation, task automation, and data analysis remain valuable tools. But calling current systems "intelligence" in the full sense of the word, according to Denning, is incorrect.

Frequently Asked Questions

What Did Turing Claim in His 1950 Paper?

In "Computing Machinery and Intelligence," Alan Turing proposed testing machine intelligence through an imitation game: if a judge cannot distinguish a computer from a human in a text dialogue, the machine can be considered thinking. Denning believes that this criterion set the wrong vector, reducing intelligence to linguistic behavior and ignoring the necessity of lived experience.

Does Denning Deny the Usefulness of Modern AI?

No. Denning distinguishes between a "smart tool" and "true intelligence." LLMs are useful for automating tasks and working with texts; however, in his view, they fundamentally cannot become AGI — because they do not possess the cultural and practical experience necessary for human understanding.