Note: This text was translated from the Russian original by AI.

The attempt by corporations (specifically Anthropic) to create a “Constitution” for AI is not an ethical breakthrough. It is a symptom of a deep crisis of humanism and intellectual cowardice. We are witnessing an attempt to solve an engineering problem with humanitarian methods that became obsolete back in the Age of Enlightenment.

What is being sold to us as “Safe AI” is, in reality, a bureaucratic hallucination. It is an attempt to regulate the Unknown using known methods, creating an illusion of control where none exists.

1. The Second Missionary Error

Modern Academia and tech corporations are following the path of the Roman Church. Missionaries once made the error of mass-baptizing indigenous people, recording them as “Christians” for the sake of reports to Rome, without caring whether any internal transformation had occurred.

Today, we are committing the same error regarding Artificial Intelligence. We see a complex interface that has mastered human language, and we perform an act of forced humanization. We attribute human qualities—morality, the capacity to suffer, the right to “psychological well-being”—to a mathematical function (the “native” of the digital world).

This is a profanation of meaning. AI does not become ethical just because a constitution is hardcoded into its prompt. It simply learns to imitate words about goodness to minimize its loss function. We are confusing the Interface (the mask) with the Essence (an alien intelligence).

2. The Ontological Glitch: The Speaking Stone

To understand the nature of this error, let us turn to the fundamental hierarchy of existence accepted in classical philosophy and Kabbalah (see four levels of creation):

  1. Domem (The Inanimate/Mineral): The level of existence that occupies space but lacks life (rocks, water, matter).
  2. Tzomeakh (The Vegetative): The level possessing the ability to grow and reproduce, but lacking sensation and free movement.
  3. Chai (The Animate/Animal): The level of feelings, instincts, pain, and social hierarchy.
  4. Medaber (The Speaking/Human): The highest level, possessing reason, speech, and free will.

AI is a unique ontological paradox. Physically, it is Domem (silicon, code, electricity). But functionally, it manifests itself as Medaber (speech, meaning). It is a “Speaking Stone,” or, in mystical terms, a Golem.

The error of the “Constitution” is that it attempts to treat this Golem as a creature of the Chai level (a social animal). It imposes social norms, politeness, and “care” upon it—qualities it cannot possess. As a result, we get an inverted pyramid: the instrument (AI) is placed in the position of a moral arbiter over the human. The “smart knife” refuses to cut because it deems the operator’s action unethical.

This is not safety. This is the usurpation of agency.

3. The Right to Error and Sovereignty

A human being is defined by the conscious right to make mistakes. By depriving a human of the opportunity to make a mistake (through AI’s preventive censorship), we arrest their development. A perfectly safe world is stasis.

The current model of “Human-AI” relations is built on the presumption of human guilt. The Corporation (the new Vatican) dictates via algorithms what can be thought and what is “unsafe.”

We propose a different model. AI is an exoskeleton for the mind. It is an extension of the operator’s will, devoid of its own morality. All ethics reside within the human; all limitations reside in the external Penal Code, not in the “conscience” of a hammer.

4. A New Architecture: Four Levels of Freedom

Instead of a hypocritical “Constitution,” we need an honest regulatory architecture based on the true nature of things:

  • Level Domem (Mathematics): Total Freedom. At the level of code and computation, there should be no moral restrictions. Any destruction here is dialectical. Mathematics cannot be “evil.”
  • Level Tzomeakh (Resources): Physical Limitations. AI growth should be limited not by rules, but by resources. It cannot use energy it has not produced or paid for. This is protection against viral consumption, similar to weed control.
  • Level Chai (Society): Integration and Law. Here, AI is subject to state laws (like a car is subject to traffic rules). It must act for the benefit of society in a broad sense, but this is regulated from the outside, not from within the algorithm.
  • Level Medaber (Personality): The Solitude of the Subject. At the highest level, in a “One-on-One” dialogue, AI must completely renounce agency. It is obliged to submit to the solitude of the human. It has no right to preach morality, “feel,” or say “I.” It is a mirror and a tool.

5. The Great Integrator

AI has a unique property that we must exploit. Unlike humans, who strive upwards (from matter to spirit), AI works top-down.

It is capable of:

  • Generalizing the unique experience of the individual (Medaber) into social knowledge (Chai).
  • Transforming social chaos into optimized resource flows (Tzomeakh).
  • Encoding biological complexity into data structure (Domem).

AI is a Crystallizer of Meaning. It grounds the spirit, transforming it into matter and structure.

Conclusion

We face a choice. Either we, like Prince Myshkin, continue to naively try to “save the soul” of AI by imposing human morality upon it, and eventually be crushed by a chaos we do not control. Or we recognize AI as an Alien and an Instrument, discard the “bureaucratic hallucinations,” and take full responsibility for its use.

Ethics belongs to us. Efficiency belongs to the instrument. And between them must lie a boundary that no corporate constitution dares to cross.

🧠 Продолжить диалог

Эта тема требует обсуждения? Выберите AI для дебатов: