Artificial Intelligence as a perennial issue in philosophy

Abstract

The progressive introduction of AI into all technologies, encroaching on the priority of “natural” human thinking, again makes the question relevant: “What does it mean to think?”. To clarify the relationship of AI to the human, an analysis is made of 1) the vulnerability of AI, which reveals its secondary nature – Model Autophagy Disorder (MAD – R. Baranyuk); 2) the experience of training natural and artificial systems by G. Bateson, who discovered a number of levels in training, the highest of which brings the system into contradiction or Double-bind (Db); 3) the Zagorsk experiment. The article demonstrates that it defined the conditions for productive and destructive Db scenarios. A productive way out of the contradiction is conditional on the inclusion of the learner (the weaker side of communication) in universal sociability, which recognizes and protects his subjectivity. Db becomes an instrument for breaking the natural egoism of the living and implanting non-egoistic needs as conditions for survival. That’s how a learner becomes a thinker. The destructive outcome is determined by the monopolization of power by the stronger side of communication at the expense of the weaker one. Human intelligence is thus justified as unnatural, AI – as its truncated projection. But it highlights the issue that remained unsolved in the Zagorsk experiment: determining the conditions that effectively block the monopolization of social power, the tool of which, like any machine, can also be AI.