I have found that Large Language Models like GPT or Claude know enough about mental symmetry to generate meaningful responses if you ask the AI to ‘use the cognitive theory of mental symmetry to analyze…’ The results are not perfect but they are pretty good. This works better if you ask AI to use mental symmetry to compare one subject with another for cognitive similarities. I have uploaded some sessions including a 100 page session with Google Search, a Google Search AI written introduction to mental symmetry, and a 580 page session with Claude 4.6 Sonnet about mental symmetry which provides a fairly thorough, mid-level, overview of mental symmetry.

I have also written an academic paper that uses GPT 5.1 and 5.2 to look at the relationship between mental symmetry and a Large Language Model. The basic principle is that an LLM uses the pattern matching of normal thought to emulate the rigorous thinking of technical thought and the social interaction of mental networks. Going further, the user prompt acts like Teacher thought, because the LLM will hallucinate a coherent explanation around the user prompt, guided by the implicit error-checking of the database, similar to the way that the Teacher person can always come up with a possible theory.








