Could you please cite these papers. If by AI you mean LLMs, that is not supported by what I know. If you mean a theoretical world-model-based AI, that's just a tautological statement.