Cognitive Mechanisms of Large Language Models: Interaction with GigaChat
Концепт: философия, религия, культура June 26, 2025 N. P. Martynenko 1 citation
Modern large language models like GigaChat demonstrate high ability to interpret information, generate text, and adapt to conversation context, including distinguishing between categories of subjects such as 'you', 'we', and 'I'. However, they still cannot reproduce key features of human consciousness such as developed self-awareness and subjective experience. The scientific community has made progress in discussing the Turing test and Chinese Room paradoxes in light of new AI advances, but unresolved questions about the criteria for consciousness remain. Creating a conscious AI continues to be a subject of intense debate and uncertainty.