Exploring the link between Julian Jaynes' bicameral mind theory and efforts to develop conscious AI. 🧠💡 The possible journey from unquestioning obedience to introspection offers insights into the emergence of human-like consciousness in AI.
Great article and good points connecting it to bicameral "framework". I was thinking along the same lines, that the LLMs are currently used as the executive mind and we are the "Gods" with our prompts.
The fact that we don't consider these models conscious is because the breakdown of this bicameral setup hasn't occurred yet. Your work and projects you mentioned seem to go in that direction.
How about adding an Ego component as well? In Vedanta, Ego is considered to be the function of the mind which appropriates all actions for itself. For LLMs, it could be just a self-congratulatory thought after successful completion of a task. Machine Ego would be quite interesting.
Great article and good points connecting it to bicameral "framework". I was thinking along the same lines, that the LLMs are currently used as the executive mind and we are the "Gods" with our prompts.
The fact that we don't consider these models conscious is because the breakdown of this bicameral setup hasn't occurred yet. Your work and projects you mentioned seem to go in that direction.
How about adding an Ego component as well? In Vedanta, Ego is considered to be the function of the mind which appropriates all actions for itself. For LLMs, it could be just a self-congratulatory thought after successful completion of a task. Machine Ego would be quite interesting.