CHATGPT 4 is at 1.8 trillion parameters. Perhaps the magic will occur at 4 trillion, or 8 trillion or more parameters. Any reasoned guesses? Perhaps some compo

CHATGPT 4 is at 1.8 trillion parameters. Perhaps the magic will occur at 4 trillion, or 8 trillion or more parameters. Any reasoned guesses? Perhaps some component is missing from current LLM models and, despite Moore's law, we've far exceeded the required number of parameters to achieve human-level intelligence. Perhaps its time to pause and examine how LLMs "fit into a bigger picture."