You start off by disagreeing with the GP and end up basically reiterating their point.

Their statement wasn’t that AGI is impossible, more that LLMs aren’t AGI despite how much they might emulate intelligence.