Ironically the argumentation here is a bit fuzzy and hand-wavey and lacking in logical thinking (abstraction, extension) and references; much like the current s

Ironically the argumentation here is a bit fuzzy and hand-wavey and lacking in logical thinking (abstraction, extension) and references; much like the current state-of-the-art in LLMs. Hmm... Edit: sorry to say basically "no u"... But a lot of the weaknesses you mention are implementation details. For example Chat GPT can already browse the web, should be able to give references and follow a "golden thread" quite well. They hallucinate? Well, so do people - conscious experience is pretty much a hallucination guided to a greater or lesser extent by sensory inputs. They're limited and based on training data? Well... So are people