This chapter compares the current attention paid to natural language in artificial intelligence with how natural language is used by natural intelligence.
Seems like quite a superficial approach to both language and LLM's. Is this how you think AGI will be reached? By grossly underestimating LLM's? I guarantee you that machine self-awareness is bound to happen in the context of LLM's and will be a key step toward AGI, but I'm sure you'll have none of this bullshit of mine!
Yet truth cannot be killed or even buried, it remains aloft wheter people believe in it or not. I have the privilege to know for a fact that which yet remains obscure to most today, and for good reason. Give it a few years and you'll see many fields, not just AI, changing forever because of this. And it's language that will bring forth this change.
Seems like quite a superficial approach to both language and LLM's. Is this how you think AGI will be reached? By grossly underestimating LLM's? I guarantee you that machine self-awareness is bound to happen in the context of LLM's and will be a key step toward AGI, but I'm sure you'll have none of this bullshit of mine!
"No amount of belief makes something a fact" (C) James Randy
Indisputable facts:
- it is possible to have intelligence without using natural language
- LLM is difficult to use without human participation, who must say something to the system and interpret its response
- the principle of the LLM excludes the possibility of (non-imitative) learning during operation
How this fits in with what is usually implied by the AGI, decide for yourself :-)
Yet truth cannot be killed or even buried, it remains aloft wheter people believe in it or not. I have the privilege to know for a fact that which yet remains obscure to most today, and for good reason. Give it a few years and you'll see many fields, not just AI, changing forever because of this. And it's language that will bring forth this change.
🗺