Yeah. That’s my point. You shouldn’t hold the opinion you have presented if you did not have the philosophical discussion

But consciousness may not be a prerequisite for knowledge and truth. I don’t see why a set of computer neurons couldn’t hold the information « the sky is blue » the same way brain neurons can. And in any case, if an AI can perfectly mimic human intelligence and experience, it’s AGI in my eyes.

You are then going into a discussion way more deeper about the « consciousness » of these agents and the philosophical meaning of « knowing the truth ». If you think that an AI need have a conscious experience of telling the truth to be AGI - then AGI might be impossible.

This is actually already included in our current definition of AGI : being able to learn and adapt at the level of the average human. If the agent cannot tell the difference between truth and lie, it does not act like the average human.

Knowing the difference between telling the truth and a lie is very vague. Most AI agents right now can tell the difference between an obvious lie and an obvious truth. If we are talking about less obvious truths statement, humans are not very good at that either.

I’m gonna get my hamster.