AGI is tied to the consciousness debate because it is a perfect contemporary equivalent to the age-old question of 'How many angels can dance on the head of a pin?" The science of consciousness is still not fully developed/understood, and there is an ongoing debate about whether artificial systems could truly be considered conscious. Some argue that consciousness is a unique product of biological brains, and cannot be replicated in silicon. Others suggest that consciousness is not binary, and that machines could have a form of consciousness without it being identical to human consciousness.

It's grounded in philosophical/theoretical considerations with key emerging arguments:

1. Functionalism: This philosophical view holds that what is important about the mind is not the physical substrate (like the neurons in the brain), but the computational processes and patterns that take place within it. From this perspective, if an AGI were performing the same kind of information processing as a human mind, it would be reasonable to consider it conscious.

2. Generalization: If an AGI can learn and generalize from one domain to another, it might be argued that it has a form of understanding, which is a component of consciousness. This is related to the concept of "qualia" - subjective experiences that accompany our information processing.

3. Self-awareness: Some argue that for an AGI to truly match human-like intelligence, it would need to have a degree of self-awareness or introspection - understanding its own internal states. This self-awareness is a key component of consciousness.

4. Ethical considerations: From an ethical standpoint, if an AGI has the equivalent cognitive and emotional capabilities as a human, it could be seen as deserving the same moral consideration as a conscious being.

Hope this provides some context. Here are a few opinions of much smarter and better informed people: https://www.researchgate.net/post/Does-artificial-general-intelligence-implies-consciousness