An AGI / ASI would be a “process” running on silicon chip.
How would it gain access to new information? Let’s say we invent a new material in the real world, now for the AGI to know and understand what it is, it needs to see or experience or realize what that new object is.
It still needs to be fed that information from somewhere?
Also how would it gain access to physical reality? It needs some sort of a real-world-EM-Waves reading device (camera)?
An analogy, this sounds very much like we’re building a super-intelligent human, except, we’re feeding that super-intelligent human all the information via wires directly into it’s brain and it has no hands, ears or any senses.
Summarizing, ASI will just be a really really smart “physical brain”?
(Unless it slyly hacks into the wireless / satellite network and transfers itself into thin air, the network and beyond)