The author says “this must be emergent behavior, can we call this AGI now?”
But there are no details on how the agent assumed control of the computer. Like, if it found the guy’s IP and used a zero day to hack his computer, sure, it’s AGI. But I think it’s much more likely the guy gave the agent access through some integration/API beforehand.
It’s still cool, and creepy, but it’s not as crazy as the guy is making it out to be.