The problem is that the conversation gets sidetracked on ASGI while ignoring the very real actual harms that are already currently happening - like AI "companions" addicting and harming kids, deepfakes, IP-theft, etc.
I also worry about malicious AI used to socially engineer and attack people at scale. We have an aging population that is already getting constantly getting scammed by really amateur attacks, what happens when the AI can perfectly emulate your grandkid's voice?
I also worry about malicious AI used to socially engineer and attack people at scale. We have an aging population that is already getting constantly getting scammed by really amateur attacks, what happens when the AI can perfectly emulate your grandkid's voice?