Of all of the AI technologies, I think voice cloning is probably the most dangerous.
That’s not what happened here, obviously, but I think it’s an example of how much impact it could potentially have.
Even a voice which just sounds sorta like someone famous is attracting national attention. Like deepfakes, we are definitely moving into dangerous ground.
When you take all the AI tech together, it’s going to start generating quite a bit of confusion as to what is real. And it won’t be the fake stuff that will confuse us, but the doubt we will begin to have for the real stuff:
This seems to me to be one of the bigger and more immediate AI risks. Curiously, it’s not really listed here - AI Risks that Could Lead to Catastrophe | CAIS