An engineer type does a pretty good job of breaking it down, in somewhat vague terms, as an existential threat.
“Robots plus online AI is a different threat. Online AI could possibly shut down or sabotage human defences while turning our defences on ourselves in coordination with a robot uprising.
“Possible, but hopefully unlikely.”
Commenting on the likelihood of his warning, he added: “Only if regulators are stupid and allow them to be superhuman or to access superhuman powers.
“On the other hand, our regulators often are stupid.”
Yeah! The regulators! If that’s the defense, then we’re as good as dead.