[84 / 7 / ?]
Quoted By: >>51021423 >>51021471 >>51021550 >>51021560 >>51021653 >>51021748 >>51021849 >>51022676 >>51022983 >>51023680 >>51023746 >>51026636 >>51027679 >>51027992 >>51028027 >>51030876 >>51033472 >>51034990
For the sake of argument, let's presume that in future super intelligent AI will be integrated into almost any device around the world.
And again, let's presume that those "famous science person says he is afraid of what AI could do" PR stunts turn out to be correct and we find out that AI does present danger to survival of humanity.
In that case, you would probably want to disable it as easily and as quickly as possible.
Now I ask you, /g/, to design an AI killswitch.
Since it is super intelligent, it is better coder than any human, so solution should probably be on physical level.
optional hard mode: AI has decentralised infrastructure, you can't just disconnect power from it's central server farm.
And again, let's presume that those "famous science person says he is afraid of what AI could do" PR stunts turn out to be correct and we find out that AI does present danger to survival of humanity.
In that case, you would probably want to disable it as easily and as quickly as possible.
Now I ask you, /g/, to design an AI killswitch.
Since it is super intelligent, it is better coder than any human, so solution should probably be on physical level.
optional hard mode: AI has decentralised infrastructure, you can't just disconnect power from it's central server farm.
