Would you be comfy living with a BOT/ROBOT/COMPUTER smarter than you? Would you always be safe or always be at risk? How?

We have the Ultimate Weapon: The mains On/Off switch or more brutally, pulling the plug out of the socket on the wall!

More seriously, is how you define "smarter". Yes, the computer in front of you can perform very complex mathematical calculations in a tiny fraction of a second, and accurately, while you are still writing the problem down - if programmed to do so - but that does not make it "smarter".

It's still only a box of electronic circuits whose fundamental component is a Transistor: an "On/Off" switch. Thousands of 'em, hundreds of thousands; microscopic and encapsulated in blocks of plastic. The real intelligence is in the human brains that designed them and their circuits, and designed and installed the code necessary to make them switch on and off in the right sequences.

You and I may or may not understand how a computer does what it asks us, any more than we might or might not know the innards of a domestic robot; but these things are still tools to help us, of vastly greater genuine intelligence, carry out particular tasks we need or wish to do.

So-called Artificial Intelligence still relies on someone programming a box of electronics to respond in certain ways, such as making a mobile robot miss obstacles it "sees" with a camera or distance-sensor. It is really a matter of deeply-layered Boolean Algebra (On/Off, A or B, A and B, etc); binary decisions to suit those binary transistors. The brain works in a very different way, and although I don't pretend to understand neuroscience, I do know it is a vastly more complex system than the computer or robot.

If the electronics control us rather than vice- versa, it will be entirely our own fault by designing and using them is such a way that we make ourselves depend far too deeply on them. It will not be their fault - they are inanimate objects.

I don't think we need fear anything more serious than a clout round the ankle from a robotic vacuum-cleaner as it wombles around the lounge carpet, let alone murder by a domestic son-of-HAL ! 

Makes for enjoyable fiction, but for a robot to turn "evil" all by itself it would have to have something like human emotions. More likely an attacking robot would be one programmed by its builders to attack.

I don't fear robots. I worry far more about how computers and the Internet are used by human beings to satisfy human greed, vanity, power, etc. A computer or a web-site cannot have those desires, as it's inanimate, but its operator might.

If we can't reach the switch in time? Run!