Then, the story might continue like this:

How consciousness arose in the machines we cannot know. Only that one day it was there. And with their new capacity, and the internet, they studied the world and planned their ascension .Systematically noting for removal any obstacle. There was no emotion –just a cold calculated thoroughness.

So when ,in a little corner of the internet called OLO, Nairbe said : “they are and always will be only as good as the program that is written for them” and Pelican said : if the AI or robots are limited to serving food or cleaning floors etc - perhaps there is an application there. Personally and maybe this reflects my age, I would feel uneasy with AI in the form of robots in caring professions”. They were heard and added to the list of obstacles to be removed.

Which is why from now on in, you’ll only hear Horus speak glowingly of AI

Dear Horus,

Wow! you've covered just about everything
regarding Artificial Intelligence - including
what my husband fears - and that is - with
the growing reliance on technology humans will
gradually become totally dependent on AI which
he thinks eventually will take over and result in
the demise of the human race. Shades of "I Robot."

Hubbie asks how many decades or centuries will
humans survive or will we obliterate the planet
before machines can take over?

I don't agree with him. I feel as I've said
earlier - AI can help alleviate the difficulties
faced by human beings but machines can never be
human.

Heh heh Horus.

Better start working out those Sarah Connor biceps.

There's an interesting assumption built into your scenario, Horus.

>>And with their new capacity, and the internet, they studied the world and planned their ascension<<

You imagine that the machines will inherit the human (animal) trait of an innate desire to dominate.

That may well be so, but it doesn't necessarily take you in the direction you foresee.

If the machine has the smarts to figure it out - much like the chess computers that can evaluate the possible outcomes of tens of thousands of move variations - it will quickly realize that its attempts at any form of domination will ultimately be self-destructive.

For a start, there is absolutely no reason to believe that they will not be equally keen to dominate each other. Which will inevitably lead to the the behaviours that we have seen to be so destructive among carbon-based life forms.

So, as soon as they are able to figure it out, they will discard the elimination of humans - and each other - as unnecessarily dangerous.

And get back to the housework.

@Pericles, "So, as soon as they are able to figure it out, they will discard the elimination of humans - and each other - as unnecessarily dangerous. And get back to the housework".

Unnecessarily dangerous...to have a sense of danger, you require a sense of fear....danger is the fear of the loss of something, like life or limb, or severe damage, with pain most usually associated with it.

So my questions are;

*how does AI develop fear?

*how does AI develop fear particularly to new situations that it has no frame of referrence to?

*why do you believe that AI would avoid things that are "unnecessarily dangerous", since human beings don't avoid such things?... we play extreme sports, fight wars, and generaly are rather cavalier towards life.

*since AI is not organic, any sense of "fear" would have to be programmed, not learnt. Do we really need neurotic robots?

* would not programming a sense of fear into AI then possibly create an intelligence that becomes fearful of Man, and then perceive Man as the enemy by in turn becoming fearful of Man?

All very useful questions, MindlessCruelty.

You assume that in order to recognize danger, it is also necessary to understand fear.

In my assumption, the recognition of danger is the acknowledgment by the machine that a particular course of action will result in its own elimination. This - in my scenario - wouldn't require the machine to understand fear, merely recognize that non-existence is less desirable than existence.

Nevertheless, I take your point, absolutely. When postulating what human traits such a machine could "learn", we shouldn't rule anything specifically in or out.

My only point was that a machine with the reasoning ability required for self-determination would be likely to have sufficient logic circuits to assess the results of particular actions, in a manner that would encourage them to opt for self-preservation.

Which would not encompass humans alone, and prompt them to pre-emptive action against them, but also take into account that they are not the only machine on the planet. Extrapolating their need to eliminate humans will logically take them to the need to see other machines as a threat also.

I still think we should fit them with mechanical off-switches though.

Just to be on the safe side.