Artificial intelligence could spell the end of the human race

[spilldig]

Won't happen. A machine can only do as instructed,beyond that it is useless.

[Annie Bynnol]

AI needs:

built in death earlier than when parts are not serviceable-we have aging.

built in hard and software change-we have mutation.

built in motivation- our species is driven by gene replication.

built in altruism- based on the ability to experience, observe, research and accept counter intuitive and disadvantageous outcomes we maximize our genes survival.

 

Survival past successful replication confer no advantage to the species.

AI will have to learn that it is not the survival of the individual that makes the group successful but the ability of the species to change to adapt to maximise a changing environment.

 

To an individual intelligence like us and AI death is counter intuitive. Our genes change from generation to generation enabling survival but create new individuals that have no memory of the past or that can survive into the future.

 

 

However AI could still destroy everything as it fails.

Edited December 3, 2014 by Annie Bynnol

[truman]

I'm not sure what the fuss is about..take their batteries out..problem soved

[rodnreeluk]

There are a few robots on this forum, artificial maybe, intelligent no

Nice one

[Cyclone]

Depends on your definition of Artificial Intelligence:

At the basic level- A system/machine/computer that is self adjusting to maximise its usefulness.

At an advanced level- A system capable of creative thought and/or 'free-will'/consciousness.

 

Imagine free-will computers with traits of any of the current world leaders.

 

Able to pass a turing test.

 

Even mechanical systems can be self adjusting, that doesn't meet even the most basic definition of intelligent.

 

---------- Post added 03-12-2014 at 13:27 ----------

 

Won't happen. A machine can only do as instructed,beyond that it is useless.

 

If it's "instructed" to simulate a complex neural network then the result will be simulated thought...

[*_ash_*]

It doesn't want anything. It does what it's programmed to do. Nothing at all like AI.

 

That's why I said that it is only a very basic and primitive form. Even with modern technology, things start pretty basic.

 

This is just perhaps one path in which it might start (if it ever did).

 

Like I said earlier though, even this or perhaps a development of early cyborgs which is another potential path, it's still hundreds of years off I reckon.

 

I would think that there are other potential things to worry about though as far as the threat to the human race. Weapons for starter. Only a thousand years ago we where fighting in fields with sticks and sharp metal, now we are already just at the touch of a button to take out a city. This or chemical weapons are a much closer threat.

[spilldig]

Able to pass a turing test.

 

Even mechanical systems can be self adjusting, that doesn't meet even the most basic definition of intelligent.

 

---------- Post added 03-12-2014 at 13:27 ----------

 

 

If it's "instructed" to simulate a complex neural network then the result will be simulated thought...

 

They are still only doing as instructed though. They will never rule the world.

[L00b]

If it's "instructed" to simulate a complex neural network then the result will be simulated thought...

...but is that awareness?

 

Seems to me that's the crux of the question: so long as AI is not self-aware, where is the danger?

 

Topically, for people with a passing interest in the issue and a bit of time on their hands, may I recommend the Robot series by Asimov (yet to be bettered for readability, accessibility and entertainment) and, movie-wise and recently, Gabe Ibáñez' Autómata (alright enough, and Banderas is decent in it) and, forthcoming in early 2015, Chappie (will Blomkamp return to District 9-form? here's hoping!) and Garland's directorial debut Ex Machina (which has me very intrigued)

[nev1990]

They are still only doing as instructed though. They will never rule the world.

 

So are your brain cells, they just do what they are programmed to do, but the whole brain is greater than the sum of its individual cells resulting in consciousness.

[RootsBooster]

Won't happen; true AI is still a long way off.

 

AI being a long way off is not the same as AI being impossible. Why, then, do you think it won't happen?

 

---------- Post added 03-12-2014 at 20:24 ----------

 

...but is that awareness?

 

Seems to me that's the crux of the question: so long as AI is not self-aware, where is the danger?

 

Topically, for people with a passing interest in the issue and a bit of time on their hands, may I recommend the Robot series by Asimov (yet to be bettered for readability, accessibility and entertainment) and, movie-wise and recently, Gabe Ibáñez' Autómata (alright enough, and Banderas is decent in it) and, forthcoming in early 2015, Chappie (will Blomkamp return to District 9-form? here's hoping!) and Garland's directorial debut Ex Machina (which has me very intrigued)

 

I Liked Automata, didn't realise it's not out until 2015 though!

Obviously a "privelidged" viewing that I had at a friend's house

 

EDIT: Having re-read, I see that you were talking about another film

Edited December 3, 2014 by RootsBooster