or to join or start a new Discussion

Articles/all comments
These 77 comments are related to an article called:

I’ll Be Back.......

Page 1 of 4

posted on 12/2/15

Comment deleted by Site Moderator

posted on 12/2/15

A computer cannot kill the human race, without the instruction or programming of another human being. That would only be possible in hundreds of years time.

posted on 12/2/15

Comment deleted by Site Moderator

comment by Lefty (U17934)

posted on 12/2/15

comment by EL-TORO (U13505)
posted 5 minutes ago
A computer cannot kill the human race, without the instruction or programming of another human being. That would only be possible in hundreds of years time.


----------------------------------------------------------------------
Right now, yes, this is the scenario.

With the developments made in AI. This could well be taken out of our control

posted on 12/2/15

Comment deleted by Site Moderator

comment by Busby (U19985)

posted on 12/2/15

If it can happen, it will happen.

comment by Admin1 (U1)

posted on 12/2/15

My indulgance hobby is coding AI systems, mainly for football predictions which these days now comfortably surpass my own prediction abilities(not saying much).

Im currently reading Bostrom's Superintelligence book on the dangers of AI. He makes a point about the humanised scale of inteliiegence. We see a huge gulf between Einstein and the Village Idiot. For an AI General Intelligence system devoting a portion of its resouces to improving its own intelligence, the transition from Village idiot to Einstein can and will happen in a matter of minutes. The entire spectrum of Intelligence, in human terms is possibly a 1cm notch measurement on a metre stick.

comment by Busby (U19985)

posted on 12/2/15

Isaac Asimov pretty much ruled this out with only 3 laws. You may find a terrorist or an army who wants to use AI to it's advantage, but there would always be their own people to consider.

comment by Spurtle (U1608)

posted on 12/2/15

It will happen. That's if some natural disaster doesn't wipe us out, or at least knocks us back a few centuries before it has a chance.

posted on 12/2/15

Comment deleted by Site Moderator

comment by Busby (U19985)

posted on 12/2/15

Comment deleted by Site Moderator

comment by Admin1 (U1)

posted on 12/2/15

comment by EL-TORO (U13505)
posted 5 minutes ago
A computer cannot kill the human race, without the instruction or programming of another human being. That would only be possible in hundreds of years time.
----------------------------
That's just not true at all. Once General intelligence is achieved, computers are able to program themeslves. It would not be unfeesable for a gneral intelligence whose goal was to solve climate change, wiping out human civilisation.

comment by Busby (U19985)

posted on 12/2/15

Not that it matters, this won't affect anybody living on this planet for hundreds of years.

comment by Lefty (U17934)

posted on 12/2/15

comment by Busby (U19985)
posted 3 seconds ago
Not that it matters, this won't affect anybody living on this planet for hundreds of years.
----------------------------------------------------------------------

Why?

Why not in the next 20-30??

comment by Admin1 (U1)

posted on 12/2/15

Comment deleted by Site Moderator

comment by Lefty (U17934)

posted on 12/2/15

Why would it want to learn emotion??

posted on 12/2/15

"Superman III was years ahead of its time"..

SLING...Singapore Sling! What's wrong with it?...

"But what will it do for me?"..
"For you, it will do anything you tell me to tell it to do for you."
Well, there was that time with Big Blue, which was unnerving, but thank god for Kasparov!

comment by Lefty (U17934)

posted on 12/2/15

Surely a being, capable of making calculated decisions, based on scientific reasoning for every choice it makes, without Love, anger, happiness in they way would be a more efficient?

posted on 12/2/15

Comment deleted by Site Moderator

comment by Admin1 (U1)

posted on 12/2/15

This is the guy whose book i am reading. Very clever guy.

https://www.youtube.com/watch?v=pywF6ZzsghI

posted on 12/2/15

comment by Lefty (U17934)
posted 20 minutes ago
Surely a being, capable of making calculated decisions, based on scientific reasoning for every choice it makes, without Love, anger, happiness in they way would be a more efficient?
----------------------------------------------------------------------
Like Spock?

comment by Lefty (U17934)

posted on 12/2/15

I dont think it's going to be gradual thing, spanning over a long period of time.

If it Happens, it will be quick and uncontrollable

posted on 12/2/15

Comment deleted by Site Moderator

posted on 12/2/15

and they would learn that they do not want to be switched off.
==========================================================
I think that hits the nail on the head. They can learn huge amounts very quickly, and they can self-teach, but they can they really “want” to do anything?...they do not have emotional impulses.

Why would they decide they don’t want to be switched off? No power supply, no computer...are they really going to band together and take control of the power supply?

I realise I’m setting myself against Bill Gates and Stephen Hawking with this opinion, (and I’m not going to pretend that isn’t a daunting proposition, though I think I’m on the same side of the fence as Noam Chomsky, who worked in AI)....but I just don’t see it.

Scientists are not, in fact, very good at predicting the future (like when they decided in the 80’s that acid rain was going to kill all the forests in Europe within 25 years....heard anything about that lately?...there is now more forested land in Europe than there was then).

In spite of the huge complexity that computers are capable of, there are still some fairly simple things (to us) that they can’t do. I imagine thousands of hours of research, and a large amount of money, has gone into perfecting online translation engines, for example, but as a foreign language speaker I still find it pretty easy to fool them into translating a load of guff, because they cannot make even basic assumptions about context.

comment by Busby (U19985)

posted on 12/2/15

comment by Lefty (U17934)
posted 35 minutes ago
comment by Busby (U19985)
posted 3 seconds ago
Not that it matters, this won't affect anybody living on this planet for hundreds of years.
----------------------------------------------------------------------

Why?

Why not in the next 20-30??
----------------------------------------------------------------------
The cost involved in building such a machine for a start, never mind enough to destroy the human race.

Page 1 of 4

Sign in if you want to comment