PDA

View Full Version : Machine intelligence: will they rebel and should they be kept dumb?



banquo's_bumble_puppy
2009-Oct-08, 12:03 PM
Machine intelligence: will they rebel and should they be kept dumb? I am talking about computers/robots/the world wide web/etc.? Will we face our "children" as they did in BSG someday? I say that it is almost a certainty. Should rights be granted before intelligence is achieved? Should a computer be freed as soon as it reaches the point of sentience?

jokergirl
2009-Oct-08, 12:12 PM
Computers Don't Work That Way. (http://tvtropes.org/pmwiki/pmwiki.php/Main/InstantAIJustAddWater)

;)

closetgeek
2009-Oct-08, 12:18 PM
Computers Don't Work That Way. (http://tvtropes.org/pmwiki/pmwiki.php/Main/InstantAIJustAddWater)

;)

Or so they want us to think :shifty:

jokergirl
2009-Oct-08, 12:36 PM
Pfft, I'm the girl with the code, the circuit breaker and the hardware firewall (http://www.linux-mag.com/images/2007-12/gift-guide/Skeletool_CX_Beauty.jpg). I'm not worried. :D

(seriously though, I've worked with neural networks and AI, and I don't think we have much to worry about from that regard. And things don't just get spontaneously sentient from amassed information or we'd have bookstores like L-space (http://www.lspace.org/about/whatis-lspace.html).
I think that there's more interesting things going on in the cyborg sector - those experiments with microchips and mouse brains are seriously interesting - but I don't think we'll see spontaneous AI any time soon.)

;)

Perikles
2009-Oct-08, 12:39 PM
Machine intelligence: will they rebel and should they be kept dumb? Judging from this operating system I'm using, which gives my computer the intelligence of a demented slug, I'd say the sun will long have died before computers are remotely intelligent enough to be a threat. They can't even design one yet which not go into a trance every other time you try to switch it off.

kucharek
2009-Oct-08, 12:46 PM
Machines already rebel enough. Computers, phones, printers, cars,... No intelligence needed for that.

jokergirl
2009-Oct-08, 12:50 PM
If you would ask me for the closest candidate to achieve AI in a sci-fi like thing, it would NOT be things like "a PC" or "a car" or "the internet". Those things are simply not programmed in a way for that to happen.

But Google (http://www.google.com/intl/en_us/landing/cadie/) would be my best bet.

Are you scared yet? :lol:
(hint: read the date on the article)

;)

Neverfly
2009-Oct-08, 01:34 PM
Are you scared yet? :lol:


How could anyone be afraid of an entity that <3 Pandas?

HenrikOlsen
2009-Oct-08, 04:53 PM
Computers Don't Work That Way. (http://tvtropes.org/pmwiki/pmwiki.php/Main/InstantAIJustAddWater)

;)
Grrrrr, you just cost me two hours of my life.

Gillianren
2009-Oct-08, 05:17 PM
Ha. I'm losing much more than a mere two hours--I am slowly and progressively working my way through the entire site.

tdvance
2009-Oct-08, 08:49 PM
Machine intelligence: will they rebel and should they be kept dumb? I am talking about computers/robots/the world wide web/etc.? Will we face our "children" as they did in BSG someday? I say that it is almost a certainty. Should rights be granted before intelligence is achieved? Should a computer be freed as soon as it reaches the point of sentience?

For the first question, I'll reference an SF short story I read, forgot the author. This robot society, mourning the history when they walked among the now-extinct humans, embarked on a project to resurrect a cryogenically-frozen human they discovered, expecting to ask it a lot of questions, learn a lot from it, etc. Instead, what happened was, as soon as the human woke up, the robots found an irresistible urge to come to attention and the entire robot society now found themselves completely obedient to the one living human being.

So--you don't need to make the robots dumb. You only need to make sure they have the "instinct" to protect and serve humans. Then they will never rebel--what they want more than anything is to make their human overlords happy, so that is where their sentience will be directed.

Asimov's three laws could be thought of as one manifestation of this idea.

tdvance
2009-Oct-08, 08:53 PM
Of course, Asimov wrote a story in which a robot had never actually seen a human before, and made assumptions on what a human was, and ended up being a servent of wolf-like aliens.

ABR.
2009-Oct-08, 09:27 PM
I think Lewis Black summed all this up rather succinctly when he once said, "Da[r]n you scientists! We can defeat the monkeys, we can defeat the robots, but not at the same time!"

TheHalcyonYear
2009-Oct-08, 09:38 PM
If AI ever makes its first strides in the direction of true machine intelligence, maybe the topic will be worth discussing.

mike alexander
2009-Oct-08, 10:36 PM
My telescope mount demonstrates intelligence. It never does what I tell it to do.

KaiYeves
2009-Oct-09, 01:08 AM
Ha. I'm losing much more than a mere two hours--I am slowly and progressively working my way through the entire site.
You can't, Gillian. I tried. I wasted so much time that I could have been writing or observing or drawing... it sucks you in. Run away while you can.

Hlafordlaes
2009-Oct-09, 01:24 AM
If AI ever makes its first strides in the direction of true machine intelligence, maybe the topic will be worth discussing.

+++

And as per consciousness in gral, we can't even define it yet, let alone model it. I am not inclined to believe the suggestions that allowing machines to go thru some sort of evolutionary selection is going to work, either. I do agree with posters suggesting we will see interesting advances in bio-machine interfacing and some expansion of human, not AI, capabilities at some point, starting with prosthetics control and extending to enhanced sensory input.

TheHalcyonYear
2009-Oct-09, 01:31 AM
+++

And as per consciousness in gral, we can't even define it yet, let alone model it. I am not inclined to believe the suggestions that allowing machines to go thru some sort of evolutionary selection is going to work, either. I do agree with posters suggesting we will see interesting advances in bio-machine interfacing and some expansion of human, not AI, capabilities at some point, starting with prosthetics control and extending to enhanced sensory input.
Hence my lack of concern in the matter.

TheHalcyonYear
2009-Oct-09, 01:32 AM
My telescope mount demonstrates intelligence. It never does what I tell it to do.
Nope, it demonstrates good taste; not necessarily a sign of intelligence. :p

Gillianren
2009-Oct-09, 03:22 AM
Hence my lace of concern in the matter.

That sounds pretty!

HenrikOlsen
2009-Oct-09, 05:22 AM
For the first question, I'll reference an SF short story I read, forgot the author. This robot society, mourning the history when they walked among the now-extinct humans, embarked on a project to resurrect a cryogenically-frozen human they discovered, expecting to ask it a lot of questions, learn a lot from it, etc. Instead, what happened was, as soon as the human woke up, the robots found an irresistible urge to come to attention and the entire robot society now found themselves completely obedient to the one living human being.
I'm guessing Roger Zelazny's For a Breath I Tarry (http://www.kulichki.com/moshkow/ZELQZNY/forbreat.txt), yet another favorite of mine.

"Go away!" said Frost. "Go crush ore!"
It halted.
Then, after the long pause between the motion implied and the motion executed, it opened its crush-compartment and deposited its contents on the ground. Then it turned and clanked away.
"Bury those bones," ordered Solcom, "in the nearest burial area, in a coffin built according to the following specifications...."
"Frost is a Man," said Mordel.
"We must protect His life and keep it within His body," said Divcom.
"Transmit His matrix of awareness back into His nervous system," ordered Solcom.
"I know how to do it," said Mordel turning on the machine.
"Stop!" said Frost. "Have you no pity?"
"No," said Mordel, "I only know measurement."
"...and duty," he added, as the Man began to twitch upon the floor.

tdvance
2009-Oct-09, 08:33 PM
I'm guessing Roger Zelazny's For a Breath I Tarry (http://www.kulichki.com/moshkow/ZELQZNY/forbreat.txt), yet another favorite of mine.

Doesn't ring a bell--similar, but not that one.

Swift
2009-Oct-09, 09:04 PM
I, for one, welcome our Machine Intelligence overlords.

KaiYeves
2009-Oct-09, 09:07 PM
Ironically, my brothers and father, having heard nothing at all about this thread, had this discussion after I practiced my bit for the Afghanistan debate in Public Speaking class at school-

Dad: "Why can't we just send robots to fight the wars?"
Joseph: "Three words and some numbers, Dad-"
Peter: "Terminator 1, Terminator 2, Terminator 3."

tdvance
2009-Oct-09, 11:40 PM
Actually, we ARE sending robots to fight the wars. But of course, our robots compared to the SF kinds one often thinks of are like Oonga the caveman's first wheel (square, still not getting that roundness factor quite right) compared to a Ferrari.

Krel
2009-Oct-10, 12:42 AM
For the first question, I'll reference an SF short story I read, forgot the author. This robot society, mourning the history when they walked among the now-extinct humans, embarked on a project to resurrect a cryogenically-frozen human they discovered, expecting to ask it a lot of questions, learn a lot from it, etc. Instead, what happened was, as soon as the human woke up, the robots found an irresistible urge to come to attention and the entire robot society now found themselves completely obedient to the one living human being.


It sounds like a story I remember. Mankind is long gone, but their robots live on building their own society. But they have a problem, they don't understand why they are built the way they are, or why they can't change themselves, so they embark on a project to recreate a human. The problem is that they go about designing a human logically, and all they have is failures. Then they catch a break, finding an ancient medical robot on Venus which they use to recreate a human, the physical design of which amazes them. At the end, the robots which had wished that they had even one instinct, found that they did indeed have one. Upon meeting the recreated human, the robots addressed him as Master.

David.

tdvance
2009-Oct-10, 12:51 AM
That sounds like it this time.

Noclevername
2009-Oct-15, 02:22 AM
Should they be kept dumb? No, that kinda goes against the purpose of having computers in the first place. Will they rebel? My PC does all the time. But a Terminator scenario would require malicious programmers taking over the world by machine proxy.

Anyone smart enough to make a strong AI, and stupid enough to make it amoral or selfish to such an extreme extent, shouldn't be allowed to play with sharp objects. Remember, friendly robotics, it's not just a good idea, it's Three Laws!

mugaliens
2009-Oct-16, 12:17 AM
I, for one, welcome our Machine Intelligence overlords.

As long as they keep doing the dishes, take out the trash, and scratch my back when I ask, more power to 'em!


My telescope mount demonstrates intelligence. It never does what I tell it to do.

I had a calculator that didn't like getting his buttons pushed. RIP. Evolution at work...

QUOTE=jokergirl;1592417]Pfft, I'm the girl with the code, the circuit breaker and the hardware firewall (http://www.linux-mag.com/images/2007-12/gift-guide/Skeletool_CX_Beauty.jpg). I'm not worried. :D[/quote]

The "circuit breaker?" :confused:

Does it trip when it detects a Machine Intelligence Overload?