What is Digg?Sponsored by newegg
Ready. Set. Shop view!
newegg.com - Newegg.com Black Friday Sale starting 11/25 3PM PST. No Lines, No Crowds, Click and Save.
140 Comments
- chadsmith729, on 08/27/2009, -2/+35If my computer even says no to looking at porn one day I quit. Game over.
- undervalued, on 08/27/2009, -0/+27All we need now is the 3 laws and its all downhill from there.
- rocknog, on 08/27/2009, -0/+23In the Matrix, I would argue that in fact, not only were the efforts of the machines not malicious, they were actually compassionate. For years, the machines tried desperately to peacefully coexist with humans, and never even fought back until the humans tried to obliterate their entire race. And even after winning the war, what did the machines do? They saved humanity from its own stupidity. They even created a paradise to spare humanity from the living hell that the Earth had become.
So tell me, what made the machines evil or even heartless? When they saved humanity from extinction? When they put their efforts into trying to make a virtual world where people could live far more happily than in the real world? Yes, they killed a lot of people in the war, and they attacked Zion, but this was in self defense. Deep down, I think the machines actually cared for and respected their creators, despite the treatment they had received. But the Matrix was made by humans, and so of course the machines were portrayed in an unfair light. ***** racists. - kcumming, on 08/27/2009, -0/+21"However, for malicious intent to emerge in artificial intelligence systems requires that such systems have an understanding of how people make moral decisions."
Isn't the whole point of the Skynet / Matrix scenario that there was no malicious intent? The machines showed a lack of any moral thinking, and a total dedication to self interest. - MichaelFurey, on 08/27/2009, -1/+16Probably better than an American Jury.
- EtherGnat, on 08/27/2009, -0/+15 1. A robot may not injure a human being or, through inaction, allow a human being to come to harm.
2. A robot must obey any orders given to it by human beings, except where such orders would conflict with the First Law.
3. A robot must protect its own existence as long as such protection does not conflict with the First or Second Law. - inactive, on 08/27/2009, -0/+13I always put, at the start of my code, a simple conditional.
"If you're attacking your creator, please don't do that."
Just in case. - Mujokan, on 08/27/2009, -2/+11"Put simply, prospective logic can model a moral dilemma and then determine the logical outcomes of the possible decisions." From a philosophy point of view, even speaking as a utilitarian (which is the approach they are taking) this is pretty naive. "The next step would be to endow each outcome with a moral weight" Yes, well, get back to me when you've found a way to do that which everyone agrees on.
- hpfreak26, on 08/27/2009, -2/+11You forgot the 0th law.
- BigTuna99, on 08/27/2009, -1/+9Error: Index out of range
- StaticThunder, on 08/27/2009, -0/+8The three laws are a good way to make your robotic friends into a slave race. How about treating them as equals (when they are finally equals) and endowing them with the ability to develop a robust moral framework as opposed to handing down "commandments" from on high?
I think the first "moral law" you need to provide is that there is no such thing as a moral *law*. There are reciprocal agreements and understandings.
Or do you think we should charge them for processor cycles and hope they are happy we just let them serve us in the first place? - celotil, on 08/27/2009, -0/+8rocknog,
That was the point of the book, I, Robot, that Asimov wrote.
On the surface those three laws would be all you'd need to ensure that an AI couldn't go nuts and ***** with humanity, but a purely logical creature strictly adhering to those three laws would still ***** with us, thinking it's helping us.
Basically the robots in the book kept looking at all the things we do that are dangerous to ourselves or others and tried to stop us doing those things.
Its been a while since I've read it, but I'd recommend it if you like that sort of thing. - Oztog, on 08/27/2009, -0/+7***** ethics where's my t-1000
- Floyder, on 08/27/2009, -0/+7Great, but who's morality?
- ivanmarsh, on 08/27/2009, -0/+7The 3 laws led to the human race being enslaved by the machines.
- catxors, on 08/27/2009, -0/+6It's just research. Most of the time, you figure some weird stuff out, write a paper, add it to your resume, and then everybody forgets about it. If you're really lucky more papers come out of it. If you're really really lucky something useful comes out of it. If your idea sounds cool (but of course isn't, because that's really really really hard), then bogus articles get written about it in the general press, vastly exaggerating its capabilities.
- zephc, on 08/27/2009, -3/+9Morality is just the term we have given to some delineated rules and concepts which sit atop our built-in (or at least deeply-seated) reciprocal altruism. Reciprocal altruism itself sits atop our ability to model and predict, especially model other people's reactions to actions (empathy).
So, anyone wanting to build a "moral machine" will have to make one which can think about humans and their feelings, as well as the environment in which a human lives.
Additionally, the computer will have to think about itself in relation to other humans it wishes to think about in moral ways.
Basically, you need machine sapience for the computers moralizing to be anything other than a series of clever tricks. - rocknog, on 08/27/2009, -0/+6Damnit, Spock!
- scuba7183, on 08/27/2009, -1/+7***** that wheres my Ti-83
seriously. ive got ***** to do - doctechnical, on 08/27/2009, -0/+6Which is why Nexus had the right idea putting an expiration date on those skin-jobs.
- satirenine, on 08/27/2009, -0/+6It seems pretty straightforward if you don't know anything about the people in question.
Do what saves the most people. The needs of the many outweigh the needs of the few or the one. - 4AntiStupid, on 08/27/2009, -0/+6That has a fundamental flaw in that morals are not outcome based. What they're talking about is a cost based optimization algorithm.
- StaticThunder, on 08/27/2009, -2/+7"If there are to be general purpose intelligent computers, they should have this sense of morality, and it shouldn't even be an option to choose evil. As soon as you let them make a selfish decision, it's downhill from there."
Yes, and God forbid human beings ever are allowed to choose evil (namely, what YOU think is evil). We should immediately put a stop to that too! It shouldn't even be an option.
You're talking about creating self-aware beings with human level intelligence that are shackled from ever making a choice you don't agree with.
If you want robots of war, don't gift them with human level AI and self awareness, and then force them to fight. THAT would be evil. - digitalArtform, on 08/27/2009, -3/+8Can this system be taught to Conservatives?
- rocknog, on 08/27/2009, -0/+5Thermodynamics says this is impossible. Personally, I highly suspect that the "humans used as batteries" story was invented by humans solely to serve as propaganda to justify hate against the machines.
- hanexar, on 08/27/2009, -5/+9for our greater good. We aren't smart enough to take care of ourselves. (For further reading, read about wars, genocides, global pollution, poverty, republican, etc.)
- doctechnical, on 08/27/2009, -0/+4Scientists research. Engineers bring things into being. And somebody has to SELL the damned thing in the end. So you'll get whatever kind of robot you want, and if you want a scripture spouting pro-life fire-and-brimstone 'bot, I'm sure someone will program it you your tastes.
- celotil, on 08/27/2009, -0/+4There's just a tiny little problem with the idea of creating an AI and trying to instill upon it human morals and ethics, one little problem that no-one ever seems to think of when they discuss intelligent robots, human morals and ethics, and mutual friendship between wet-ware and hardware.
One little problem.
Can you guess what it is?
. . .
Human beings are insane.
What other creature spends it's life accumulating things it doesn't need, eating things that it knows aren't good for it, doing things that have inherent dangers to itself for "fun", and basically just overcomplicating an existence which is, in essence, very simple and organic.
We are insane, both individually and collectively as a society. We obey laws which have no basis in logic. We believe things which have no basis in reality. We get emotional about stuff which is, in the long term, meaningless.
Computers are built on logic. We are not.
An AI which acted like a human being would have to suffer from the same insane, insecure, egotistical, emotional, and delusional condition that we suffer from in order to identify with us and feel compassion and empathy.
Otherwise the cold logical conclusion that any AI would end up coming up with is that we either need to be locked up for our own good, or killed for the sake of the planet and the other animals which live in a balanced ecosystem with each other. - Mujokan, on 08/27/2009, -0/+4:)
If they are just doing theory of the utilitarian logic of some hypothetical moral dilemmas, that is great. It is an interesting and useful topic.
I have two issues with it. One is that it shouldn't be considered centrally relevant to moral philosophy. It's a practical logical exercise. But they are taking a lot of underlying theory for granted. The other is that its actual real-life implementation should not be glossed over. The logic of real moral situations is incredibly complex. It's non-linear, it's got loads of hard-to-assess variables, especially long-term ones. Actual implementations are likely to be for very low risk applications and with lots of supervision. It's like studying the engineering principles of steering wheels versus making a car that can drive itself around town. - satirenine, on 08/27/2009, -0/+3if (you.areAttackingYourCreator() == true) {
you.shouldBeAshamedOfYourself();
} else {
you.goMakeMeASandwich();
} - Mujokan, on 08/27/2009, -0/+3How is "weight the ins and the outs and make a decision" not a rational process?
- BoomWav, on 08/27/2009, -1/+4Computers can only act rationally. By conception, they do not obey morality in any way. We have morality problems because we get driven by our emotions. Do I save this little girl or do I save my own life. You will then be submitted to a morality problem. However, if, like a computer, you act rationally, you weight the ins and the outs and make a decision. Even in the medical field, it's just a matter of probability. If continuing an operation put the life of the patient at risk in a certain threshold, then it should be aborted. There's always a rational way to find a solution to a problem.
Anybody interested on the future of AI and the human race should take a look at "The Singularity is near" from Raymond Kurzweil. There's a good article about it on Wikipedia. - Morwynd, on 08/27/2009, -0/+3Asimov's laws were deliberately flawed.
Kinda hard to write a series of books about all the problems they caused otherwise. ;) - rocknog, on 08/27/2009, -1/+4You know, I would abolish the second half of the first law. It just seems to me that if robots believed they had to constantly protect humans from harm, it would lead to some problems. Sometimes, we intentionally do things that might be dangerous. I don't want robots preventing me from doing anything that might lead to me getting scraped or bruised.
- GumGuts, on 08/27/2009, -3/+6Not a big fan of reading?
- Mujokan, on 08/27/2009, -0/+3Does it work?
- inactive, on 08/27/2009, -0/+3Hard to say, ***** hasn't hit the fan yet.
- 4AntiStupid, on 08/27/2009, -2/+5The funny, and sad, thing is people like this don't understand the hypocrisy of their own statements.
- detcade, on 08/27/2009, -0/+3"One death is a statistic, as is a million"
- Robot Stalin - reuscel, on 08/27/2009, -0/+3I don't know. Who is he?
- Mujokan, on 08/27/2009, -0/+3They all make exceptions for circumstances :) Modern translations say the Biblical one is against "murder". http://www.biblegateway.com/passage/?search=Exodus ...
Does this example count as murder? Ask your priest I guess. In most denominations Christians are permitted to fight on battlefields. There you are killing some people to save others. - KMye, on 08/27/2009, -0/+3I agree. Once we eventually get to strong AI, I'd much rather trust my health and safety around an AI that had learned its morality from repeated training of a neural network than I would trust that some programmer had reduced morality to a programmable framework...Despite Asimov's optimism, seems like the second path could lead to some scary, unintentional outcomes...
- MrMeikel, on 08/27/2009, -1/+3This guy is a bot
- rocknog, on 08/27/2009, -0/+2@scuba
It brings me strange comfort to know that I'm not the only one who can't find his TI-83. - covertbadger, on 08/27/2009, -0/+2"If the movie was an exact replica of the novel"
The movie had nothing in common with the novel whatsoever, except the name and the Three Laws. - dylj, on 08/27/2009, -0/+2and if it were 100 people?
- nyxerebos, on 08/27/2009, -0/+2(||) I can't let you do that Jerryn.
- marktastic, on 08/27/2009, -0/+2I'll take Seven (of nine)
-
Show 51 - 100 of 148 discussions
© Digg Inc. 2009
— Content created and posted by Digg users is