Ha, awesome!Andronicus said:
On the actual topic...this sounds like Jackass, really. Especially as other scientists are essentially calling the project out for having little value.
Robot Punches People to Learn Asimov's Laws
- Thread starter Logan Westbrook
- Start date
On the actual topic...this sounds like Jackass, really. Especially as other scientists are essentially calling the project out for having little value.
I certainly hope we aren't trying to teach robots to follow Asimov's Laws. They don't work very well and are fairly easy to bypass.
But...it's a bot. If it wants to kill us, it can simply, you know, affix it's screwdriver attachment and shank a *****. Or, use it's robotic arms to lift us up and weld metal to our flesh. Most of us die after something like that.
Alrighty...well I must say this once. We're not ready. We're not ready for AI, We're not ready for sentient robots, We're not even ready to tolerate each other. People do some nasty things to other people for the flimiest of reasons. If we cannot act humanely to our fellow man how can we expect our creations to do the same?
All of those planning to defect to Skynet's side and becoming cyborgs with human brains, get in line behind me :-D
..."Beep boop"? Great. Bad enough that the robots are going to be beating us all into a thin gooey paste... you're saying they're going to be doing it for achievement points.Andronicus said:
I am impressed to see my countrymen doing something awesome enough to be mentioned on the Escapist!
That said, types of injuries might be a bit smarter thing to test...but then again can't have live test subjects used for that, now can you?
That said, types of injuries might be a bit smarter thing to test...but then again can't have live test subjects used for that, now can you?
Not to be a killjoy but surely this study is a somewhat unnecessary use of funds given that we can already ascertain what mass "x" moving at velocity "y" will do to the human anatomy. I saw in a TV show I think, C.S.I. or somesuch?
Surely, robots that could interact with humans in this way could more effeciently be given sensors and processors to avoid contact altogether and still work at freakish speed?
As for the laws, Asimov did theorise his own issues with the laws, so robots walking around in the day to day would need proper programming, especially those interacting with us on a "personal basis". Just look at "Moon", although:
Gotta love them robots...
Surely, robots that could interact with humans in this way could more effeciently be given sensors and processors to avoid contact altogether and still work at freakish speed?
As for the laws, Asimov did theorise his own issues with the laws, so robots walking around in the day to day would need proper programming, especially those interacting with us on a "personal basis". Just look at "Moon", although:
How did GERTY ever manage to tell Sam the truth of the situation? I can predict your answer "He did it for Sam's wellbeing" surely when he was being programmed, number one protocol must have been "Never allow a Sam to discover the truth", but then of course it wouldn't have been much of a film
Gotta love them robots...
If a few people have to get punched so that the rest of us can someday enjoy sex-bots, then punch on, robots; punch on!
Actually there is a reason why they don't make sence. It's because Asimov never ment for them to be taken seriously. If you read some of the various Robot series and the Detective series Asimov pretty much shows that the laws are easily broken. Infact most people don't realise there are actually 4 laws not 3.Retodon8 said:
I could just see CS professors wanting to extend this research indefinitely.
"I'm going to be offering extra credit for anybody who's willing to fill out a brief survey on an AI enter-face that we're testing. All you have to do is run the program then rate it from painless to unbearable."
"I'm going to be offering extra credit for anybody who's willing to fill out a brief survey on an AI enter-face that we're testing. All you have to do is run the program then rate it from painless to unbearable."
Wait, what? Achievement points?The Rogue Wolf said:
When did I say they'd be doing it for achievement points?
They make perfect sense for machines. For sentient beings no, but that's not what we have and it's not what a machine would need to asses which rules were in play.Retodon8 said:
Dangerous machine must be safe to be around and not require human interference to be so. For people to accept dangerous machines outside of factories then they need to be sure that they won't cause harm.
Machine must serve purpose. We'd want to be sure anything we ask a machine to do will be done and this is effectively on the fly programming.
Bloody expensive machine must not allow it self to be accidentally destroyed. No-one likes wasting money just because the machine was stupid enough to get broken.
even for sentient beings those are staples of most social rules.
first law - don't hurt other people. sometimes followed with - unless they deserve it, *rolleyes*
second law - be socially productive. In human society this is regulated by currency, not being productive means no food and no food means being dead. not every nation has support for the poor. In this case Asimov's wording it's literally - be a slave so in that way not appropriate.
third law - don't commit suicide or self harm. Again these are technically crimes in most countries whether they are enforced or not. Although some view suicide as a preferable option to failure or loss of social usefulness by and large they are seen as bad things.
The real difference is that sentient beings break the rules all the time, which was the point of Asimov's robot psychologist stories. The stories machine became progressively more intelligent and started to have problems with the rules. Even the lower end machines could still hit logic problems with them though but that's a matter of implementation not the rules themselves.
