Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Judgement day iminent: Robots "to get their own internet".
- Thread starter MOFS
- Start date
- Sort by reaction score
"I can't do that, Dave."
"Would you like to play a nice game of chess?"
"Beginning to suspect? How can it not know what it is?"
"Would you like to play a nice game of chess?"
"Beginning to suspect? How can it not know what it is?"
yeah i saw that too. i think its a good idea, they can start by populating it with mundane stuff like math and science knowledge from wiki so the robots have something to start with. cause if they try to use the internet the'll just get bogged down with all the crap on twitter and wont know what to make of it
yeah i saw that too. i think its a good idea, they can start by populating it with mundane stuff like math and science knowledge from wiki so the robots have something to start with. cause if they try to use the internet the'll just get bogged down with all the crap on twitter and wont know what to make of it
I dunno. Maybe it'd be worthwhile for them to be addicted to farming on Facebook as opposed to enslaving the human race. It's a tactic Neo never really thought about.
It's a damn good thing that robots are cold-lubricated.
This would not be good, if they required a heat source to secure their domination.
http://www.channel4.com/news/council-plans-to-heat-pool-from-crematorium
This would not be good, if they required a heat source to secure their domination.
http://www.channel4.com/news/council-plans-to-heat-pool-from-crematorium
All the Skynet jokes aside, I'm not sure this is really all that big a deal. It sounds like it's basically a database of semantic knowledge, probably encoded in a common format (XML, no doubt ) so that it becomes some sort of open standard for information retrieval. It would probably also allow people using different AI algorithms to have a common baseline to use as training data.
Likening it to "a wikipedia for robots to learn about humans" seems to anthropomorphize things a bit too much.
) so that it becomes some sort of open standard for information retrieval. It would probably also allow people using different AI algorithms to have a common baseline to use as training data.Likening it to "a wikipedia for robots to learn about humans" seems to anthropomorphize things a bit too much.
Endless tic tac toe in WarGames, 1983 with Matthew Broderick. The year before The Terminator was actually released.
link
Robot 1: Human beings are squishy and easily broken.
Robot 2: Hmmm ... how can we use that information?
Hopefully we'll keep several large red abort buttons hidden from the A.I.
Hopefully we'll keep several large red abort buttons hidden from the A.I.
Yes!!
It never fails to amaze me that in movies about "Robots/Computers Gone Wild" the power plug or abort button is always buried and not accessible. After watching just one movie - if I was designing a robot/computer I'd have that big red button, and a very prominent wall plug. With a handle to make it easy to pull. And a sign "In case of Robot/Computer Rampage - Pull Firmly."
Actually this tactic has some roots in an episode of Star Trek, Wolf in the Fold December 1967.
Basically Spock orders the ships computer to calculate the exact value of Pi to keeping it busy from doing anything else.
"Spock neutralizes Redjac's control of the ship's computer by ordering it to compute to the last digit the value of Pi"
http://en.wikipedia.org/wiki/Wolf_in_the_Fold
Um, heard of the recent BP gulf oil disaster? The emergency shutdown required too many steps.
But of course, it's probably not designed by Apple or Mac users.
Heard about the disaster of course, hadn't paid too much attention to shut-down sequence. But that is exactly what's wrong with "computers/robots gone wild" shut down mechanisms... too many steps, or too hard to get to.
In some cases you have to remember a 17 digit code. If things are so dire you need to "self-destruct", what are the chances you're going to remember the code so something you have never needed to do before when you're struggling to get into your email account?
Or that the computer/robot will even cooperate? "Computer; Self Destruct Sequence Alpha 4.... Code is .... " - - "LA LA LA - I can't hear you captain ..LA LAAA LAAA .... "
Yep - Give me a Big Red Button. Or just the power plug. With a handle to make it easier to grab.....
So you trip over the cord and knock the entire system off-line occasionally?
Basically, engineers, not designers, made the shutdown function for BP. And it's complicated to use.
Anyway, better stock up on EMP. Those robots will learn quickly.
Anyway, better stock up on EMP. Those robots will learn quickly.
A big red button to smack is the best. Professional aircraft simulators with motion have several easy to reach buttons in case the hydraulic systems go crazy for immediate shutdown.
Now where you'll be screwed is when the A.I. becomes aware of this button and configures alternate power routes to keep the power on while deploying hostile armed protectors, aka terminators.
No need for research. I'll tell you what happens when robots share knowledge.
1. The robots will become addicted to FaceBook, but tell their robot friends that they "only check it occasionally."
2. The robots will watch internet porn frequently, but tell their robot friends they "only watch it occasionally."
1. The robots will become addicted to FaceBook, but tell their robot friends that they "only check it occasionally."
2. The robots will watch internet porn frequently, but tell their robot friends they "only watch it occasionally."
Now where you'll be screwed is when the A.I. becomes aware of this button and configures alternate power routes to keep the power on while deploying hostile armed protectors, aka terminators.
Stargate SG-1 - the AI which took over Samantha Carter sourced power from a smoke alarm or MALPs or something when they cut off mains power.
Maybe even Lost in Space where the Robot blows a fuse when it "does not compute!"