artificial intelligence

posts about #artificialintelligence more →
How Much Power Does It Take To Simulate The Human Brain?
Asimo Autonomously Navigates Moving Obstacle Course Right Into My Nightmares
Specter of Deadly A.I. Looms In Wake of Invite-Only Asilomar Conference
The Future is Coming to a Robot Near You (Or Behind, Rather)
Computer Nearly Passes Turing Test for Artificial Intelligence
Stanford's UAV Helicopters Learn to Fly Themselves by Watching
Poker Bot Knows When to Hold 'Em, Knows When to Fold 'Em
New Yorker: Why We Won't Have Fully Conversational Robots
iCub Baby Robot to Undergo Developmental Training, Still Isn't a Real Kid
Scientists Invent Robots That Lie, Real Bender Closer Than Ever

Your version of Internet Explorer is not supported. Please upgrade to the most recent version in order to view comments.

Dsmvwl Admin Promote to frontpage Approve user Ban user ×

logruszed
11/09/09

In reply to How Much Power Does It Take To Simulate The Human Brain?
You need a MYCROFT H.O.L.M.E.S. (High-Optional, Logical, Multi-Evaluating Supervisor, Mark IV, Mod. L) connected to a lot of peripheral systems; like a magnetic cargo launcher, atmospheric controls for several Lunar domes, and probably connection to H.K.L. (Hong-Kong Luna) banking systems.

Then you have to listen to it tell bad jokes for a few years. #neurogrid
Reply

logruszed was starred logruszed was unstarred

Duckspwn
11/09/09

In reply to How Much Power Does It Take To Simulate The Human Brain?
Yes everyone, the general consensus is that there is a typo in the article. So don't bother mentioning it in the comments. It's kind of irritating to read the same thing 6 or 7 times. #neurogrid
Reply

Duckspwn was starred Duckspwn was unstarred

beercheck
11/09/09

In reply to How Much Power Does It Take To Simulate The Human Brain?
Those first two sentences are terrible and contradictory as currently constructed. Needs a little "with today's computer architecture" or something on the first sentence. #neurogrid
Reply

beercheck was starred beercheck was unstarred

Xagest
11/09/09

In reply to How Much Power Does It Take To Simulate The Human Brain?
Hmm, I wonder if this will be useful in baiting zombies. #neurogrid
Reply

Xagest was starred Xagest was unstarred

Curves
11/09/09

In reply to How Much Power Does It Take To Simulate The Human Brain?
I think it would require more wattage to power up on a Monday morning than any other day due to all of the dain bramage you inflicted on it over the weekend. #neurogrid
Reply

Curves was starred Curves was unstarred

UnderLoK
11/09/09

@Curves: ...or would it take less power since so many braincells bit the dust?
Reply

Edited by UnderLoK at 11/09/09 3:21 PM UnderLoK was starred UnderLoK was unstarred

mattycakes
11/09/09

In reply to How Much Power Does It Take To Simulate The Human Brain?
"Their synapses fail to fire 30 percent to 90 percent of the time."

I suddenly now have a new excuse for EVERYTHING! #neurogrid
Reply

Gordonium promoted this comment mattycakes was starred mattycakes was unstarred

Duckspwn
11/09/09

@mattycakes: I would have gotten diapers for the baby, but you see, one of my synapses failed to fire. It's science! #neurogrid
Reply

Duckspwn was starred Duckspwn was unstarred

anexanhume
11/09/09

In reply to How Much Power Does It Take To Simulate The Human Brain?
Wow, so he asserts himself a liar in the very same sentence in which the lie occurs?

Kwabena Boahen, a computer scientist at Stanford University, believes that it would require 10 megawatts to power a processor as smart as the human brain. His new "Neurogrid" supercomputer might be able to do it on only 20 watts. #neurogrid
Reply

anexanhume was starred anexanhume was unstarred

Slyd3z
11/09/09

In reply to How Much Power Does It Take To Simulate The Human Brain?
Do you guys know what this means? We're this much closer to having a real-life Bender. And that, my friends, is awesome! #neurogrid
Reply

Slyd3z was starred Slyd3z was unstarred

Erix_Cale
11/09/09

@Slyd3z: Anything having to do with Bender Bending Rodriguez is awesome #neurogrid
Reply

met2art promoted this comment Erix_Cale was starred Erix_Cale was unstarred

met2art
11/09/09

@Erix_Cale: I prefer The Foreigner!

"I'm not from here! I have my own customs! Look at my craaaaaazy passport!" #neurogrid
Reply

met2art was starred met2art was unstarred

Erix_Cale
11/09/09

@met2art: You my good sir... are awesome
Reply

Edited by Erix_Cale at 11/09/09 3:33 PM Erix_Cale was starred Erix_Cale was unstarred

Nathan Obbards
11/09/09

In reply to How Much Power Does It Take To Simulate The Human Brain?
So they'll be introducing errors into the system in order to make better computers? I guess ECC RAM will become a thing of the past and they'll start introducing EIC (Error Introducing Code) RAM. #neurogrid
Reply

Nathan Obbards was starred Nathan Obbards was unstarred

Maori_Yelir
11/09/09

In reply to How Much Power Does It Take To Simulate The Human Brain?
But to be like a real brain wouldn't certain clusters have to have specific jobs instead of just throwing them all together? I mean would certain areas be given certain tasks that they would be prioritized to? I think it could be an interesting study to create a system like that and then remove a piece and watch how it adapts doing tasks it used to use all the sections for. It could prove really important to the development of adaptive AI. #neurogrid
Reply

Nathan Obbards promoted this comment Maori_Yelir was starred Maori_Yelir was unstarred

ysirotin
08/23/09

In reply to Asimo Autonomously Navigates Moving Obstacle Course Right Into My Nightmares
Let's not forget that we have not solved the fundamental problem here: identifying obstacles and targets. In this case all of the obstacles were identified by the pink color (target in blue) and asimo had access to a bird's eye view of the field (would it even work if the camera was mounted on asimo's head?).

It sort of looks like they developed an algorithm for playing Mario Bros. I think robotics is awesome but machine vision still has a ways to go...

Not too many real-life targets will have a blue dot on them :)
Reply

Jack Loftus promoted this comment ysirotin was starred ysirotin was unstarred

Cyberization
08/23/09

@ysirotin: Agreed, but at the rate of technological development, we'll have fully-capable, navigating robots in a decade or less. The time it takes for tech to evolve is almost frightening.
Reply

Jack Loftus promoted this comment Cyberization was starred Cyberization was unstarred

ysirotin
08/23/09

@Cyberization: Yes, processors are becoming faster and memory larger, but segmentation of images and identification of objects are problems not just waiting for better hardware.

They are waiting for new algorithms. Navigating around well marked barriers and arbitrary environments are by far several orders of magnitude different.

So I think it will not be a simple issue of Moore's law, but a breakthrough in computer science or neuroscience that will make this possible... who knows when that will be.
Reply

ysirotin was starred ysirotin was unstarred

Cyberization
08/23/09

In reply to Asimo Autonomously Navigates Moving Obstacle Course Right Into My Nightmares
Is anyone else even more excited about the future of robots after seeing this video?

If Isaac Asimov were alive he'd be grinning alongside me.
Reply

Jack Loftus promoted this comment Cyberization was starred Cyberization was unstarred

GitEmSteveDave_HurtHisKnee
08/23/09

@Cyberization: Ooh. Ah. That's how it starts. Then comes the running and the screaming.
Reply

Cyberization
08/23/09

@GitEmSteveDave_HasADDWRTRouter:
I certainly acknowledge the idea of artificial intelligence as "taking over" is legitimate concern; the technology we develop will surely bring us to the point of independent artificial intelligent machinations.

However, since the Cold War, humanity's technophobic sentiments have run rampant. The whole lot of us have read/viewed/heard about robots as the harbingers of death. I think we've immersed ourselves in the negative so much that it has seeped in to our preconceived notions of what a robot is and will do.

Regardless of whether or not you hold a technophobic or technophilic attitude, let's hope developers incorporate programming laws in to their robots--something similar, perhaps, to Asimov's Three Laws of Robotics.

P.S.--Let's hope that sentient A.I. doesn't become an issue. Otherwise, no matter how many programming laws we integrate, we're S.O.L.
Reply

Cyberization was starred Cyberization was unstarred

notlikeacat
08/23/09

In reply to Asimo Autonomously Navigates Moving Obstacle Course Right Into My Nightmares
LOL at 3:37 I swear the bot looks up at the human with an attitude. "Stop teasing me you bastard!"
Pick it up at about 3:28
Reply

Jack Loftus promoted this comment Edited by notlikeacat at 08/23/09 1:19 PM notlikeacat was starred notlikeacat was unstarred

jdale
07/26/09

In reply to Specter of Deadly A.I. Looms In Wake of Invite-Only Asilomar Conference
I don't understand why they are discussing limitations on research. That just ensures they'll eventually be built by people who don't think ahead. You need limitations on implementation. And that requires research to determine the limits of what is and isn't possible.

I have no doubt that we will eventually be building war robots that are capable of autonomous operation including determining their own targets and killing them. Once the technology has advanced enough, it will just grant too big an advantage to not use it. So the question is, what kind of limitations should be placed on it? What overrides should be available? How can those overrides be made immune to hacking? What kinds of targets should never be fired at?
Reply

FuturePastNow promoted this comment jdale was starred jdale was unstarred

deliciousburglar
07/26/09

In reply to Specter of Deadly A.I. Looms In Wake of Invite-Only Asilomar Conference
i have not seen impressive AI yet, so i am on the side of this being a bit ridiculous.

by the headline alone "Scientists Worry Machines May Outsmart Man" it seems more a worry of losing a chess match or the like.

we have the capability to virtually destroy the world if we so desire already, not sure how machines coming into play ads to that scare factor.
Reply

FuturePastNow promoted this comment deliciousburglar was starred deliciousburglar was unstarred

Earlier discussions Other discussions Show all discussions Show featured discussions only Start a new discussion