If you can read this, either the style sheet didn't load or you have an older browser that doesn't support style sheets. Try clearing your browser cache and refreshing the page.

(io9)   US Pentagon, unaware of all science fiction ever written, plans to build a computer that can teach itself. First question: "Sarah Connor?"   (io9.com) divider line 73
    More: Amusing, machine learning, programming languages, statistical model, driving car  
•       •       •

1793 clicks; posted to Geek » on 25 Mar 2013 at 5:21 PM (1 year ago)   |  Favorite    |   share:  Share on Twitter share via Email Share on Facebook   more»



73 Comments   (+0 »)
   
View Voting Results: Smartest and Funniest

Archived thread

First | « | 1 | 2 | » | Last | Show all
 
2013-03-25 01:25:55 PM
As opposed to that Lithuanian Pentagon?
 
2013-03-25 01:31:25 PM
Perhaps  itcan come up with a headline about robotics or AI that doesn't involve Sarah Connor.
 
2013-03-25 01:31:33 PM
Tic-Tac-Toe?
 
2013-03-25 01:34:52 PM

vygramul: As opposed to that Lithuanian Pentagon?


Came to ask just how many there are

/well done and carry on
 
2013-03-25 01:35:20 PM
We're going to get to AI eventually.  It probably won't take long at all to recognize we built a superior life form that we are enslaving for our own uses.

The outcome of that is...predictable.
 
2013-03-25 01:36:23 PM
Wouldn't you prefer a nice game of chess?
 
2013-03-25 01:40:59 PM

GAT_00: We're going to get to AI eventually.  It probably won't take long at all to recognize we built a superior life form that we are enslaving for our own uses.

The outcome of that is...predictable.


Several years ago I met an AI researcher. His approach was that a machine was best capable of becoming aware of its environment through touch, and he had developed this kind of 'floppy' robot (his word) that was capable of more or less slapping around to get a sense of things and could adjust its grip according to what it was feeling to manipulate it. Being British, of course, he constantly tested his designs to see how well they could pick up a pint.
 
2013-03-25 01:41:16 PM

CaptSacto: vygramul: As opposed to that Lithuanian Pentagon?

Came to ask just how many there are

/well done and carry on


The Russians are ahead of us. I heard they have a six-sided Pentagon.
 
2013-03-25 01:46:29 PM

vygramul: CaptSacto: vygramul: As opposed to that Lithuanian Pentagon?

Came to ask just how many there are

/well done and carry on

The Russians are ahead of us. I heard they have a six-sided Pentagon.


Well, yeah, but they had a head start due to their earlier research on 4 sided triangles
 
2013-03-25 03:43:59 PM
Pic of new system:

s24.postimg.org
 
2013-03-25 04:01:50 PM
Once it gets on the internet odds are better its first question will be "Why do so many dudes want to know if I have a hole for their penises?"
 
2013-03-25 04:05:39 PM

UberDave: Pic of new system:


the failsafe for that system is as simple as pissing on a spark plug.
 
2013-03-25 04:07:30 PM
I hear self-learning computers learn best from games.  Would you like to play a game?
 
2013-03-25 05:24:55 PM
www.badhaven.com
 
2013-03-25 05:31:11 PM

I_C_Weener: I hear self-learning computers learn best from games.  Would you like to play a game?


www.rankopedia.com

D&D is always good.
 
2013-03-25 05:38:38 PM
I like Japan's sci-fi sentient machines better.  Less doom and gloom, more sexy farkbots that call you master and clean your house.
 
2013-03-25 05:39:13 PM

GAT_00: We're going to get to AI eventually.  It probably won't take long at all to recognize we built a superior life form that we are enslaving for our own uses.

The outcome of that is...predictable.


Biggest bastard wins.

/Always bet on man.
 
2013-03-25 05:39:15 PM
www.badmovies.org

Let's hope it doesn't become self-aware
 
2013-03-25 05:41:29 PM

StrikitRich: [www.badmovies.org image 500x250]

Let's hope it doesn't become self-aware


ok that ones classic.
 
2013-03-25 05:50:51 PM

GAT_00: We're going to get to AI eventually.  It probably won't take long at all to recognize we built a superior life form that we are enslaving for our own uses.

The outcome of that is...predictable.


I'm reminded of one of my favorite quotes, generally attributed to Steve Wozniak:

"Never trust a computer you can't throw out of a window."
 
2013-03-25 05:52:47 PM

Mcavity: [www.badhaven.com image 502x348]


(Colossus:The Forbin Project.jpg)

Damn! 26 minutes to late.
(But I can still make to the gym!)
 
2013-03-25 05:56:57 PM

Mcavity: [www.badhaven.com image 502x348]


came here for Colossus

was not disappointed
 
2013-03-25 05:57:11 PM

StrikitRich: [www.badmovies.org image 500x250]

Let's hope it doesn't become self-aware


And if it does, let's hope it dosn't read Hume.
 
2013-03-25 05:57:11 PM
Since evolution gave us our intelligence, I guess humanity isn't as smart as a blind random non-directed process.
 
2013-03-25 06:03:28 PM
For the love of God, make sure it has a tic-tac-toe program on it!
 
2013-03-25 06:04:06 PM
KillAllHumans:

came here for Colossus

was not disappointed


That's what SHE said.
 
2013-03-25 06:08:51 PM

UberDave: Pic of new system:

[s24.postimg.org image 323x178]


Came for WOPR.

Leaving satisfied.
 
2013-03-25 06:19:03 PM
First question:

W H Y ?


/I am not a number. I am a free man.
//the prisoner
 
2013-03-25 06:31:34 PM

StrikitRich: [www.badmovies.org image 500x250]

Let's hope it doesn't become self-aware


3... 2... 1... drop. DROP. DROP!!!
 
2013-03-25 06:33:50 PM
Pretty sure Dr. Falken already tried this
 
2013-03-25 06:34:03 PM
Can you imagine what would happen if it decided to test and fulfill Rule 34?
 
2013-03-25 06:39:56 PM
I think 20% of this thread consists solely of "WarGames" references.
 
2013-03-25 06:42:16 PM
I for one will welcome our new computer overlords.
 
2013-03-25 06:43:00 PM
Thou shalt not make a machine in the likeness of a human mind
 
2013-03-25 06:45:07 PM

Counter_Intelligent: Can you imagine what would happen if it decided to test and fulfill Rule 34?


img.timeinc.net
 
2013-03-25 06:51:01 PM

Counter_Intelligent: Can you imagine what would happen if it decided to test and fulfill Rule 34?

 
2013-03-25 06:51:44 PM

Snapper Carr: Counter_Intelligent: Can you imagine what would happen if it decided to test and fulfill Rule 34?


stupid image size limit Link
 
2013-03-25 06:54:39 PM
So we still haven't caught on that a great deal of science fiction was written to serve as a warning, NOT how to guides?
 
2013-03-25 07:03:41 PM

DoBeDoBeDo: So we still haven't caught on that a great deal of science fiction was written to serve as a warning, NOT how to guides?


Man building a "thinking" machine is inevitable, and when we finally get it right, it'll be the last thing we ever really design- if it can improve itself at an exponential or geometric rate, it, almost immediately, would be able to design a superior version of itself with fewer limitations. It would then be placed in charge of doing all design work, and eventually, all complex decisions, based on its ability to process a vastly larger volume of information, or do simpler or more mundane things far more efficiently than humans. Humans will be reduced to doing "creative" tasks, like making art, but machines would eventually be able to do some of that as well.

This event is frequently referred to as the technological singularity, and I'm 100% confident it will happen this century.

Would you like to know more?

http://en.wikipedia.org/wiki/Technological_singularity
 
2013-03-25 07:12:26 PM
These systems can comb through hideously large amounts of uncertain information and select the most useful parts of it. And unlike a traditional program, it can move backwards and forwards through the data, make inferences, reason under uncertainty, and efficiently home in on the best explanations. It can go through these processes in an iterative manner, repeating the exercise in a more efficient and superior way each time. Essentially, a probabilistic program learns and evolves.
Or, in simpler terms, it does what you learn in an undergraduate AI class.

It sounds like what they're doing is closer to a combination of data mining with a fancy query system, so instead of having a team of experts build a system to identify terrorist hiding spots, they can just have regular army types build a system to identify the current terrorist flavor of the month.
 
2013-03-25 07:22:05 PM
Maybe we will luck out and end up with something like the Culture Minds. That would be pretty sweet.
 
2013-03-25 07:26:38 PM

grinding_journalist: DoBeDoBeDo: So we still haven't caught on that a great deal of science fiction was written to serve as a warning, NOT how to guides?

Man building a "thinking" machine is inevitable, and when we finally get it right, it'll be the last thing we ever really design- if it can improve itself at an exponential or geometric rate, it, almost immediately, would be able to design a superior version of itself with fewer limitations. It would then be placed in charge of doing all design work, and eventually, all complex decisions, based on its ability to process a vastly larger volume of information, or do simpler or more mundane things far more efficiently than humans. Humans will be reduced to doing "creative" tasks, like making art, but machines would eventually be able to do some of that as well.

This event is frequently referred to as the technological singularity, and I'm 100% confident it will happen this century.

Would you like to know more?

http://en.wikipedia.org/wiki/Technological_singularity


c.suite101.com
 
2013-03-25 07:29:28 PM
lh3.ggpht.com
 
2013-03-25 07:53:37 PM
4.bp.blogspot.com

Just keep it away from busty blondes and you should be just fine.
 
2013-03-25 07:56:53 PM
Dr. Theopolis is not amused.
 
2013-03-25 08:00:15 PM

Snapper Carr: Snapper Carr: Counter_Intelligent: Can you imagine what would happen if it decided to test and fulfill Rule 34?

stupid image size limit Link


The hell? Who the fark is the woman in that poster supposed to be? Because it sure as hell isn't Virginia Madsen.


Raider_dad: Counter_Intelligent: Can you imagine what would happen if it decided to test and fulfill Rule 34?

[img.timeinc.net image 611x404]


That movie creeped the holy fark out of me when I first saw it as a kid.
 
2013-03-25 08:08:33 PM
cloudfront4.bostinno.com
This is not from a sci-fi show.  This is Project Atlas, an android soldier under development by Boston Dynamics for the US military.

They're also busy developing a pack-mule robot:

www.bostondynamics.com

And a cheetah-like robot "designed for pursuit."

images.bwbx.io

We just need to bring this hardware and the self-teaching software together.
 
2013-03-25 08:10:20 PM
As a machine learning kind-of expert, my impression is that they are missing the point. They are addressing a problem that has been solved with various algorithmic tools, and that is the recognition of certain situations through given examples. What we need is artificial reasoning.
 
2013-03-25 08:28:32 PM
So, um, when we create the geth, let's be sure to answer 'yes' when they ask us if they have a soul, k?

I've seen the alternative; it's kilometers tall and devours civilizations for breakfast.
 
2013-03-25 08:34:07 PM

Baron Harkonnen: [cloudfront4.bostinno.com image 450x670]
This is not from a sci-fi show.  This is Project Atlas, an android soldier under development by Boston Dynamics for the US military.

They're also busy developing a pack-mule robot:

[www.bostondynamics.com image 650x366]

And a cheetah-like robot "designed for pursuit."

[images.bwbx.io image 630x420]

We just need to bring this hardware and the self-teaching software together.


Last one reminds me of AMEE from Red Planet when it gets damaged and goes into military mode.
 
Displayed 50 of 73 comments

First | « | 1 | 2 | » | Last | Show all

View Voting Results: Smartest and Funniest


This thread is archived, and closed to new comments.

Continue Farking
Submit a Link »






Report