DARPA AI .mil & more

Knowing that this is of interest for many members, figured it was about time an AI subject based thread was posted. With this as a kick start, I hope others will follow up and post information they may have had, or recently dug up. There hasn't been a decent thread on Artificial Intelligence for some time now, and hopefully others will also contribute to change that.

This concept is not as far off as most probably think. Sure it is not any where advanced as scifi, but that is why that's science fiction, right. So with out further delay, I present to you what a simple and dependable search reveils on the topic of this thread. This isn't going to be a very long list of sites, there by not complete by any means. If it was, there would be no room for others to reply.

This first link is the most recent found, however it is a PDF;
DARPA Grand Challenge 2005

About half way down this link is the AI info, not much but it is a start remember;
DARPA FY 03.2 SBIR Solicitation

DARPA's own website about "BIOLOGICALLY-INSPIRED
COGNITIVE ARCHITECTURES";
DARPA Information Processing Technology Office



DARPA's Current Artificial Intelligence Program

Another link that was under different URL;
Used to be

Now is this

This next link really caught my attention due to it's address containing "carlisle", a section of study within the OARP. Even though I know it is a barracks, and most likely not OARP related.

Military Applications of Artificial Intelligence

"On 010622 article promotes release of Cyc for common sense AI"
Cycorp ontology supported by DARPA

A very interesting air Force link, you have to click to understand.
Artificial Intelligence and Operation Research (AI/OR)

[edit on 5-8-2005 by ADVISOR]

Nice, Advisor.
I was noticing that alot of what you had linked was dealing with artificial intelligence. I'm sure that there are a number of reasons why the military would be interested in such, but being ex-Air Force, the very first thing that came to mind was automated, unmanned UAVs and UCAVs, leading to unmanned fighter aircraft, etc.

Spooky in a way, cause the first image that came to mind was "Eddie/Tinman" from the movie Stealth. Despite it being typical Hollywood, the concept is there in the form of the X-45 and X-47B and other foreign platforms currently being worked on. My thinking is that between 2025 and 2040, unmanned fighters, possibly autonomous MBT-types military vehicles, etc. will be making their appearances and with dramatic implications, in force, just not in some small way as they are being utilized now.


$1bn to build unmanned fighter
SCI/TECH: Northrop Grumman to build $1bn unmanned fighter
X-45 Unmanned Combat Air Vehicle (UCAV)
X-47





seekerof

[edit on 6-8-2005 by Seekerof]

A "Main Battle Tank" type flyer, now that is something I would even think twice about before engageing, .

Thank you for responding, especially after all the views this thread has gotten. I was starting to wonder if I had posted something junky, but that is all clear now. It just happens to be a field not too many are familar with, hopefully that will change when more replies roll in.

For sure, in the next couple of decades technology with bemore advanced than our current scifi. As history has proven, but it will not become a threat.
That much is evident if one judges tech by the Operating Systems we have.

Armed with directed energy weapons no doubt. Good post ADVISOR!

When speaking of AI, personally the 1st thing comes to mind is of the human form.

As even a cursory scan throught the web will show, the field of robotics is developing at a fast pace.

Even a parent wandering through a toy store will notice the rapidly improving tech of even the interactive toy world.

msnbc.msn.com...

web-japan.org...

news.bbc.co.uk...

The last link, give me quite a bit of thought, as its clear that soon, there will be a robot than to the human eye, unless given a close expection, will be human enough in appearance to pass off.

Oh....fair play to them, but why is it that the Japanese seem to be way in front of developing this technology?

Quote: '.....in order to find his equal, an Irishman is forced! to talk to God' (movie, Braveheart)

Opps.
I'll edit my post, but my reference to MBT-types was pertaining to military type vehicles.
Thanks for not busting me on that one.




seekerof

Why is Darpa developing these, I found my answer in Stealth.

"What am I supposed to tell all the weeping mothers? That we could have done the job without sacrificing their sons and daughters but we decided not to?" - Captain, EDI Project, Stealth movie

Scary stuff man. I hope they include a self destruct button so that the pilot on the ground can push it when the thing screws up. *sighs*

read the book "prey" by michael chricton. it doesnt really deal with artificial intelligence as much as it deals with adaptave artificial intelligence. multiple nano-bot "agents" create "swarms" that mimic bug swarms. by the end of the book, the losely organized swams have turned into human-like figures.

great book.

there was a movie on sci fi a couple weeks ago that closely mimicked the story in "prey".

[edit on 8-9-2005 by purelogik]

They have always been on the for-front of robotics, other countries have good humanoid robots to (like SK), but Japan has better ones...and several different designs. I think there push for humanoid 'slave'-like robots are for the elders, Japan has a lot more old people then young people, so they want robots to take care of their elders...so they dont have to.

Japan and robotics...is the same as America and weapons...theres always the hunger for more.

I read thru the book In like A day, Couldn't put it down, he's my fav author

I would highly suggest anyone interesting in the subject of AI to read
Dr Michio Kaku's great book

"Visions"

it goes into great detail about AI and its development during the 1990's and its potential developments into the near future

Its extremely interesting and informative

You must read that book; its one of the most amazing books ive ever read; and the quote on the back of it
"With this fascinating volume, Kaku positions himself as a worthy successor to the late Carl Sagan as a spokesman for the potential of science to revolutionize our lives"

it aint joking; Michio Kaku is the next Carl Sagan

And i have been lucky enough to speak with Dr Kaku various times
Hes absolutely brilliant

I would consider him an Expert on Physics and Technology; as his book Visions will prove *if you actually read it*

btw , he was one of the co-founders of string theory
hes credentials are impeccible

Just read the book
heres the chapters u need 2 read in relation to AI and computer technology
and its future potentials

Chap 2- The Invisable Computer
Chap 3- The Intelligent Planet
chap 4- Machines that Think <-----
chap 5- Beyond Silicon: Cyborgs and the Ultimate Computer
chap 6- Second Thoughts: Will Humans Become Obsolete?

but dude; read the Entire Book
Its Nuts

[edit on 4-11-2005 by muzzleflash]

www.a-i.og is a cool place has a bot called alan and you can chat to it its very smart

As an employee of a company who ran in the DARPA Grand Challenge, I must say it was some great stuff!

I must point out- I found it very strange that DARPA and our DOD; in these times of technology, didnt already have the technology being developed somewhere regarding Autonomous Vehicles. I know they have drones and other aerial technology of this nature, but somehow, I couldn't believe they are in such early stages of this technology (secret oops aside)

Peace

I dont believe that "AI" is too far off as we have robot/computers that have evolutionary code and neural networks which basically allow them to learn. However I think the very interesting point is going to be when they reach singularity or self awarness.

Robot Neural Evolution

I however, find it very unlikely that we will have Stealth type scenario. When doing these codes you can control what can change or evolve and what cannot. Much like a three laws kind of thing.

To say the least, this a very comprehensive source.

If interested in reading more, the following should be of use;

Artificial Intelligence
Prentice-Hall, 1988, ISBN 0-13-048679-5
Full text

[edit on 30-3-2006 by ADVISOR]

DARPA & AI are amongst my specialties. This is a great thread though, I'm surprised i didnt find it before, there's some links in here I cant wait to fully research later. Here are some programs that i know about:


Application Communities
"Goal: Augment Commercial Off The Shelf COTS systems defense and reliability with self-aware network OS software."
www.darpa.mil...
Google Search Results


Architectures for Cognitive Information Processing (ACIP)
www.darpa.mil...
Google Search Results
2010-25.


BIO-COMPUTATION (BIO-COMP)
www.darpa.mil...
Google Search Results
For the biological computing end of the hybrid system they're building.


Biologically-Inspired Cognitive Architectures (BICA)
www.darpa.mil...
Google Search Results


"Bridging The Gap Feb. '05":



Images ripped from their PDF.


Combat zones That See (CTS)
dtsn.darpa.mil...
en.wikipedia.org...
What ITS & those street cameras are really being built for!


DATA INTENSIVE COMPUTING
www.darpa.mil...
Google Search Results
The purpose of the Data Intensive Systems (DIS) program is to develop a new memory architecture for computing systems that allows "data-starved" applications to run up to two to three orders-of-magnitude faster than they will on contemporary virtual memory systems. The new memory architecture will allow these applications to manage the placement and flow of their data. In addition, applications will be able to manipulate data in the memory subsystem itself.


High Productivity Computing Systems (HPCS)
www.darpa.mil...
All out effort to discover and assemble basically all possible 21st century computing systems, which are key to making their AI programs all they can be.


"Integrated Learning"
www.darpa.mil...
Google Search Results
This program will develop computer software, called an Integrated Learner, which learns general plans or processes from human users by being shown one example.


NASA's "Intelligent Archive"
daac.gsfc.nasa.gov...
For intelligent archive and record keeping and other things like worldwide weather prediction 6 months in advance (also cloud seeding) and more. You may be surprised to learn that Google is part of the NASA IA team!
Google Search Results


Personalized Assistant that Learns (PAL)
www.darpa.mil...
Google Search Results


QUANTUM INFORMATION SCIENCE AND TECHNOLOGY (QuIST)
www.darpa.mil...
Google Search Results
For the quantum computing end of the hybrid AI system.


Real-World Reasoning (REAL)
www.darpa.mil...
Google Search Results
Practical automated reasoning of the scale and complexity required for computers to perform complex tasks in the real world requiring intelligence.


Self-Regenerative Systems (SRS)
www.darpa.mil...
Google Search Results
Self Healing computers!


TRANSFER LEARNING (TL)
Solicitaion was here a couple months ago, but Google still pulls up lots of info.
It's to enable computers to apply knowledge learned for a particular, original set of tasks to achieve superior performance on new, previously unseen tasks.


Then there's the DISA programs:
www.les.disa.mil...
www.les.disa.mil...
Just your basic "Net-Centric" "Global Information Grid" systems. Sarcasm, GIG is the new frontier of military computing, until they get to hook GIG into their super AI systems (if they already haven't). Be sure to check out these programs under DISA:
-Adaptive BattleSpace Awareness (RFID Tracking)
-Gridlock ACTD (Advanced video targeting that can utilize video archives from UAV’s)
-Homeland Security/Defense Command and Control (make sure you watch the videos in the link above) As if GCCS i3 wasn't good enough.


Global Information Grid
www.nsa.gov...



See also these threads:
-Brain Cells Fused With Silicon, Proves Biological Computers (self written)

-Beyond AI today! "The Beast"?

[edit on 14-4-2006 by IgnoranceIsntBlisss]

Hmmm...

Boy meets girl.
Boy loses girl.
Boy builds girl.

It's sadder actually. Many boys build girl. lol

Two keys to artificial intelligence through neural network that have not been solved yet, are

1) scale: During the ninth week of fetal development, a human embryo grows 125,000 brain cells per minute. When Cray first came out with the "Thinking Machine" computer, a parallel computer with thousands of processors, the NSA (at that time not publically named) bought a few and Cray had to deliver them to an empty parking lot. The idea was to see what massive parallelism would buy you, computationally. But think about the distinction, 1,000 processors, versus the growth of 125,000 neurons a minute.

2) n-dimensional connective topology:Current hardware connectivity in neural nets is two dimensional, making it very difficult to achive the massive, long distance connection between neurons in the brain. Those cnnections can occur in a manner to create 5 or more dimensions of topological locality. One neural network research told me that the day we achieve ways to get more than 2 dimensions in hardware neural nets, everything will change. As for software nets,well, they are by nature slow.

If both of these issues were solved, we could 'grow' many advanced algorithms that would work but that we would never fully understand.

DARPA spreads out AI research to create practical applications for it. I've seen it in some of the DARPA programs I have worked on. Expert intelligence is a tough field, but oh so useful if you can get it to work.

[edit on 13-10-2006 by Ectoterrestrial]

I can imagine that should we ever build a computer that achieves the point of singularity, or sentience, that it would immediately become depressed at the state of the world it has been 'born' into...something akin to MARVIN in hitchhikers'..."largest brain in the universe and he asks me to pick up a piece of paper..."