Military Robotics….Deep in the Singularity Zone
I’m as big a fan of technofuturistic science as the next reader of Danger Room but National Defense Magazine ‘s article really is breezily optimistic:
Reverse Engineering the Brain May Accelerate Robotics Research
….Machines that walk upright will assist civilians and the military alike, said Stefan Schaal, associate professor of computer science and neuroscience at the University of Southern California.“We should at some point be able to create an artificial human being and I think humanoid robots are currently the first step toward that,” he said at the Army Science conference.“This is going to happen,” he predicted. “And it’s going to happen in this century.”It may not be as “polished” as the iRobot movie, he added.While other experts noted that there are huge technological hurdles to overcome, basic research continues on several critical technologies such as vision, movement and computational models that will allow robots to “think” like humans.
A parallel effort to map – or reverse engineer – the human brain is going to give robotics experts inspiration that will allow them to create these advanced models, researchers at the conference said.The National Academy of Engineering is spearheading this “Grand Challenge.” Just as researchers successfully mapped the human genome earlier in the decade, the engineering community – not normally thought of as being a part of the life science discipline – says there will be a clear benefit to a Herculean effort to figure out exactly how the human mind works.“If we could determine the software of the human brain, we could embed all sorts of systems so as to provide human like quality for machines,” said John Parmentola, director of research and laboratory management at the Army office of the deputy assistant secretary for research and technology.Neural models will enable robots to better perceive, think, plan and act, said James Albus of the Krasnow Institute at George Mason University, Va.
“Significant economic and military applications will develop undoubtedly early in this century and in fact are already developing,” he said.
Read the rest here.
The part that makes me a tad skeptical is the “reverse engineering” of the brain. This is no small task. “Wetware” isn’t hardware and the wetware here is dynamically adaptive and to an extent individualized within parameters we do not yet fully understand. Unless I am missing something ( please correct me if I am) in terms of difficulty, reverse engineering the brain would appear to be harder than almost any other question that could possibly be related to the whole field of robotics itself.
While scientists have learned more about the human brain in the last 10 years that the previous 10,000, brain science is still in it’s infancy. The exciting MRI scan studies are primarily exercises in positively identifying correlation of brain activity with specific cognitive and physical tasks; what these studies mean in terms of application requires extrapolative speculation and experimentation.
By all means guys, go for it! I’m behind the effort 100 % as the spillover benefits are going to be enormous. However, I’d wager that this strategy is not the fastest route to functionally useful, autonomously acting, robots on a societal scale.
ADDENDUM:
I just picked up P.W. Singer ‘s new book Wired for War: The Robotics Revolution and Conflict in the 21st Century
ADDENDUM II.
Jeff Hawkins at TED.com on the revolutionary potential of brain science:
February 18th, 2009 at 2:02 pm
Absolutely right that it’s easier said that done!One of the ironies is that in cognitive psychology, ‘neural networks’ refers to modeling the function of the mind on the computer science models that themselves are models of the functions of the computer based on the presumed architecture of the brain! It’s not quite recurssive, but it’s pretty close!I don’t think we have much use for artificial humans — they real thing complains too much, and causes all sorts of problems. We’re looking to create autonomous entities that do our bidding, which is much closer to an insect/hive model.
February 18th, 2009 at 3:35 pm
Thanks Dan. I figured as much but I hadn’t realized how self-referential the concept had become until your comment, aside from the sheer complexity of the task ( "Hey…to make a functioning, independent, nation-state, let’s first reverse engineer a galaxy!" ).
.
I also agree with you -I’m not entirely sure why a robot would have to think like a human in order to be useful. Seems like a large, unwarranted, assumption.
February 18th, 2009 at 3:43 pm
"…don’t think we have much use for artificial humans …"
.
It does not have to be absolute, just close enough for certain functions where a close appearance to humans might suffice. Possibilities (1) prostitutes/concubines, (2) security guards, (3) hospice workers.
February 18th, 2009 at 4:54 pm
Mark: Your use of Jeff Hawkins’s TED talk in this post prompts this comment.
Jeff’s company "Numenta is creating a new type of computing technology modeled on the structure and operation of the neocortex. The technology is called Hierarchical Temporal Memory, or HTM, and is applicable to a broad class of problems from machine vision, to fraud detection, to semantic analysis of text. " http://www.numenta.com/
To quote Jeff: "From the beginning, I knew that one day we would be able to build machines that work on principles used by the brain, and that doing so would benefit people all around the world. "
Take a look at the Vision4 demo to see how HTM "learns" to identify objects in images – a significantly difficult categorization/classification problem.
Also Jeff’s book OnIntelligence provides the theoretical background on HTM.
and this link outlines his business strategy http://www.numenta.com/about-numenta/strategy.php
I have been watching the development of Numenta for some time, and feel that its technology offers a lot more than the "military robotics" applications that your post dicsusses.
I think Numenta is emerging as THE breakthrough innovator in the area of applied brain research and that Numenta will one day achieve Jeff’s goal of "benefitting people all over the world."
February 18th, 2009 at 8:45 pm
Right, Lex. Even easier than your three: receptionists, women’s and ethnic studies instructors, journalists. A mouse brain equivalent would work. Give them ten, maybe twenty years.
February 18th, 2009 at 9:07 pm
Presumably, future humans could engineer insect-like humanoids by taking the neanderthal genome, selectively replacing genes related to docility, obedience, and appearance, and then implanting it in a chimpanzee, creating a worker-bee type creature that can be used for the purposes that Lexington outlines, without having a fundamentally human core (or soul??)
It probably would be easier to build a remote and cover it with skin (plus whatever specific organs are desired), as aside from difficult-but-straightforward issues like balance control, everything else is essentially a fine motor/software issue.
February 18th, 2009 at 9:20 pm
Dan, yes. The combination of robotics and genetic engineering will give us … replicants.
.
I picked the three I did because they each would need only limited human-like functionality to do their jobs. Sexbots would not get bored or otherwise object to their owners’ needs, security guards would not get bored walking the perimeter and keeping alert, and hospice workers would be able to devote persistent care and attention to people whom it is too expensive to have large numbers of humans watching with that degree of persistence. The sexbots would probably be primarily biological in structure, the security guards you may want to have made out of metal and looking menacing, the hospice bots might look like friendly cartoon-figure humanoid robots.
.
Of course, for warfare, all kinds of stuff is possible. A fembot-assassin that seduces a an enemy then detonates the explosive in "her" torso? All kinds of combat bots. Robot EMTs to go into hazardous places — burning buildings, combat zones.
.
All of this science-fiction stuff really is going to happen. A lot of it will happen in Japan, first, from the little I read.
February 19th, 2009 at 4:10 am
Hi Dave,
.
I think what you are describing, where certain brain structures are used as an analogous/inspirational starting point for engineering design makes a whole lot more sense to me than trying to reverse engineer a functioning "brain" in toto. Will check out the links.
February 19th, 2009 at 12:21 pm
[…] Military Robotics….Deep in the Singularity Zone […]
February 19th, 2009 at 2:22 pm
Providing a power source would be a problem for any human shaped robot.
Let alone building a useful AI, the problem you would encounter in powering the things are completely beyond our capabilities.
Balance, useful reaction times, sensors, etc, placing this in a human shaped package will be impractical, trying to integrate that with an AI package, near to impossible.
It took 350 years for the helicopter to make it off the drawing board.
September 13th, 2009 at 5:33 am
[…] https://zenpundit.com/?p=3033 […]
September 28th, 2009 at 2:04 pm
" I think the idea of making an Ai for robotics is absolutely insane. Simply put, we all saw terminater and irobot and if there is any malfunction in the AI program the robots could potentially rebel against us. I think a more practical way to go with robotics is to create large manned mech units with the cockpit in the chest of the mech and possibly large bars connected to the feet and wheels attached at the ends of the bars and on the sides or bottoms of the feet for fast movement with such a large mech. I’m still unsure of how to possibly get them air born but this is a dream of mine that i want to persue and any info on the subject would be greatly appreciated.