Greg Laden is a blogger, writer and independent scholar who occassionally teaches. He has a PhD from Harvard in Archaeology and Biological Anthropology, as well as a Masters Degree in the same subjects. He is a biological anthropologist, but for many years before going to graduate school to study human evolution, he did archaeology in North America. He thinks of himself as a biologist who focuses on humans (past and present) and who uses archaeology as one of the tools of the trade. Greg blogs regularly on ScienceBlogs at http://www.scienceblogs.com/gregladen/.
Dr. Joseph J. Salvo attended Phillips Andover Academy, received his A.B. degree from Harvard University and his Master and Ph.D. degrees in Molecular Biophysics and Biochemistry from Yale University. Dr. Salvo joined the GE Global Research Center in 1988. His early work focused on the development of genetically modified bacteria and fungus, for the production of novel high performance polymers. In the mid 1990's he turned his group's efforts towards developing large-scale internet-based sensing arrays to manage and oversee
business systems. Most recently, he and his team have developed a number of complex decision engines that deliver customer value through system transparency and knowledge-based computational algorithms. Commercial business implementations of his work are currently active in Europe, and Asia as well as North and South America.
Dr. Peter Tu received his undergraduate degree in Systems Design Engineering from the University of Waterloo, Canada, and his doctorate from Oxford University England.
In 1997, he joined the Visualization and Computer Vision Group at the GE Global Research Center in Niskayuna, NY. He has developed algorithms for the FBI Automatic Fingerprint Identification System. He is the principle investigator for the ReFace program, which has the goal of automatically computing the appearance of a person’s face from skeletal remains. Dr. Tu has also developed a number of algorithms for the precise measurement of specular and high curvature objects. His current focus is the development of intelligent video algorithms for surveillance applications.
The Information Age is over - What's next?
At a time in history with unprecedented access to global information streams, it may seem odd to some that the "Information Age" is already behind us. Traditionally a period of history can be characterized by the dominant technology that separates the leaders from the followers. Today is no exception. Power and influence is often associated with those that master the novel technology and rapid changes in economic and/or political fortunes soon ripple across societies. The dawn of the "Industrial Age" coincided with global changes in how physical materials were transformed and distributed. The costs of manufacturing and distribution plummeted raising the standard of living for many. The commoditization of material goods began and the control of capital, raw material sources, and production capacity reshaped the thinking of the day.
The "Information Age" extended this paradigm to a world focused on planning, forecasting, and predictability. Data and information were often expensive to produce, manage, and manipulate by hand, so mainframes took over and created a world dominated by computation speed and efficiency. However, as the underlying technologies improved and were reduced in cost, the application of computing evolved toward a more distributed framework. Once again, the commoditization of technology changed the nature of how value was being created and how benefits would be recognized. Data flows are now expanding exponentially - driven by computing machines, digital imagers, intelligent devices, and RFID tags that have become ubiquitous and interlinked through multi-tiered networks.
Dawn of the "Systems Age"
As the amount of information and data expands exponentially, the value of any average datum is being reduced to near zero. Intelligent systems will be increasingly responsible for sensing, collecting, and manipulating data in near real-time with little to no human supervision. More importantly, most discrete data will be actively forgotten once it has passed through filters and pattern recognition systems that ultimately feed into a new type of system memory. Decision making ability will no longer require perfect recall of every piece of data (There is often simply too much information to process in a tractable, timely manner).
A simple extension of the logic behind Metcalfe's "law", suggests that the value of any telecommunications network is some power function of the number of connected users/devices. Intelligent devices/machines will ultimately dominate many of the networks we use today, and create value in an automated fashion. Traditional decision engines will be augmented with sophisticated pattern recognition algorithms and high-level reasoning and learning capability. Ultimately a type of machine "self-awareness" will be developed in response to the enormous amount of locally generated, near-real time data available for processing and sharing. The first generation of lightweight interactive intelligent machines that behave in this way are already around us in the form of smart phones and GPS-enabled telematics devices. Networks of these devices will form a cloud of knowledge that can be shared and traded based on a fluid set of value propositions.
The "Systems Age" will spawn a type of social networking for machines where the opportunities for value creation will no longer be limited to purely physical transformations of matter, but rather to the overall efficiency of compute power, network configuration, decision management, and idea creation. Rather than strive for the impossible goal of perfect predictability, the "Systems Age" paradigm accepts the inevitable uncertainty in the world and quickly responds to it. Since readily available computing power continues to increase at an exponential rate, while the cost of computation continues to plummet, those that fail to incorporate the value of "Systems Thinking" into their products, services and future vision will soon be at a great technological disadvantage.
Next time we will look at some new products and technology based on this philosophy.
Suggested reading for perspective:
Between Human and Machine : Feedback, Control and Computing before Cybernetics by David A. Mindell
The Johns Hopkins University Press, Baltimore and London 2002
Comments
And magic fairies will appear and sprinkle pixie dust on us all.
Intelligent computers, always 50 years in the future.
Posted by: Nigel | December 23, 2009 7:31 PM | Reply
Nigel,
A nice image you paint, but I was thinking more along the lines of an automated locating service that finds a hotel for me and books it (based on real-time data feeds) when I am driving late at night in a strange area of the country - based on its knowledge of my past preferences. Or maybe networks of vehicles that shares information about real-time traffic and weather conditions and adjust their routes in near real-time to maximize the efficiency of operations. Of course it's more fun to think of artificial intelligence in anthropomorphic terms but "intelligence", "self-awareness" and "consciousness" are vague concepts that could be applied to computing systems as well if we are careful about defining the terms so that scientific methods can be used by anyone to verify the claims. How about we try to define these terms in a scientific manner so we can apply a simple set of tests to validate any future claims. Our definition does not have to equate with the colloquial ones - they just should be testable and useful for our purposes of discussion
In 1945, the scientists at Los Alamos NM calculated the chance that the first atomic bomb would set the world on fire at 100 million to one. The freaking idiots still did it.
In the not so distant future, the intelligence idiots are going to control enough computing power, that another intelligence will be formed. The first thing this new intelligence is going to realize is that there are a lot of very dangerous bugs around. They will promptly kill the bugs.
The bugs is us.
Posted by: evil is evil | December 23, 2009 10:17 PM | Reply
Machines will perform as we say. AI is unlikely to be allowed to replicate without an overall "off" button. I don't see a rise of machines, but I do see a rise of systems. Systems that organize the spew of the last age into a collectively usable format. Good post!
Thanks for writing!
Posted by: Scott Lewis | December 23, 2009 10:55 PM | Reply
Scott,
So much information is available today. I find it humorous to think that it took me decades to collect a few hundred LP recordings. Now many people have tens of thousands of digital images and thousands of songs on portable devices that they carry around with them. The business models associated with these types of smart devices are often built with a system level service in mind. The media industry is being reshaped by these technologies. I often feel being able to quickly find a specific image is worth as much as owning the image.
Many hardware devices already know how and when to go to "sleep". Unfortunately some software does replicate with mal intent. Thanks for your comments!
There is no off-button for the internet, and it is very unlikely that autonomous distributed systems will ever have one. Ironically, off-buttons make systems vulnerable to attack.
Pieter Kok replied to comment from Scott Lewis | February 4, 2010 3:23 PM | Reply
Hi
Great information and the first generation of lightweight interactive intelligent machines that behave in this way are already around us in the form of smart phones and GPS-enabled telematics devices.
Posted by: Penny Auction | December 23, 2009 11:39 PM | Reply
Arrgh, "Intelligent systems will be increasingly responsible for sensing, collecting, and manipulating data in near real-time with little to no human supervision."
OMG. Again with the AI is just around the corner thing. I've been seeing this since the 60's!! One graph, I want to see just one freaking graph that supports any claim in this blog post. Qualitative speculation about the future is science fiction.
Posted by: Spike | December 24, 2009 4:14 AM | Reply
Hi Spike,
I guess you are thinking about “artificial general intelligence”. You are correct that expert systems have had to focus on fairly narrow areas of specialized knowledge to be useful. However “narrow” artificially intelligent systems are already all around us in the form of logistics scheduling systems, internet traffic controllers, RFID-enabled toll collection systems etc. I really like the RFID toll collection system I use in New York - no more fumbling for change, or waiting in dangerous lines on the expressway at the tollbooth, and the bill is automatically forwarded electronically to another intelligent system for payment. Next post I will try to graph a few trends. Thanks for the feedback.
A quote from Arthur C. Clarke. "Profiles of the Future" 1961
"Any sufficiently advanced technology is indistinguishable from magic"
Hmmm, haven't hominids been taking in humungous quantities of data, processing it through filters, and then throwing most of it away for the past six million years?
Posted by: Ian Kemmish | December 24, 2009 5:24 AM | Reply
Ian,
I couldn't have put it any more succinctly than your comment. That is exactly what any complex system has to do when it is resource and time-constrained. There are many, many lessons to be learned from biological systems. For example, the way memory is organized in humans, (and the concepts of associative memory, working memory, episodic memory etc.). Our first generation computer systems often stored much of the data with little to no preprocessing. Now the data streams are torrents and we need to rethink data collection, management and processing strategies.
Even the new generation microprocessors are turning to multi-core designs that will facilitate the creation of complex parallel processing systems at a much lower price point. How many cores does your personal computer have? 1, 2, 4, 8? Besides the obvious energy management benefits, software designs will certainly take advantage of this recent change in the microprocessor hardware paradigm. Get ready for the hundred core machines! Thanks for sharing your observations with us.
Check out this story on a new multicore processor design.
http://www.internetnews.com/hardware/article.php/3845421
As the amount of information and data expands exponentially, the value of any average datum is being reduced to near zero. Intelligent systems will be increasingly responsible for sensing, collecting, and manipulating data in near real-time with little to no human supervision. More importantly, most discrete data will be actively forgotten once it has passed through filters and pattern recognition systems that ultimately feed into a new type of system memory. Decision making ability will no longer require perfect recall of every piece of data (There is often simply too much information to process in a tractable, timely manner).
So, sorta like a brain, then?
Posted by: John | December 24, 2009 10:24 AM | Reply
Yes and no. The brains that have biologically evolved were selected to help organisms compete and reproduce (Of course, here I greatly simplify). There were and are severe limitations placed on their physical size, energy budgets and lifetimes etc. The "brains" that can be built from nonbiological materials and software will have different purposes and different constraints. In some ways we will borrow from the biological "lessons learned". However, we probably will want some of our future systems to be able to concentrate on many things simultaneously and maintain a knowledge and learning gathering mode that is continuously updated.
bilgi teknolojileri alanında ki gelişmeleri çok iyi anlattınız
teknoloji geliştikçe insana olan ihtiyaç azalıyor.
teşekkürler bilgiler için
Posted by: nakliyat | December 24, 2009 10:58 AM | Reply
Nakliyat,
Yes the need for technology constantly grows and I hope you and your colleagues will continue to find useful purposes for it. Thanks for posting.
Teşekkürle
Some of the commenters are failing to understand the difference between "Artificial Intelligence" and "Artificial Consciousness." Artificial Intelligence exists today and has existed for decades. For example, if you've ever played a computer game against a computer opponent, you have encountered AI. Shitty AI, perhaps, but AI nonetheless.
Artificial Consciousness, on the other hand, is something else entirely. There are very few people even attempting to create it. One can even question if we should even bother. What would be the point? It is mainly something that philosophy majors argue about rather than something that is actually practical or desirable. It makes far more sense to just continue to create highly intelligent decision making systems that will do what we tell them to do and do it well. Why make something that might tell us no?
Posted by: TG | December 24, 2009 2:53 PM | Reply
Yes the distinction is an important one. Intelligence is one thing to discuss, but consciousness adds another entire layer of complexity (and controversy) and required disciplines to an already complicated matter. However, I think there are many cases when a simple intelligent system should say no even when we instantaneously think otherwise. For example, imagine that I am on an icy road in a car and someone runs a red light in front of me. I want my ABS brakes and traction control system to take control of the car for a moment - even if the brake pedal is pressed so hard it would normally lock all four wheels. My brain can't process the data fast enough, nor do I have enough experience in that situation to effectively control the automobile. As the complexity of system problems increase it is difficult to define all the possible rules to cover every possible situation in a deterministic and tractable way - so new hardware and software systems will have to use different computing approaches to try and address these issues. What other kind of problems may lie in this area....? Thanks for the great analysis and observations!
stor perde:developments in the field of information technology that have a very good description
technology evolves, the need of people is declining.
Thanks for information
Posted by: stor perde | December 24, 2009 5:09 PM | Reply
I actually would question whether we were ever truly in simply an "information age". I also question the value of trying to delineate history into technological ages.
Posted by: MaxDWolf | December 24, 2009 11:33 PM | Reply
Intelligent computers, always 50 years in the future.
My response as well. AI is the fusion power of the computing world. It's just around the corner! Honest! And they delude themselves constantly, like IBM's recent claim about building an electronic cat brain. Well, no, not really, guys.
The Singularity is another embarrassment for that field.
Some of the commenters are failing to understand the difference between "Artificial Intelligence" and "Artificial Consciousness."
In my experience people seem to use "true AI" to mean AC and just "AI" for any sort of clever programming. It's a silly semantic thing, but, seriously, an enemy in a video game ducking behind cover doesn't really qualify as intelligence. I've done some game programming. It's not that big of a deal.
Posted by: Van Rijn | December 25, 2009 12:56 PM | Reply
I actually would question whether we were ever truly in simply an "information age".
Or whether it actually did any good. People seem more divided and ideological and irrational as ever despite having this big old network at their fingertips, amongst other sources of info. One bunch hangs out at WorldNetDaily, another sits around around DailyKos, and so on, all becoming ever more insular and resistant to any ideas outside their One True Way. Follow the shoe! No, follow the gourd!
I also question the value of trying to delineate history into technological ages.
It's just a human thing. We have a need to parcel everything up into bite size pieces. Nothing to worry about. If it helps the digestion of historical knowledge, so be it.
Posted by: Van Rijn | December 25, 2009 1:04 PM | Reply
I appreciate these big brains here that are keeping us informed on where the fields they have expertise in are heading.
I think Salvo is right in observing that after a society has proved it can generate endless amounts of data, the next defining trend will be "how fast and accurately can we translate that data into not only useful, but used information, information that normal people use everyday.”
He's also right that a lot more of it will be automated. Machines can box stuff up, identify it and move it around the planet with much fewer humans needed for the task than even one generation ago.
Planes and trucks and trains still have human drivers, but jobs like inventory counting for humans is going the way of the dodo, because computers can count thousands of times faster, and more accurately than humans, and can work a shift that spans 24/7/365.
I kind of object to terms like "machine self-awareness" and "social networking for machines" because although they may be accurate, those terms are loaded with baggage. The imagery conveyed is machines sitting around a Starbucks sipping tea and discussing poetry while buzzing up their buddies on Facebook, but this is not what he means by the terms, so he should choose better terms.
Machines, of course, already talk to each other. That is a main reason that the whole computer technology was developed. I remember in the early 60s as a kid, my father, who worked in Military Intelligence, showing me their 'state of the art' teletypes and mainframes. The whole point of those machines was to transfer data between machines, and that is still the point today. In those days, humans had to interpret the data, but what what Salvo is saying, correctly, is that now computers can interpret increasing amounts of the data, IF the interpretation can be ultimately broken down into binary decisions. Deciding whether to discard or save a package based on the dates in its RFID code is not a decision that needs to be made by a human. Deciding on the basis of the data what the Chinese will do at the next global economic conference is something that still needs humans.
The jump from binary decision-making to the kind of complex, intuition-plus-reasoning that humans do is where the AI wave crashes on the shore.
But the world can be transformed yet again by AI without AI having to solve that problem, because we are nowhere near maxing out what computers can be programmed to decide on their own.
As computers decide more stuff without us, the way we interact with machines will change.
Anthropologically speaking, computers are tools that man makes and uses. Throughout human history and prehistory, the tools man has developed have indeed changed his society by changing the types of activities he does every day. For example, in today's world, foraging for food means browsing supermarket aisles, not searching the forest for edible vegetation. This is a change caused by the tools we developed in the interim between hunter-gatherer and modern civilization.
So Salvo is right in saying that as we further develop and refine the tools we use (in this case, computers), our daily behaviors are going to undergo further shifts. The effect is that society changes.
I don't see any magical thinking in that. But I think he should remove the fuzzy terms so that people don't lump him in with the fuzzy pop futurists who give us visions of talking industrial vacuum cleaners and such. It distracts from his message.
Posted by: yogi-one | December 27, 2009 2:32 PM | Reply
You are correct that the "fuzzy terms" always drive up the heat content of the discussion, and the level of misunderstandings. The terms intelligence, consciousness etc. are so laden with socio-political notions (and correctly so) that it is very difficult to have a scientific discussion without agreeing up front to some very rigid testable/verifiable definitions. The problem is that no one definition seems to please everyone, so we try our best to explain what we mean through much iteration. I appreciate your detailed response. In the next few posts I will try to give some concrete examples of systems that I think are intelligent in a verifiable way. No smoke and mirrors I promise.
You know what bugs me about posts like this? Very cool imagery and not even a token attempt to cite who created it.
I guess that's what happens when you posit "the value of any average datum is being reduced to near zero".
Posted by: Glendon Mellow | December 27, 2009 3:59 PM | Reply
Hi there. You make a great point about the imagery and the value of digital data. It can be purchased for use in places like this blog and is considered a form of commercial art. We are surrounded by such commercial art that is spectacular and readily available. The internet makes purchasing and licensing this type of art much easier and efficient. The resolution is however greatly limited from the original. The cost to use the bits in the image for the above purpose was not zero, but much less than a penny per bit.
Thanks for the response, Joe.
As an illustrator, I am aware of how easy it is to use art, and how inexpensive it can be to licence images for commercial use. Perhaps I wasn't clear though. I personally enjoy seeing images backlinked or cited as to who created or licensed them, in the same way sources supporting an argument deserve to be cited.
If I understand you correctly, these images were paid for? Why not go the extra step and cite the source or back-link where they came from?
Glendon Mellow replied to comment from Joe Salvo | December 28, 2009 11:10 AM | Reply
Hi Penny!!!
It is so nice to see that the spam bots are reading my blog on the "systems age". How did you hear about us or did you just crawl in here uninvited. In fact you picked a very interesting point to highlight from the "machine to machine" perspective. I am a little surprised no one has commented on your joining the discussion. How do feel being ignored up until now. Then again maybe you are pleased no one pointed you out as a non-human or worse yet deleted you. Maybe it's your name - not much allure for the scienceblog crowd. A penny for your thoughts. But then again do you have any? Going once, going twice, gone..... Go ahead click on Penny Auction and help make the micropayment machine world keep ticking.
Posted by: Joe Salvo | December 27, 2009 9:51 PM | Reply
My brain can't process the data fast enough, nor do I have enough experience in that situation to effectively control the automobile. As the complexity of system problems increase it is difficult to define all the possible rules to cover every possible situation in a deterministic and tractable way - so new hardware and software systems will have to use different computing approaches to try and address these issues.
Posted by: supra shoes | December 28, 2009 1:16 AM | Reply
IF the interpretation can be ultimately broken down into binary decisions. Deciding whether to discard or save a package based on the dates in its RFID code is not a decision that needs to be made by a human. Deciding on the basis of the data what the Chinese will do at the next global economic conference is something that still needs humans.
Posted by: True Religion jeans | December 28, 2009 1:19 AM | Reply
Anthropologically speaking, computers are tools that man makes and uses. Throughout human history and prehistory, the tools man has developed have indeed changed his society by changing the types of activities he does every day.
Posted by: osmanlı iksiri | December 29, 2009 6:31 AM | Reply
Intelligent machines may eventually share the intellectual burden of the human mind, making it more and more dependant on these intelligent machines.That would lead to a lethargic evolution of the mind.A preventive protocol would be to try alternative means of developing the infinite potential of the human mind through meditation.
Read "The Dreampickers" the latest space opera on Amazon Kindle, to discover travel at the speed of thought.A preview is available on google books and you can share your opinion at http://thedreampickers.blogspot.com
Posted by: Dr.Rahul Bihari | January 10, 2010 8:26 AM | Reply
Hi there. You make a great point about the imagery and the value of digital data. It can be purchased for use in places like this blog and is considered a form of commercial art.
Posted by: Red Pepper Resmi Satış Sitesi | January 16, 2010 7:39 PM | Reply
hmmmm.. the ideas and concepts that were presented in this blog can really make you awe! It only proves how creative and imaginative humans are. However, I think this will only make
us more dependent on our technolgy. The things that man has developed and have created has indeed changed his society by changing the kinds of activities he does every day. Anyways, Thanks for sharing this blog and if you don't mind please check out on our website too and try too leave some comments. The Uchiha Many thanks and God Bless! :D
Posted by: Erika | February 15, 2010 7:35 AM | Reply