The Age of Hyper Information society
| Koichiro Tamura Director of the Electrotechnical Laboratory |
|
| Koichiro Tamura Director of the Electrotechnical Laboratory |
We are now on the verge of meeting the 21st century. This is not only the turning of a century but also the end of a millennium. Some would say that people do not always believe in Christianity and this turning point is merely a change of numbers. It may be partly true. Yet, I cannot dismiss the idea that the turning of centuries could symbolically and radically change our views and perspectives.
President Clinton and a noted Japanese baseball player Ichiro are coincidentally saying "We must change." I feel that, now, we are in times of drastic change. I have named this period the "hyper information society."
It is widely argued that three great revolutions of technological and civil importance have taken place in history. The first revolution was the agricultural revolution believed to have occurred some 10,000 years ago. The human population has been continuously increasing ever since. The innovations regarding food production fostered the development of mass production, mass distribution, and mass consumption food technologies which developed into indispensable elements of human life.
The second revolution is the Industrial Revolution which started in the mid 18th century. It enabled the mass production of goods. Today, we can see a great number of industrial products everywhere. It can be correctly said that the origin of the present situation, which is crowded with goods, dates back to more than 200 years ago when the Industrial Revolution began.
The third revolution is the ongoing changes in mass production, mass distribution and mass consumption of information. In my opinion, the origin of these present-day innovations regarding information technology was the advent of computers just before 1950 rather than Johan Gutenberg's invention of movable type in the 15th century.
I believe all of you are aware of that point; I mentioned it just for the sake of confirmation. These ongoing innovations started around 1950 and will probably continue till around 2050. It could therefore take some 100 years for these technological advancements to settle in our society.
Taking a closer look at this projected 100-year period, the present time can be positioned as the "inflection point" of the century-long revolution. Now is the time of the most drastic changes. From the late 1960s to 1970, people were expecting the coming of the "information society." In the 1980s, people found this name not appealing enough and invented a new phrase, "advanced information society." The reason I coined a somewhat peculiar expression, the "hyper information society," is that I think an entirely new phase of information technology innovation has just started and it should be distinguished from what was previously called the "information society" and the "advanced information society." I believe we should clearly recognize this.
I mentioned that the present time is the "inflection point" of this ongoing era of innovation. Completely different types of information technology innovation are beginning right now. It should be noted that we are indeed facing it.
What is the difference between the information society/advanced information society and the "hyper information society"? (Fig. 1)
Information Society
|
Hyper Information Society |
| 1980 | 2000 |
| Rigid Systems -> | Flexible Systems (interaction with enviroment) |
| Pursuit of quantity -> | Pursuit of quality |
| Efficiency -> | Creation |
Fig. 1
Rigid systems are now rapidly changing to flexible systems. In this context,"system" means "information system." In the RWC Project, "flexible information processing " was chosen as a project objective. In my opinion, the difference between rigid and flexible systems is whether or not a system can interact with the outside world or correspond to changes in the outside world.
A flexible object can change shape when it touches other things while a rigid one does not. In other words, the flexible object supports some kind of change in response to the outside world. In this sense, "mainframe computers"- which already sound obsolete-are rigid systems; the general user cannot make slight modifications on them. Figuratively, a mainframe used in a conventional integrated processing system is like the principal object of worship enshrined deep inside a holy building and attended upon by a number of maidens. The general public sit upright in front of terminals in order to access the sacred power of the host computer. We have no other word but "rigid" to describe such a system; this is what we call the "Third On-line System" which is extensively developed in Japan. The Third On-line System is one of the notable successes that the Japanese electronics industry has made. Ironically, however, further development of this particular system is somewhat prevented by its own success and rigidity.
On the other hand, a form of flexible systems constituted by personal computers and networks started emerging in the 1970s, and still continue to increase in number at a remarkable pace. In fact, even a new "distributed integrated system" has gained popularity among general users and has triggered an explosive growth of its user base. In reviewing all these changes and developments, it is obvious that we are entering an era which cannot be correctly described by the conventional phrases - "information society" and "advanced information society."
After flexibility, the second characteristic of the hyper information society is to shift from quantitative efficiency to quality. In the Third On-line System, individual bank systems were designed to meet, as quickly as possible, the needs of customers who wanted to withdraw money from their accounts. The answer for how to provide such quick service was mechanization. Quantitative improvements of information processing capabilities, distributing capabilities, and storage capacity have been extensive. Such quantitative approaches for system design will not disappear, rather, they will be carried out in a more extensive and thorough manner. At the same time, we have recognized the importance of information quality.
Today, the Internet provides us with access to an immeasurable amount of information scattered worldwide. Most of this information however is rubbish. It is important to know how to obtain valuable information from this huge pool of data. Data often loses its value even when it becomes only slightly old. We should consider how we can get the latest and correct information. It is necessary, given the present times, to clarify the relationship between the real world and information; that is to say, we are already in an era where the quality of information should be re-examined from its roots. In short, Real World Computing is an initiative intended to establish such a relationship between information and the real world.
Differing from the quantitative improvements which counted so much in conventional system designs, a special emphasis was exclusively put on the pursuit of better efficiency. The third characteristic, I believe, is how high-quality information can be created in future systems. Since research and development activities can be correctly described as the creation of information, the question which follows is how it is possible to create correct information and useful information, not only in research and development activities but also in the course of our daily lives. In other words, the age of information creation is emerging; it is often pointed out that the contents of information will weigh more and more.
Regarding the subject of this session, I coined a new phrase,"hyper information society," and maintained that our society is now developing from an "advanced information society" to a "hyper information society." Now, let's get back to a very fundamental question "What is information processing?"(Fig. 2)
| Generation/input of information Transfer/distribution of information Processing/calculation of information Storage/retrieval of information |
|
| Network computing |
Fig. 2
The core of information systems is the generation and input of information. The generation of information is the act of putting information in a form which can be handled in the system. It thus deserves the word,"generation." Your ideas become "information" once you write them down on research paper. Ideas cannot be called "information" if they only remain in your brain. In other words, after ideas are put in a system called a "journal," it is understood that information has been generated in that system. This is not all of the story. Recently, information written on paper has little significance or value. Only information available on a network is valuable. So, generating or inputting information means inputting information into a network.
Information gains value when it becomes known and/or used by other people. To this end, methods for information transfer and distribution will be needed. This is what we call "information communication."
Information is also processed and calculated. Processing and calculating information mean reducing the volume of information. For instance, if you inform others of the result of an expression "2+3+5," information will be reduced from "2+3+5" to "10." In other words, the information "2+3+" will be omitted. People always want to gain information which has been processed or calculated. Information is processed and calculated because people want to have only the necessary part of information since they are facing too much raw data to handle by themselves.
Also, information is usually stored as memory in a form which can be withdrawn any time on demand. Each of the following four elemental areas (generation/ input, transfer/ distribution, processing/ calculation and storage/ retrieval) are integrated to constitute an information system. As a matter of course, network computing will be required to operate such an information system.
Office automation is rapidly spreading throughout the world. The particular phrase, "office automation," was coined around 1975 by researchers at Xerox's Palo Alto Research Center (PARC). As one of the underlying motivations for office automation, they intended to integrate the four elemental areas I previously mentioned. In their proposed system implementation, every user was to be given a personal computer with an effective user interface; all PCs would be interconnected. With such a basic plan, PARC's researchers actively promoted the research of a "distributed integrated system" for some 10 years. As a result, almost all aspects of "distributed integrated systems" were invented in that institute. Unfortunately, however, we did not fully understand the meaning of "network computing" and "distributed integrated systems" in Japan. Most PCs were used as stand-alone word processors and could not be expanded into a network environment. This narrow-mindedness continued for years and delayed the development of OA in Japan by 10 years. It was a regretful tragedy that we did not understand very fundamental things.
In the Real World Computing Project, I believe all research activities are being conducted partly or fully based on the four elemental areas. In the generation/input area for example, interfacing will be necessary. Are clicking icons on a Macintosh a true interface? Driven by such a very basic question, researchers continue their activities toward the creation of a better user interface.
In the area of transfer/distribution,"transfer" refers to data transfer, not only among people, but also among computers. So, researchers naturally find it necessary to study networked computers which can work as a combined single computer.
I think the phrase,"information highway society," is already obsolete (Fig. 3). Rather, "Internet society" seems more appropriate.
Various additions will be implemented to form a global network. IMS systems of plants, LANs in offices, and intelligent electrical appliances in private homes will be connected to the global network.
Fig. 3
I drew this figure (Fig. 3) in 1980 and recently added "intellectual artifact." In short, anything in any environment will be linked to the Internet,"Anything" in this context means literally any thing. I have recently heard that washing machines and toasters can be connected to the global network; meaning that Internet-connectable models of washing machines and toasters have already been put on the market. Frankly, I didn't quite understand their use in the beginning. So, I asked many friends what they would think the possible use of networked washing machines and toasters might be. Among their responses, the most interesting and unique answer was that there would be recipe or toasting method applets written in JAVA and stored on network servers in the future. One would just need to explore recipe applets which would best fit his/her tastes. Once you have found one, just download the recipe so that you would be able to enjoy toast that best suits your palate, all the time. Developing this story further, manufacturers of rice cookers wouldn't have to go through a long series of experiments to determine cooking programs and could virtually burn trials on ROMs. By making a rice cooker with Internet connection capabilities, a user can obtain the best cooking program from any server on the network.
An employee of a Japanese manufacturing company told me that the company's employees were directed to think how to possibly connect any of their products to the Internet. This is quite an interesting story.
He also told me that a company-wide direction was issued to consider the possibility of integrating new technologies into their products when ICs (integrated circuits) were invented almost three decades ago. The same directive was given when microcomputers became standard electronic components some years ago. At the time, I personally didn't think of any use for ICs other than for computers. As a matter of fact, the first practical application of ICs was color TV. Since then, IC technology underwent an incredibly rapid evolution.
The next example is microcomputers. Today, microcomputers are found in a variety of Japanese industrial products. Japan has been very strong in this area.
After microcomputers, expert systems followed, which then led into the development of fuzzy systems. Now, the Internet is a great concern among corporate activities.
In 1995, we had a big earthquake in Kobe. Reflecting on that disaster, research was started to make architectural data available via a network, so that monitoring could be done to identify seismically vulnerable buildings and areas.
Well, I know a lot more examples. One thing I can say for sure is the computer is no longer a calculator, but rather a kind of medium (Fig. 4).
|
| Computer: "Calculator" ->Media,intellectual engine |
|
| Ultra huge network |
|
| Social/industrial activities will be put into "virtual spaces" |
|
| Product of labor:"good"->"information" |
|
| Scale of richness "physical" --->"spirtual" and "intellectual" |
Fig. 4
The word "medium" or its plural form "media" has a wide range of meanings. In a sense, the word can be used or interpreted in an arbitrary manner. Once integrated in a system, it can work as an intellectual engine.
Furthermore, an ultra large network like the Internet can be constructed on this medium.
Our social and industrial activities will be moved to "virtual spaces." This phenomenon of "virtual spaces" such as virtual labs, virtual companies, and electric commerce is already emerging through networks.
More importantly, the primary product of labor will be "information" rather than tangible goods. The share of intangible products is overwhelmingly increasing. In other words, software has established its position as a commercial product and are being traded at higher prices than most commodities. Strangely however, Japanese opinion leaders scarcely accept the value of software. It seems to me that they regard computer software as something extra, and often say we'd better work harder to produce better commodity goods in larger quantities. It is quite difficult to persuade them to change their ideas. I know many people saying,"We can manage this 'eperceived problem' by purchasing software." They are not aware of the fact that software is also the product of rigorous labor. It's an awful situation. Software is the product of our brain work. As I mentioned before, we are now in the time of an Information Revolution, following the two great revolutions-Agricultural and Industrial. The Information Revolution means the coming of a new era in which people start creating software information.
Japan is not rich in energy and natural resources. The only resource we have an abundance of is human brain power. So, we have to radically change the direction of our industrial activities. Otherwise, Japan will have a very hard time in the 21st century.
In the past, we believed we were rich as long as we could eat a lot and possess a lot. But that concept of wealth turned out to be wrong, and we are now discovering for the first time the fact that information provides intellectual and spiritual nourishment."Information" has various meanings here. Religion and art are "information" as well. Thanks to such information, we can enrich our spirit and intellect, and feel satisfaction that goods or food cannot fulfill. In other words, it is an era in which "spiritual" and "intellectual" values are much more important than physical values. Technology has developed to support such an era. It is not an era of superficial change but rather fundamental change. Although it is a stale word, I would like to say a "paradigm shift" is taking place (Fig. 5).
|
| Centralized control | --> | Autonomous Distribution |
|
| Closeness | --> | Openness |
|
| Homogeeity | --> | Diversity |
|
| Supplier-leading | --> | User-Leading |
Fig. 5
As shown in the figure, the first shift is "from centralized control to autonomous distributed control." This can be true of information systems as well as social structures. We often think of dismissal when we hear the word "restructuring." However, many organizations are in the process of "restructuring." They are trying to shift from a rigid system, constituting a vertical hierarchy, to a new system of a flatter and more dynamic organizational structure. The Electrotechnical Laboratory, our institute, is also conducting such reforms. It is true that a paradigm shift is occurring in the area of organizational structure.
Organizations were usually closed to other organizations and individuals. However, network-type activities are increasingly developing and expanding. For instance, Company A has been doing everything necessary for its business by itself. But Company A finds that some things are no longer possible to carry out by itself, and has to ask for a certain technology from Company B and technical support from Company C.
This cross-organizational networking is also true of research activities. The percentage of research activities that are closed to cooperation from outside organizations is decreasing in the Electrotechnical Laboratory. It is statistically proven that the number of papers produced jointly by members of the Electrotechnical Laboratory and other institutes has been steadily increasing during the last decade. The traditional boundaries of individual organizations is becoming unclear.
Information networks are also losing their clear-cut boundaries. It is not true to say there were no global networks before the Internet. Some large companies in the United States, for instance, had well-organized global networks. Such corporate networks were, however, confined to their own companies. These company-wide networks linked major cities like New York, Paris, and London. Yet, they were closed to the outside world and not connected with any other companies or universities.
The Internet is quite different from those company-owned networks. It is a completely open system that anyone can access. At the same time, you can limit access if you want, and it is theoretically possible to add other networks such as Intranets and Extranets to the Internet. This expandability is a good characteristic of the Internet's openness.
Diversity is more important than homogeneity. This shift of value is also a result of the "Information Revolution." The true value of information is "changeability" and "difference." Goods and foods can keep their value even when the same goods and foods are produced in large quantities. On the other hand, information does not retain value in the same way. Once you know a certain piece of information, it is useless to hear that same information again. So, difference counts here.
Human groups also produce value from their existence, as long as they maintain diversity. If two people share exactly the same opinion, no value is created. In conventional "object-oriented" industrial activities, combining the labor of two people might double the amount of products. In a new information-oriented society, however, value will be created when discussion takes place among people having different ideas or opinions. So, diversity counts much more than homogeneity.
Next is the shift from supplier-leading to user-leading systems. In a mass-production economy, a limited number of suppliers often dominate the market share and can set the overall direction of user behavior. In future systems, emphasis will be placed on each user's demand. This can be true not only in regards to goods, but also in regards to information. Radio/TV stations and newspaper publishers collect information and evaluate each piece of information so that they can purposefully deliver news or information of marketable value to their customers. On the Internet, however, a huge number of individual users around the world are transmitting information on their own. So, the user is beginning to have the initiative over the supplier.
Now, we are going to review some issues we may face in constructing a "hyper information society."(Fig. 6)
|
| "Papyrus computer" | |
|
| Super high-performance server | |
|
| High-speed network | |
|
| High-performance chip | |
| Where is the limit? |
Fig. 6
Professor Tanaka has already provided clarification on these matters. So, I would like to discuss them from a different perspective.
Around 1975, the notebook PC was accurately predicted. Furthermore, even network connectivity was predicted to be implemented on notebook PCs. Many researchers, including myself, predicted the advent of notebook PCs more than 20 years ago. I have been often asked,"If you can foresee the future, what will come after notebook PCs?" My answer to that question is "papyrus computers" -which are sheet-type computers. Notebook PCs are rapidly improving, and I believe the ultimate form of notebook PCs will be a very thin paper-like computer. We will not need such a troublesome multi-window processing system once we have a papyrus computer. Bound like loose-leaf, papyrus computers would allow you to take a look through multiple pages (screens) by turning them. There will probably be very interesting and effective application areas. I'm really looking forward to the development of a papyrus computer. I hopefully believe that the sheet-type computer will be developed in ten years.
My next point of discussion is in regards to "super high-performance servers." Like NCs and net PCs, software burdens are rapidly migrating to servers in an effort to reduce the workload of clients. This is quite a natural way of development. In a sense, we are approaching another age of central large computers. Future large computers must be different from old mainframes, which were very rigid systems. The primary form of implementation will be through network servers, and they will be required to have extremely high performance. To meet the need for extremely high performance, massively parallel computers will be a natural choice.
When used in an organization, a client/server system will become uncontrollable if clients go their own way and use different versions of software applications. There are strong demands to control such arbitrary use of software. To the best of my knowledge, the first solution to this problem was the Plan Nine system developed by the Bell Labs. I think they have been working on this subject for more than ten years.
"High-speed networkd" is my next point of discussion. A Gigabit-Ethernet is already available; an ethernet with a 1-Gbps transfer capability will soon be extensively installed in offices, institutes, and universities.
Just five or six years ago, people suspiciously asked what the use of such a high-speed network would be. Some argued that 100 Mbps was fast enough. However, it is now commonly accepted that 1 Gbps is the minimum speed requirement in the multimedia age. High-performance chips in DSP or the like, are the primary concern regarding the achievement of essential processing speeds for the multimedia age. We know that the evolution of hardware is not endless, however there are no bounds on demands deriving from the software side.
As the Fifth Generation Computer Systems Project produced a good deal of achievements, the Real World Computing Project will be a big success as well. However, if you plan to put the results of the RWC Project into practical use, the new computing capabilities will use up all available computing power and hardware resources found in today's technology. I'm not sure when it will be, but software demands will eventually exceed hardware capacity around 2030, 2040, or some other time in the future. What shall we do then? That is the question. Well, I can personally say it's none of my business because I'm not sure if I will still be alive then. Yet, this is a very serious problem that we must consider in depth.
I think some RWCP members were motivated by this particular issue to join the Real World Computing Project. It is essential to generate a paradigm shift at a very fundamental level of present computer technology. I do not mean I expect significant shifts in the fields of genetic algorithms and chaos so suddenly, but a paradigm shift will be required at an unexpectedly fundamental level.
In reviewing the temporal development of software from the present into the future, now is the time for object-oriented systems to reach their maturity (see Fig. 7).
|
| Maturity of object-oriented system |
| | | |
| V | |
|
| Agent |
| | | |
| V | |
|
| Cooperative agent |
| | | |
| V | |
|
| Formation of "information field" |
Fig. 7
We now have a lot of object-oriented products, including JAVA, on the market. I predicted that object-oriented systems would become popular in practical areas during the 1990s. It is already at the end of the decade and a lot of object-oriented products are appearing on the market.
The "object" has an image of something rigid. The "agent" is a flexible object with the capability of interacting with the outside world. This "agent" concept will be widely accepted. Even now, we already have "user-interface agents" and "network agents." Agent-oriented systems will soon make a big breakthrough, and will be extensively used in various areas.
If agents work individually, they could not cause great effects. Agents have to work in a cooperative manner.
A project on cooperative agents that the Ministry of International Trade and Industry started seven or eight years ago will be completed in the next fiscal. For the purpose of demonstration, they made a football game - football is a typical example of "gcooperative agents." The manager do not instruct each player, but each player thinks what to do towards the final purpose of scoring a goal. The cooperative agent program was thoroughly beaten by another football game program developed by students at the University of Tokyo. I asked them what the main reason for their defeat was. They answered that their program didn't have a strategy. I didn't quite understand how to make up a football strategy but I'm sure it's not an inherent defect of the cooperative agent system.
I believe that the concept of "field," such as "information field," will make an impressive comeback, and an entirely new software system will be created to utilize the huge number of computers connected via networks. I understand RWCP is vigorously working on that kind of software system.
The word "information field" his very vague (Fig. 8). The vagueness in meaning gives more possibilities to the perception of this concept. Conventional networks are all communication-oriented. Conventional networks will be changed over to "field computing" or "hybrid computing". In other words, the network will provide a field for computation.
Fig. 8
The information field is a fundamental structure which will be an established infrastructure in the real world, the information world in which people live.
I said that anything can be connected to the Internet. Strictly speaking, however, anything which is computer integrated can be connected to the Internet. As a whole, those computers will support us. I think the fundamental structure can be called "information field" or "information world." The point here is whether or not the information world correctly reflects the real world. If the information world falsely reflects the real world, we would have an outrageous infrastructure. In this sense, real world computing has a very important significance.
We have to clearly define both the information and real worlds, and deliberate over how the two worlds should be related to each other (Fig. 9).
Fig. 9
In reviewing the pursuit of information quality (Fig. 10), the value of information, specifically where you obtain it, where you send it, and your format of expression will be important.
Next is accuracy. Regarding industrial standards, confusion reigns if the definition of a word such as, "one volt," is different between two people. Standardization is an initiative designed to eliminate such complications. People will pay far more attention to whether or not expressed information coincides with each other.
|
| Value From where To where Expression of required information |
|
| Accuracy Commonality of information expression (standardization) |
|
| Freshness Response to change |
Fig. 10
Another element is "reshness." All these factors constitute the quality of information. We will probably face other fundamental problems which may need our thorough consideration.
Since we are in the Internet age, the Real World Computing Project poses great significance for future societies. I really hope that all researchers and other people involved in the RWC Project will make utmost efforts to make this project a big success.