Computer Intelligence Assets presents its new proprietary technology: INTELLECTRONICS [NEW YORK] "The INTRON is a new device for automation of tasks which routinely require complex problem solving," said Dr. James Keene. "Advanced in concept and performance, the INTRON is in a league by itself and may therefore be distinguished from computer-based artificial intelligence (AI)." Thus began Dr. Keene's announcement of an exciting new discovery: the INTRON. Let us look at the INTRON, and where it might take us. The INTRON has numerous applications in both service and manufacturing industries. The most obvious military and space applications alone suggest that the nation which develops these projects would have an enormous advantage. The invention of the INTRON suggests specific investment opportunities and applications projects. Investors are sorting out the outlook as follows: --- The economic and national security implications of this advanced technology (THE BOTTOM LINE), --- A functional description of the invention (WHAT IS AN INTRON?), --- Some applications of Super-Intelligence (HOW ARE INTRONS USED?), --- How the invention might be developed into marketable products (THE DEBUT OF INTELLECTRONICS). THE BOTTOM LINE Please consider that... (1) The previous lead of the United States in technology has been diluted by significant innovations by foreign companies. (2) Japan alone has committed substantial sums of money to R & D efforts to develop the "fifth generation" of computers. Regardless of the amount spent on R & D, there is, of course, no guarantee that the desired product will be created. However, it may well be that the INTRON is the key invention required in this quest. Since the INTRON is now a reality, eliminating this particular element of R & D risk, any reasonable equity investment may be considered to be a bargain. Let us assume that our competitors have not yet developed the "fifth generation" of computers. They could not know, then, that a key invention required to fulfil their goals, the INTRON, is not even a computer! In fact, the INTRON invention in Super-Intelligence provides immediate leverage in planning because the structure of these "fifth generation" systems becomes clearer. They will be INTRON systems in which arrays of INTRONS perform tasks using computers and input/output devices as assistants. (3) If performance and cost alone are evaluated, most current users of computer technology will shift to INTRON CONTROL systems to obtain or maintain competitive positions. In brief, the market for INTRONS is as large as that for computers. In addition, since INTRON CONTROL systems have more applications than computers, it is reasonable to assume that the forthcoming INTRON market is larger. (4) Because these advanced systems will usually use computers in auxiliary functions (such as input/output interfacing, data management, calculations, etc), the manufacturers of computers will be modifying their products to make them more cost effective in these INTRON applications, and thereby maintain or increase their customer bases. This would apply equally to the makers of large mainframe computers as well as the microprocessor industry. (5) As the manufacturers of computer hardware adjust to the appearance of INTRON technology, so will a large segment of the computer software industry. For example, much software (from simple accounting to AI programs) would include subroutines to interface the programs through the hardware interfaces of their host computers with the INTRON CONTROL devices. More specifically, much of the current AI software (whether it be programs to guide a lunar land rover or simple speech synthesis) are appropriate tools to place under INTRON control. (6) These considerations lead, of course, to the active fields of AI and robots. For the former, INTRON technology does not represent direct competition. In the first place, Super-Intelligence as implemented by INTRONS performs higher order functions than do AI methods. Hence, direct competition between INTRONS and AI is not in the cards. Rather, the two will complement each other in hundreds of applications. In the case of robots, the situation is somewhat different since INTRON-controlled robots can perform a far greater variety of tasks of much greater complexity than computer-driven robots. Considerable funds are currently dedicated to development of improved automation in a variety of manufacturing, space, and military situations. These projects have shown the elegance of AI methods as well as their limitations. The invention of the INTRON now provides further perspective in how these projects might best be developed. In a great number of cases where AI cannot deliver the desired results, INTRON technology should be considered. In sum, the INTRON has important economic and national security implications. WHAT IS AN INTRON? The word INTRON is a short form of INTELLECTRON, which contains the word "intellect" and the sound of "electron". Hence, the INTRON is an electronic device which displays Super-Intelligence. As such, it can perform tasks requiring more intelligence than current artificial intelligence can in principle perform. While the intelligence of INTRONS is clearly greater than AI systems, it should be considered less than that of humans, although INTRONS perform better than humans in numerous applications. To distinguish the level of intelligence involved, the term "Super-Intelligence" has been introduced to avoid confusion. The INTRON can learn how to use complex information provided to it to solve a problem in situations where both the possible actions and the criteria of successful problem solution are both well defined. However, the key feature is its success in solving problems when the method or procedure for using the information provided to produce successful actions is not known, or not clear, or even changes radically over time. Thus, INTRONS are continuously learning how to use the information provided pertaining to the problem at hand, much as animals and humans do. The concept and inner design of the INTRON is unique, bearing little, if any, resemblance to AI or statistical techniques (which are often mistakenly employed as mentioned above where the tasks exceed the capabilities). Naturally, the performance of INTRONS verifies this statement. Indeed, many of the applications environments mentioned above would be challenging even to human experts, if required to produce successful results without knowledge of how the information should be utilized. More specifically, the INTRON's intelligence can be applied in any situation which meets the following four criteria: (1) PERTINENT INFORMATION can be delivered to the INTRON. Not even human intelligence can solve problems or produce adaptive behavior without relevant information. However, the word "pertinent" is important here. It need not be known before hand which information is IN FACT relevant. It could well be that much of the information provided is simply not used by the INTRON. The INTRON can perform its functions when a sufficient amount of pertinent information is included in the total body of information provided. There are several important implications of this fact. First, the designers of new products are NOT required to possess the scientific or technical knowledge concerning what information is in fact necessary. This knowledge need not even exist. The designer need only be familiar enough with the application to provide a wide variety of potentially pertinent information to the device. Second, all of those applications which have required such knowledge but have remained dreams since this knowledge did not yet exist, may now be implemented with INTRON technology. Third, it should be clear from the foregoing that the INTRON can be used as a RESEARCH INSTRUMENT, since the knowledge which the device develops in the performance of its assigned task can be "dumped" and examined. Thus, the INTRON will perform the task and in addition, reveal precisely which of the available information was used and how it was used. Once this is known, products and applications can be improved by discarding information which proved to be irrelevant (not used) and providing new information which might improve performance. Some concrete examples of information which would meet criteria 1 are: *** All of the outputs of sensors on satellites, vehicles, or robots where either automated actions (e.g., guidance, manipulation) or analysis (e.g., crops, minerals, weather, troop movements) are desired. *** Market, price, and general economic information in applications involving any kind of purchase and sale transactions. *** In game applications (e.g., electronic games, sports, racing, military), information concerning player characteristics and current game position could be used to produce intelligent actions, develop strategies, etc. (2) The second requirement is that POSSIBLE ACTIONS be well defined. This requirement of INTRONS is no different than any other automated system. Examples include specific signals to guide a craft or robot arm; specific results of analysis, such as "the crop below the satellite is wheat" or "the weather will be cloudy in X location"; buy or sell X commodity, currency, equity, or real estate; change to pitcher X in a baseball game; bet on horse X in the fifth; move X number of Y type ships to Z location; increase bank reserves; decrease inventories by X amount; increase or decrease production; perform X surgery; etc. (3) Third, the SUCCESS OR FAILURE of the action must be well defined. INTRON technology can be used when it is known what is success or failure in the particular application. Examples include: if wheat is indeed the crop below the satellite when the INTRON says that this is wheat; whether or not any type of INTRON-produced forecast proves to be correct or not; whether a purchase-sale transaction resulted in profit or loss; whether or not the game, battle, or war was won or lost, etc. Note that this success/failure feedback is not required at all times. In some situations, the feedback information would not be always available. Hence, once INTRON-manned satellites exceed the performance of existing systems (the learning process) in pin-pointing troop movements in practice maneuvers in New Mexico, for example, similar pin-pointing anywhere in the world can be used with a high degree of confidence. (4) Finally, the METHOD or procedure for using the available information must be NOT KNOWN, or not clear, or if known, may change at any (unexpected) time. What situations are these? These are the common situations which decision makers must face daily. However, this criteria may sound strange. Consider the opposite. If the manner of problem solution under all conditions were known, and if the actual condition at all moments was also known, then neither INTRONS or human intelligence would be necessary. The solution could simply be programmed and used with confidence as is the case in artificial intelligence applications. But in those applications where this cannot be done for whatever reason, then INTRON technology should be tested. It has been mentioned that the proper method may not be known or that the method which "works" may change in a rapid or unpredictable manner. Other reasons why a fixed problem solution may be difficult, costly, or impossible to program on computers include the presence of an intrinsic random element in the task (breaking news may cause market prices to suddenly fluctuate; a volcanic eruption may suddenly destroy the credibility of carefully developed analyses of satellite sensor information; a pinch hitter might strike out; etc). In other words, INTRON technology should be considered in situations where "human operators" are currently employed because the complexity of the task or the necessity of adaptive responses to unanticipated conditions requires this level of intelligence. Humans can be replaced by INTRON controlled robots where there may be a high element of risk (operating tanks, constructing space stations, purchase-sale transactions of large dollar amounts, etc) and/or because INTRONS perform the task better or at lower cost. To summarize, the INTRON is physically a device to which four types of lines (e.g., wires) are attached: (a) STIMULI -- the incoming information (b) RESPONSES -- the outgoing control signals which produce actions (c) FEEDBACK -- inputs of the success or failure of the RESPONSES, (d) FUNCTION -- input signals which indicate which of several function should be performed at any moment. Two of these Function (or Clock) inputs cycle the device through four functions, which are: 1. Read (input) incoming Stimulus information. 2. Generate outgoing Responses (output). 3. Read (input) the Feedback caused by the Responses. 4. Learn is an internal function (no inputs or outputs). HOW ARE INTRONS USED? First, a brief note to the designer-engineer: Use of INTRONS in new applications simply require the proper interfacing of signals from the real-time environment. Note that INTRONS do not require computers (although most computer systems will probably include INTRONS); namely, the INTRON I/O can be connected directly to compatible input and output devices. A simple external clock circuit is used to synchronize INTRON functions with external events. These functions should be clocked cyclically 1-to-4. In some instances, abbreviated 1-to-2 cycles can be used. However, this shortened cycle should not be used excessively since conditions (or the dynamic relationships in the situation) might change. Remember, this is one of the major reasons for adding INTRON CONTROL to the product or application in the first place. The INTRON discovery implements a process of a very general nature. Considering the history of similar general developments, such as the ability to transmit information via electromagnetic waves (e.g., radios, TVs, satellite communications), or the ability to perform calculations electronically (e.g., the computer, and now microprocessor chips), it seems that appropriate applications will be numerous. Some plans for the INTRON are presented below. Perhaps the reader will use imagination concerning potential applications where INTRONS will probably be used, such as: *** In any device which presently contains a computer or microprocessor (CPU on a chip) or other large-scale integrated circuit chips. You name it -- video games, home computers, all sorts of automated devices, calculators (which will be "super-problem-solvers"), automobiles, airplanes, scientific and medical equipment, weapons systems, and so on. *** In any object sent into space. Here the economics is simple. The cost of putting any object into space is high; therefore, one wants to get the best return possible on the performance of devices sent into space. Indeed, it is not difficult for the lay observer to appreciate that INTRONS will probably be managing more that 90% of the critical decisions made by such devices and that humans in space will largely be supervisors, "riding herd" on INTRON robots. *** In any work site where experts use computers. Since INTRONS can also use computers, a large number of computer operation functions which require critical decisions can be farmed out to them. Many of these applications would be transitory in nature (perhaps a decade), since it is difficult to imagine how computer manufacturers could avoid the economic pressure to incorporate INTRON CONTROL in their products. *** In situations where the sheer volume of data is so great that it is difficult to extract the relevant information where practical decisions must be made. Such situations include economic forecasting, logistics of materials shipment, materials procurement, foreign currency hedging, analysis of data collected by satellites and space probes, weather forecasting, military maneuvers, medical diagnosis and treatment, marketing decisions, management of monetary aggregates, etc. *** In applications where practical decisions must be made RAPIDLY which means, of course, that failure to do so bears a heavy cost. Examples are purchase-sale transactions, military operations, medical practice, etc. In such applications where computers are often used to increase speed, one often finds various models (often statistical in nature) designed to assist in the decision-making process. The problem suffered by these installations is that their investment has been justified by the importance of the decisions to be made, but the computer model is too often one step behind the real phenomena being tracked. Consider that such models usually incorporate values (called weights) which determine the decisions produced based on previous analysis of the predictive value of the variables used. What happens is a double knock-out, something not allowed even in boxing. First, the experts who developed the model do not realize that these values in their programs must be changed (because external conditions and relationships have changed) until "it is too late", that is, until the model performs so poorly that it is clear that something is wrong. And second, by the time the experts redo all of the analysis with more current data and inaugurate their "improved model", there is absolutely no guarantee that the external system being modelled has not undergone yet another change. Thus, it would be surprising if these installations were up to date in real-time decision making. These types of computer facilities are supremely suited for INTELLECTRONICS. Now the experts are not to be faulted; indeed, the very best scientific and statistical expertise is usually present in those facilities where the decisions "really matter." Instead, this is PRIMARILY A MANAGEMENT RESPONSIBILITY to provide these people with the very best in technology to work with. And it is precisely this kind of decision-making situation for which INTRONS were designed. THE DEBUT OF INTELLECTRONICS A general strategy for the development of an Intellectronics industry will be briefly outlined. Each element of this outline would lead to specific business plans and contracts for investors and clients. (1) INTRON CONTROL CENTERS (ICC) -- ICC's are centralized locations containing INTRON arrays connected with INTRON users via an international communications network. These centers receive the required information from client users or other sources and provide the INTRON-generated decisions (actions) to the users under contract. ICC's provide a organizational structure with which literally hundreds of specific INTRON-application projects could be rapidly conducted. Other advantages include low initial capital investment, immediate cash flow for the INTRON CONTROL services provided, protection of the proprietary secrets behind the INTRON since the INTRON arrays will be maintained at secured facilities, protection of the proprietary or national security relevance of incoming and outgoing signals, and rapid appreciation by potential users of the often stunning levels of intelligence of which INTRONS are capable. At each ICC, the INTRON arrays themselves are maintained under maximum security separate from usual staff and administrative functions. An important point for users is that there is not necessarily a need that the ICC have any knowledge whatsoever of the type of decisions involved in a particular application. That is, the INTRONS work in an abstract world of numbers; what these numbers in fact represent may be entirely irrelevant to successful INTRON CONTROL. This means that government and business can test INTRON CONTROL in their own most sensitive applications without anyone operating the ICC knowing even the broad nature of the task. In such cases, ICC personnel provide orientation seminars so that users can prepare their data in an acceptable format (including feedback). Such users can find comfort in that if such maximum security is desired, the ICC would not even know if INTRON performance was better than competing methods (except that contracts were not terminated). Computer Intelligence Assets (CIA) has generated many ideas for applications. The first ICC was established at a CIA facility in San Juan, Puerto Rico. This ICC started with a specific application, which meets all of the criteria described above. This application system is the SUPER-TRADER, a device which contained an array of 180 INTRONS. The Super-Trader project used market price and general economic information to trade commodities futures contracts (currencies, financial instruments, metals, foods) on major exchanges. For background information on this project, the document, "The Super-Trader: An Invention in Robotics," may be obtained from CIA. In short, the ICC phase of the Intronics industry provides INTRON CONTROL services under contract. The market potential is tremendous. Some of CIA priorities are: --- Corporate currency hedging. --- Bank financial instrument hedging and loan portfolio decisions. --- Corporate material procurement and inventory management. --- INTRON analysis of all types of satellite sensor data. --- A variety of forecasting services, perhaps including economic indexes, interest rates, and weather. --- INTRON guidance of various types of vehicles and robots. (2) THE INTRON-ON-A-CHIP -- While phase 1 above focusses on service, the second phase is manufacturing. The technology is available to place an entire INTRON on a very large-scale integrated (LSI) circuit chip. This, of course, would be a fundamentally new product (unlike the microprocessor-on-a-chip which was simply a dramatic reduction in size of a device--a CPU--that had previously been in use for years). While this phase would be somewhat more capital intensive, the risks would seem to be minimal especially if the ICC's were widely used demonstrating the utility of the product. In light of this decreased risk and of the potential returns, it is not too early to begin specific planning of this manufacturing phase of the Intronics industry. It is obvious, for example, that the INTRON-ON-A-CHIP would experience the same competitive advantage as well as open the door for entirely new applications as did the microprocessor-on-a-chip. The general approach here is straightforward: (a) Engineering drawings showing the connections, functions, power requirements, and timing diagrams of the INTRON chip would be provided to all major companies manufacturing electronic equipment. These drawings find their way to the designers in those organizations. New product technical specifications are greeted like the tooth fairy. In other words, a process will have begin from which there is no turning back, because these individuals will dream up numerous new products or improvements of existing products. (b) Meanwhile, a prototype INTRON chip is produced. Thus, as potential purchasers of these chips are weighing these designs, many will reach the prototype stage which would require an actual INTRON chip for final testing. (c) What follows is largely a question of management of production and distribution. In conclusion, the INTRON chip would permit the construction of high density Super-Intelligence in very small light-weight spaces. CONCLUSION This document introduces INTRONICS, the world of INTRONS. Perhaps this general description will prompt interested readers to explore further the wide variety of issues, questions and possibilities which it may have raised. For equity investors, the INTRON would appear to be a very valuable property indeed. For potential users of INTRON control services or INTRON chips, specific discussions leading to business agreements are appropriate. Prospective future users may wish to explore a number of test applications to determine if INTRONS do in fact perform superior to any competing technology in terms of both technical and economic criteria. NOTICE: Patents and Trademarks pending for INTRON, INTELLECTRON, INTRON CONTROL CENTER, INTRONICS, INTELLECTRONICS, INTRON ARRAYS. WHY WAS THE INTRON INVENTED? Presumably, the answer to this question resides in part in the presence of the proper ingredients in one place at one time. The inner mechanism of the INTRON resulted from a synthesis by Dr. Keene of concepts, experience, and technical information in the following fields: --- Integrative Brain Functions --- Mathematical Formalisms used in Quantum Physics --- Computer Automation --- Artificial Intelligence --- Design of original Analog and Digital Devices --- Systems Analysis and Multi-Variate Statistics (Information on the author, including publications in internationally recognized scientific journals, is available upon request from Computer Intelligence Assets.) (C) 1982-1991 James J. Keene, Ph.D. This file may be duplicated via ECHO, uploads, downloads, etc, but only if it is transmitted unaltered. Thank-you for your interest! Please address comments & inquiries to: Computer Intelligence Assets 130 Hiawatha Trail, Medford Lakes, NJ 08055 USA phone 609-953-7205 voice/data/fax (call voice first to set param's).