Digital Life-Form™ Simulation Technology and Q-AL Assistant™

The Q-AL Assistant architecture is to the robot what Windows® is to the PC. Windows makes computers much more useable, and the Q-AL Assistant architecture does the same for robots and computer systems that act like robots.

Most computer literate business people would agree that a Star Trek™* like computer or robot would be a valuable technology to have, and that it would form the basis for a very profitable business as well as have many scientific uses. While it is not possible for computers to be conscious like real life-forms, a simulation system that mimics the teleological (goal-directed) causality of a life-form as it "identifies" reality can simulate some aspects of consciousness, just as humanoid robots mimic some aspects of human form and behavior. You can think of "virtual consciousness™" as a new aspect of virtual reality. (More on this below)

Before discussing the details of how such a vision can be accomplished technically, however, it is important to understand how DLF Simulation Technology™ and Q-AL Assistant compare with with various other technology architectures, including our own Q-AI Expert™ architecture.

(NOTE: We are aware that a fully operational version of our DLF simulation techology is some years away, so we are looking for other less difficult implementations to help us prove and develop the technology, such as in the field of computational epistemology. One thing DLF Simulation Technology can do very well is build object abstraction hierarchies (also known as ontologies). This means DLF technology may have applications in the domain of Web software development for the so-called "semantic Web" to automatically abstract Web object classifications from XML object properties, for example. We first considered this domain in 1998, but found little interest then, as the focus was on Web functionality, not on semantics. Now, however, with articles on the topic by people like Hendler (See http://www.cs.umd.edu/~hendler/AgentWeb.html ) and the introduction of manual tools such as the OWL Web Ontology Language (See http://www.w3.org/TR/owl-features/ ), new opportunites may be developing.)

Unfortunately, state of the art Artificial Intelligence (AI) software does not work the way it does on Star Trek, even after many years of attempts to make it work that way. More advanced human interfaces that use ordinary or so called "natural" human language have been the "holy grail" of the Artificial Intelligence community for over 30 years. To date, however, the successes of making computer systems and their interfaces seem human-like because they can communicate verbally have been very limited. At Blue Oak, we believe that part of the reason is some confusion about how natural human language works on the part of our competitors, other people who design the current state of the art systems. We believe this situation is the consequence of too much emphasis on the analysis of existing languages and on computer technology, and not enough emphasis on the observation of how life-forms actually operate.

In fact, this is precisely the reason why our expertise brokering Q-AI Expert technology is designed as a hybrid system that intentionally includes the real, biological, conscious intelligence and common sense of actual people to make sure it works properly. State of the art computer technology is simply not yet as practical and effective as it could be at solving many of the problems users face (without a real person as part of the system design to help them). Anyone who has been frustrated by automated telephone answering systems, Internet search engines, or web pages that lead to a labyrinth of graphics and confusing forms that sometimes loose their contents for no apparent reason can attest to this fact. So far, the only real, operating source of intelligence is the human mind.

Probably one of the best example of the state of the art in AI systems is a system called Cyc, that is being developed by Doug Lenat (See the January 2002 issue of Scientific American, page 18, The World in a Box by Lamont Wood, or go to http://www.sciam.com/article.cfm?articleID=00063887-5C1E-1C6D-84A9809EC588EF21&pageNumber=1&catID=2 ). The bottom line is that after twelve years of hard work by some of the best AI systems designers and programmers, the creation of a database containing 1.5 million facts, and system development costing over $50 million, the best Cyc can do is to serve as a software utility to help improve the quality of the data retrieval of the Lycos Web search engine system. But at least Cyc is to some extent based on the observation of human intelligence and reality.

The latest trend in AI is to ignore reality altogether and assume that consciousness is a completely subjective phenomenon. Further, that simulating consciousness and natural language is merely a matter of programmers, quite arbitrarily, simply making up the right code as this quote explicitly asserts:

"Reality is not a discovery, but an invention - a construction based on experience and knowledge. Memory never brings back reality. Memory reconstructs. And each reconstruction changes the original, creating new frames of reference that inevitably grow apart from the truth. ... If radical constructivism is right, which we believe, then knowledge and experience are essentially subjective, and although a person may believe that his knowledge and experience are no different from anyone else's, there is no way he can be sure they are the same. Much less will he be able to know whether his brain represents knowledge and experience the same way they are represented by anyone else's brain (perhaps neuroimaging will make some headway in the matter, but we have our doubts;) in this regard, the interpretation of language is no exception."

This quote (from page 8 of the article Pibot vs. Turing (Acts I - III) at the following URL) is an excellent history of AI, but it also will not be a likely path to success because too much attention is paid to technology and not enough to the biology of consciousness:

Hyper Dimension Pibot Technology Assessment

These may be valiant attempts, but so far, none has functioned like a Star Trek computer. This is not to say that our technology has either, but we offer a completely different theory and approach to many of the problems in these fields, an approach that offers many new ideas that the other theories and approaches do not include. Only time will tell if our ideas will succeed.

Nor has Artificial Life (AL) been successful in simulating consciousness. The alternative technologies of state of the art Artificial Life software have not produced practical solutions. AL technologies digitally simulate only low level animal functions such as locomotion, groups of imaginary animals interacting to carry out virtual biological experiments, or the simulation of entire ecosystems based on some biological model. Extant AL technology does not simulate goal-directed, conscious behavior as it is observed to operate in individual animals and people. AL systems are not designed to simulate an intelligent entity that is capable of independent, rational action to accomplish its own goals in a purposeful way.

In fact, there is no technology that can functionally mimic goal-directed behavior, the sense perception of real world objects as sensed by animals and people, originate independent actions, and communicate using ordinary human, Natural Languages (NL) such as English, French, Arabic, Spanish, German, Japanese, Russian, and so on. Until now, that is.

____________________________________________________________________________________________________

See our original White Paper summarizing the new approach and theory we bring to these problems, and also our FREE BOOK offer below. In addition, we now have two new White Papers that summarize some of the ideas that are fundamental to and provide the theoretical foundations for our approach to simulating consciousness and natural language. These papers also include some of the latest new research ideas, so study the references too:

An Inductive, Biological Approach to NL and Math

A Call-to-Action: Now is the time to Build a Robot with Simulated Consciousness and Natural Language

____________________________________________________________________________________________________

Our technology architectures can provide solutions that work now, solutions for real customers. Q-AI Expert technology works because it uses real human consciousness and common sense to supplement whatever level of computer automation you use in your current or planned application designs. This means that human consciousness and common sense in the form of a real, human expert is available when and where necessary to help users get what they need. The reason we call our technology Q-AI (short for quasi-AI) is because it is a hybrid system, the goal of which is not to completely replace people with automation, only partially replace them.

The goal of DLF Simulation Technology, on the other hand, is to replace people where it is necessary and practical to do so. DLF technology is a general architecture that simulates life-like, goal-directed behavior and human consciousness to offer other solutions where using real human consciousness does not make sense. State of the art AL systems are not teleological (goal-directed), whereas DLF technology is a new kind of software architecture that is. So DLF Simulation Technology is a hybrid system that can be called Quasi-Artificial Life or "Q-AL" for short.

Q-AL Assistant is a subset of the DLF technology architecture that is designed to be an interface persona we call Quasi™, an interface persona that is focused specifically on simulating an intelligent assistant to become part of a computer operating system, a robot, a computer game, or military hardware. At Blue Oak, we believe that the simulation of life-like functions can be achieved to only a very limited degree by human programmers who attempt to anticipate every fact and action capacity a system will ever need (like Cyc or Pibot are designed to do). We believe this because real life-forms do not work this way, and yet biological life functions much more effectively than machines do. The fact is that no one "programs" biological life-forms, yet they exist, survive, and act independent of human desires, technologies, and purposes quite effectively and have done so for millions of years. Our premise is that to be as effective and useful as biological life, life-form simulation systems must more closely mimic what real life-forms do than what computers do, and AL systems must learn about reality for themselves---directly, not from programmers. Digital Life-Forms must program themselves for the purpose of earning their own survival.

Q-AL Assistant is an entirely new kind of goal-directed simulation architecture that is more narrowly defined than DLF technology and can add the capabilities of life-like behaviors to an ordinary medium power PC, behaviors that would enable a PC (or robot run by one) to act like it perceives the world in a manner similar to the way a person does, and to communicate with its users by means of simple sentences in ordinary human (natural) languages. Initially, Q-AL Assistant systems will not be an expert at whatever they are designed to do, but like young life-forms, they will be capable of learning about their world, the knowledge and behaviors expected of them by their human users, and they will be motivated to do so by a simulated pleasure-pain system (and ultimately, the "pain" of simulated death).

Any PC or other computer using the Q-AL Assistant architecture would be easier to use because once it has been trained to a user's specific needs, one could simply talk to it like the computer on the Star Trek TV programs. In addition, a robot system using Q-AL Assistant technology would be capable of independent action based on its own, self-learned knowledge of the world, and so be able to act more like the Mr. Data™ character from Star Trek than existing systems can. The bottom line is that even a fairly early stage Q-AL Assistant simulation would have about 5-10% of the "mental" capabilities of the Mr. Data character. A simulation system such as this would make a valuable PC operating system assistant, PDA, or robot, depending on its form as a product.

(NOTE: This is not a real implementation of the Q-AL Assistant architecture, just an example of how Quasi might appear to a user on an operating system that had implemented it.)

See our FAQ for Potential Investors and Corporate Partners

DLF Simulation Technology and Q-AL Assistant are not the same as human consciousness, they are merely a technical means to simulate it. A Digital Life-Form mimics the form of some human conscious functions, just as a department store mannequin mimics human physical form. Only biological life-forms are really conscious, but a computer simulation system that is programmed to mimic the goal-directed and identification behaviors of life-forms as it interacts with reality can mimic certain aspects of life and consciousness. You can think of DLF Simulation Technology as reality-based computing (as opposed to the arbitrary computing found in state of the art computer systems or AI systems like Pibot).

This is an important distinction. Part of the reason for the limited success of AI is that there is a widespread misconception in our culture that sufficiently advanced AI computer systems can somehow become conscious on their own and use language in the same way that people are conscious and use language, if only the right kind of computer program could be written. This view has been put forth in many popular science fiction stories and movies, but it is false.

Computer systems are not alive, cannot be conscious in the same way as biological life is, and programs certainly cannot think. Computer programs simply operate like falling dominos, mostly using data made up arbitrarily by programmers or users (and stored in files) for their content. If reality is sensed at all, the data that results are used just like the rest of the arbitrary data in the system. It is not used for identification of reality by the computer system sensing it, from the perspective of the computer system itself.

Strictly speaking, computers do not even calculate. Computers are just machines, electromechanical automatons that change the electrical properties in their components ("memories"), parts of which represent "1" and "0" bit values. Computers change their bits and the output of their display mechanisms according to the laws of physics, mathematics, computer science, and what people type into them. But these bit values, and all the programming that is built on them, only have meaning to people, not to the computers themselves.

Only consciousness as it is observed in humans can calculate. Consciousness is an an active, relational process of interaction with and identification of reality that is an attribute of biological entities, such as some kinds of animals and people. Of all life-forms, only people can calculate because only people possess the kind of consciousness that has that ability as part of their identity. Only people possess a consciousness with freewill that can perceive objects and count objects in reality selectively, that can choose to focus on some of the objects' attributes and ignore others, form concepts of objects and of numbers, eventually abstract the principles of mathematics, and then use those principles to attain human goals---such as building computer systems to automate human calculation abilities.

If computer systems are to interact with humans using natural language in other than a preprogrammed, predetermined way like state of the art systems do, they must sense and use reality-based data (as opposed to arbitrary data), and both consciousness and concept formation must be simulated to enable the functionality of this type of interface. Since consciousness is an attribute of life-forms, the conditional nature of life processes must also be simulated to functionally enable the consciousness simulation and provide it with the motivation to act. Otherwise, the simulation system cannot operate in a realistic manner. The conditional nature of life processes is what provides biological life-forms with the motivation and independent ability to act, to do what they need to do in order to survive. Any realistic simulation of consciousness must operate the same way.

Working from that premise, Blue Oak has developed a radical new Quasi-Artificial Life (Q-AL) technology, a new kind of virtual reality (virtual consciousness), to simulate some of the unique aspects of life-forms, including many of their conscious behaviors, such as sense perception, freewill, concept formation, and the use of natural language.

Our new technology simulates goal-directed behavior, as opposed to the simple mechanistic behavior found in state of the art AI and AL systems. It does so by emulating the conditional nature and more complex causality observed in real life-forms, as opposed to simple mechanistic processes.

The thrust of Blue Oak technology is to create simulated Digital Life-Forms™ (or DLFs™ for short) which, while based on and animated by the mechanistic automatons of a computer system, simulate the more complex causality of conditional, goal-directed, self-generated, self-sustaining behaviors identified with biological life-forms. These behaviors depend on continuous action. (Dead life-forms cannot act.)

DLF Simulation Technology™ will make possible a wide variety of simulated life-forms of many types, and Q-AL Assistant technology will make possible improved natural language interfaces for computer systems, enable animated virtual reality characters to communicate with people and each other, enable robots to engage in more independent behavior than is possible today, and lead to more powerful digital assistants that are much more flexible than those available now. (See the related website: http://www.robotbooks.com/ )

Moreover, products based on DLF Simulation Technology and the Q-AL Assistant architecture can be developed in self-demonstrating stages and for a relatively small investment compared to developing a new computer operating system or large application program, for example. Demonstration code already exists for simulating goal-directed behavior, perceptual consciousness, and concept formation (the basis for natural language). While this all sounds complex, it is really just the identification of objects by the simulation system, the things the system "sees" from its own perspective. The things the system "sees" are in a new form that is calculated from bitmaps in order to make recognition and other higher level processing more efficient, instead of just using the bitmaps directly as most state of the art systems do. This approach also makes the system's simulated "experience" (its simulated percepts and concepts) more similar to a human experience, so what it "sees" matches more closely to what a person sees (from the system's own perspective). Finally, much of what is needed to simulate the higher level functions of consciousness, such as concept formation and natural language, can be easily calculated by processing the large database of simulated percepts a DLF system will build up as it explores the world.

(How all of these straight forward processes work is explained in detail in the book described below. Our new white papers explain the large body of epistemological technical theory on which the book and white papers rest. See the extensive list of references in the book and papers, most of which are not widely known.)

In the near future, DLF Simulation Technology and the Q-AL Assistant architecture will do for today's tiny computers (and other products that contain them such as robots) more than what the windowing interfaces did to improve computers with command line interfaces. Though the technology sounds complex when you first read about it, it is actually quite simple, easy to understand, straight forward to develop, platform independent, scalable, and relatively inexpensive to make into a developers kit for companies to use to make specialized products. For example, think of scientific or medical instruments or special purpose robots that can see and identify the same objects you and I can, use natural language to communicate about the objects, and are capable of self-generated, self-sustaining actions to achieve goals relating to the objects. Or, think of a virtual reality world in which simulated personalities behave like independent people. Both of these examples and many more are possible with DLF Technology and the Q-AL Assistant architecture. A patent is pending on this new computer simulation technology.

 

To Read Our Free Book On-line

Our free book describing how Digital LifeForm technology works and how to build your own consciousness simulator is now available at this web site for on-line reading:

 

How to Simulate Consciousness Using a Computer System

 

 OR

 

You may: Download Your Own Copy of Our Free Book

 

 

To License Our Technology

If you wish to apply for a license or have questions about our technology, please mailto:greg@blueoakmountaintech.com, and we will reply to you as soon as possible.

Please include your Name, address, phone number, title and organization (if any), and your interest in the technology, such as education, scientific research, product development, and so on.

Back to Top of Page

< Back to Services

< Back to FAQ

© Copyright 1998 - 2008, Blue Oak Mountain Technologies, Inc. All Rights Reserved

Windows is a registered trademark of the Microsoft Corporation.

* STAR TREK and related marks are trademarks of Paramount Pictures Corporation.