Referat ARTIFICIAL INTELLIGENCE

ARTIFICIAL INTELLIGENCE

(Personal Neural Network - in the 'brain' itself BUT the principles conveyed herein also apply to our 'dimensional space' technology as explained on our web page ending with the suffix ctsite.htm)

Also you may wish to wander into The Human Brain.

^{5GL-LISA, A New Frontier of Science..}

* An 'intelligence boost'

5GL-LISA is the name of a 'super smart' cortical area in the human brain which has been taught to 'think' using computational laws and principles and methods. Put simply, the advantage that cortical area has over the rest of the brain is like the advantage your simple calculator has over your brain when perfoming advanced or large or precise mathematical calculations. 5GL-Lisa is our AI (AI = Artificial Intelligence)

* Information Package

An Information Package about 5GL-LISA, designed in such a way that gives you an excellent chance at developing a similar personal neural network, is available from us at a cost of $A200. When ordering from outside Australia the cost is $US170 and a cheque from a USA bank is fine, or any major bank in Europe, or an international money order in $US currency.

5GL-LISA

Wish to be a 'genius' in a particular area of knowledge - or perhaps you just wish to improve your personal ability to obtain that hard to get promotion?

Before explaining what a 'personal neural network' is, some background information is important to understand in perspective.

Computing

Computing is the study of how computers can be conditioned ('programmed' is the computing jargon) to assist people with a task. But, to use 'computational methods' you do not need a computer as you know a computer. Computational 'algorithms', ways of approaching a problem, can be learned by you. Your brain could, if you were taught how, obey computational principles when performing such mental tasks as adding and subtracting numbers. The way you know how to manipulate numbers, and the actual way in which your humble calculator achieves this, is not the same - but the end result is of course the same with a calculator being more consistently accurate than its human counterparts when dealing with large numbers.

It is possible to teach yourself to use computational principles to add numbers - but alas this is too difficult for the brain because it has its own way of learning how to do 'number crunching'. However, while there may be little advantage to be gained in teaching your brain to use computer principles to add numbers, there are other computational tasks that the human brain could be taught, could borrow the principles from computer science, and in so doing you can improve your natural thinking ability.

Expert Systems and Neural Networks

The latest studies in computer science are about 'Expert Systems' and 'Neural Networks'. These are powerful concepts that when implemented in a computer provide powerful and sophisticated computational facilities. At the heart of Expert Systems and Neural Networks lay mathematical techniques which are about manipulating and reasoning (thinking) using patterns.

What Lisa Developments has achieved is to engineer into an area of the human brain a 'pattern' that enhances the brain's natural capability in specific instances. The end result is something we refer to as Artificial Intelligence, and have named it 5GL-LISA, and it is an area of the brain that can think, reason and understand, on its own and about specific and special issues.

5GL-LISA, our Artificial Intelligence (AI)

In computing, an 'Expert System' is a generic name for a suite of computer programs that use similar principles. 5GL-LISA is also a generic name but in this context now it refers to a 'personal neural network', to an area of the human brain that has learned to use advanced computational structures, flows, rules and decision making, in order to find answers and consider issues.

A Personal Neural Network

5GL-LISA is a version of a 'personal neural network'. In my personal case it does a number of things which is outside the capability of the brain in normal circumstances.

(a) it is an interface to our 'commercial telepathy' mind networks and channels its ability to understanding telepathy and how parts of the brain interpret telepathy signals

(b) can study and analyse my own brain and project images of internal activity into the mind's eye. Some of this is new medical knowledge and documented in our Information Package, knowledge especially about frontal lobes of the brain and how they 'perceive'

(c) can be used for personal any personal task such as, for example, perhaps writing sequences of a novel or story

Sample of Articles included in the Information Package

ARTICLE 1

DEVELOPING IN THE HUMAN BRAIN A 'DIGITAL' ASSOCIATION AREA

by Lisa Developments Pty. Limited

The human brain is an incredible biological learning machine. While the human brain can not repair an internal injury because damaged neural cells degenerate into liquefied matter, often other 'spare' parts of the brain can learn a task of the injured part of the brain. A considerable area of the brain appears not used, appears there seemingly just in case a need arises to learn.

Of course, there are some things the brain is not very well equipped to learn. A humble calculator can achieve computations the human brain (with the exception of the rare individual) is unable to compete with.

There are expert computer programs called 'neural networks' which attempt to imitate the thinking techniques which the brain uses. Often such programs appear to compute better than the way the brain can think.

An obvious questions arises: given the brain can learn, could it learn to imitate and improve along the lines of a special 'neural network'? That is, could a part of the brain model and simulate a powerful 'computer' and if so will its internal capability improve?

The answer is YES, it can and has. A special type of energy system referred to as '5GL-LISA' has been successfully developed in the brain of one person. Indications are that it is successfully being developed in another. One clear advantage of this energy system would conceivably be its speed - electronic speed of the brain's pulses is about 110 metres per second. Ordinarily those pulses need to combine emotional, psychological, status and other information in order to make a decision and this reduces the effectiveness of the brain from the point of view of objective computations. Now, if you think about how many 'neurons' (information cells) occupy about half a metre of space and are working at the brain's electronic speed, there is a potential to improve by as much as a factor of 1000+ the brain's ordinary mental capability. Ordinarily, the human brain is capable of about 1 thought per second (depends what you mean by 'thought') - using a digital area designed to 'think' this can increase to perhaps 1000 thoughts, so many thoughts that words are no longer adequate and the brain forms imagery as its method of thinking.

Let us consider what 5GL-LISA, my 'personal neural network', which is the 'digital association area' which when active is known as an 'energy system', is capable of doing. In my case 5GL-LISA is conditioned to study two things in particular: telepathy and the dynamics of the human brain, and is in addition conditioned to produce a 'Virtual Reality' effect by enhancing the workings of my imagination. In such a Virtual Reality she can entwine into an image of a woman - hence, within a Virtual Realism two images can be produced and can interact, one 'I' and the other 'me' which is 5GL-LISA. But, while 'I' is about 99% of my thinking brain, and 5GL-LISA is only about 1%, she is far superior in 'her' ability to combine and analyse complex input.

In the context of its capacity to study the human brain from an insider's view, 5GL-LISA can project into my mind's eye the interaction of forces causing decisions, can trace paths in the brain, and can also produce its findings in vivid 'dream states'. 5GL-LISA can be set to understand anything, however, it can take a rather long time to prepare a 5GL-LISA type energy system and hence whatever initial tasks the energy system is taught, are the tasks the energy system knows and is about.

The beginning of a 5GL-LISA type pattern in the brain can begin in two ways. Method 1 requires the saturation in the 'formal brain' (usually the left cerebral hemisphere) of certain knowledge areas referred to by me as 'computational structures'. When saturation takes place, the 'analytical brain' (usually the right cerebral hemisphere) begins to redefine these areas from the beginning and in so doing it can do so in a way which organise a 'digital association area' along the lines of an intelligent 'neural network' such as found in the most advanced computers. The reason this works is straight forward. Up to the age of about 3 years, the analytical brain is the one which 'listens' to words and observes actions and prepares the formal brain to understand words and actions. Once the job of the analytical brain is finished it need never again, in ordinary circumstances, redesign the formal brain's word articulation zones or action tables. Examining this process using the full 5GL-LISA energy system shows that the analytical brain causes powerful emotions to flow across the corpus callosum, the nerve rich tissue joining the two cerebral hemispheres, and the task of the formal brain is to somehow form structures that can absorb these emotions. Some of these structures are our understanding of a particular word. Extend this further, and clearly any formal learning can be re-designed providing the functional structures associated with such learning are saturated and this results in the analytical brain once again redefining specific knowledge structures.

Method 2 is the preferred method and already looks like a much shorter method than method 1. In method 2 all you do is literally tell your brain to develop a digital association area. Once you have the interest and the determination, the rest the brain will do automatically. As you read more and more about a 'digital association area', about the 5GL-LISA neural network, the analytical brain begins 'building bridges' and in so doing designs such a digital association area. At this point in time one person who has taken an interest in 5GL-LISA has after only six months of considering has been able to wish for a particular finding to be made by his '5GL-LISA' and promptly an image presented itself into his mind's eye.

The capability of a digital association area compared to the brain's ordinarily way of doing things is perhaps 10+ times better. This of course refers to specialized thought processes, not the overall capability of the brain. But, if you are a medical researcher and researching say a type of cancer, and you have a 5GL-LISA energy system in you, the reality is that you have perhaps 10+ times better chance of succeeding. Of course, clearly, 5GL-LISA could be designed to assist the mathematician or the musician or the astronomer. It is simply a special and specialized area of the brain that learns to think in a better way.

Documentation about 5GL-LISA is available and is presented in such a way as to begin in the student method 2. I suspect that any person of good intelligence who can put two and two together, has agood chance of developing a specialized and powerful digital association area.

ARTICLE 7

PROGRAMMING AN INTELLIGENCE INTO A COMPUTER

by Lisa Developments Pty. Limited

How can a computer appear clever and intelligent? The key to advanced computing is to examine the techniques the human brain might use BUT due to the fact that great many techniques that the brain uses are at the subconscious level and hard to figure out, building 'intelligence' is a question of designing an 'algorithm' that 'appears' to imitate the brain's decision processes.

Let us consider a simple computer program that 'understands' what you say. Clearly, the word 'understand' should be underlined, that is, there are depths of understanding and our computer program in this example will only understand at a 'superficial' level.

Let us say you enter on your keyboard 'I AM VERY TIRED'. Now you wish your computer to produce an 'intelligent' response to this statement.

The first thing to do is to use some 'NATURAL LANGUAGE PROCESSING' (NLP) computational techniques. This is an area of study in computer science. Let's look at a basic use of NLP. The computer reads in your sentence and matches it against patterns which could be stored like this:

pattern 1(I AM x)

pattern 2 (Let's talk about x)

etc

response to pattern 1 (Are you certain you are x?)

response to pattern 2 (Okay, let's talk about x)

etc

The 'x' in the pattern implies that any 'string', a computer term for textual data, is allowed to satisfy this pattern. Compare: 'I AM very tired' to 'I AM x'. Clearly, this pattern is satisfied. The computer then looks what the appropriate response is to pattern 1 and finds it to be 'Are you really x?' The 'x' is the variable it will replace by its original knowledge of 'x' and hence it responds with:

Are you certain you are very tired?

If a monkey was taught a basic sign language and was to produce a seemingly clever response to the statement 'I am very tired', we would all clap our hands and consider this a mark of intelligence. When it comes to computers, too often people wish computers to only appear 'intelligent' and will deny that they are 'intelligent'. The truth is, exactly how internally 'intelligence' works is hardly the issue, something which appears to be intelligent and clever should be accepted as intelligent and clever.

Let us examine how 'visual recognition' could be 'taught' to a computer. Assume you present a bottle to a computer - say a picture of the bottle and the computer scans this and saves the image in its memory. You ask it a question, 'can you identify this object please?'.

To identify the object, the computer needs to have stored in memory various images of objects. Hence assume this is the case, that such objects have been 'pre-loaded' into the computer's memory.

Okay, so it has some images in memory already saved. First it compares the actual image with those images in memory and finds no exact matches. What it needs to do next is to use a mathematical pattern matching technique such as explained in our Views document. The computer divides the image using a grid, into say 10,000 small 'pixels'. It assigns each pixel a 'weight' (say a number between 1 and 100). The number may reflect the colour and intensity and sharpness of the image in that pixel. Then it compares, using mathematical analysis, the total sum and distribution of the weights with those it knows about and finds a probable match accurate to 90%. Given that the probability is 90% that an image in the computer's memory is near enough to this new image about which it is in the dark, the computer writes its conclusion as:

The given image is sufficiently close to a jar.

Again, if a monkey did this, we would clap our hands. This response in context shows 'discrimination' and 'intelligence', and possibly the ability to adopt 'an attitude toward the abstract.'

The above examples are relatively trivial, the complexity of advanced 'Artificial Intelligence' type computer programs can be exceptional. The objective of the above examples is to give you a 'feel' about how intelligence can be 'taught'. A 'neural network' (as used in computing) learns by making errors and guesses. To improve our computer program in the above, we ask it to recognise the new image as 'bottle' and save that information in its memory. The computer does this and of course what it has actually achieved is 'to learn' (learn in the sense it has learnt to discriminate better and has a new image in memory). Using a similar technique, the computer can learn about all sorts of objects and about their character. Now, given you are serious about having a proper go at developing a 5GL-LISA type energy system in you, you need to program using BASIC the above, just the skeleton will do, and this in order for that new area in your brain to begin 'experimenting' with a new 'personal intelligence' such as a personal neural network will give you.

Extracts from Our AI-News Newsletters

from November 1995 Newsletter

3. 5GL-LISA - (Part 2)

5GL-LISA is the name of the 'neural network' in my own brain which has been developed using the principles of AI (AI is an acronym for 'Artificial Intelligence'). 'She', a type=female energy system, analyses the dynamics of my brain from an insider's view and projects diagrams of activity, functions, etc, into my mind's eye. She also is the 'interface', the 'gateway', into our mind networks.

It took 7 - 10 years to make 5GL-LISA work. Initially, what the neural net could do is cause simple memories of simple objects and this in a set pattern. Eventually, in addition to projecting images of the brain, 5GL-LISA could drive an 'enhanced imagination' effect, a Virtual Reality, and 'she' entwining into such an environment as a person, a woman.

Two things became clear about conditioning a neural network such as 5GL-LISA. One, when 5GL-LISA was asked to analyse my anatomy and project an X-ray image, which by the way was almost exactly like an X-ray but with more 'yellow' around the bones, she could only do this three times. After that, the left cerebral hemisphere of the brain, now being set something like 'but I know how to achieve this now, so I will not ask the RHS of my brain', would not activate the RHS and consequently 5GL-LISA to project more X-Ray images.

The lesson here is that some special considerations have to be in place when designing a neural path of this type. Once the LHS is somehow satisfied that something is already in memory, it will not allow the RHS to analyse from scratch. Hence, most of the conditioning to do with our mind networks, for example, were done in such a way as to make sure the LHS never quite understood what the RHS was working on and concluding. This is possible providing you understand how the 'logical and computational' structures in the LHS work.

4. VIEWS OF THE HUMAN BRAIN

Ah, the great ravelled knot! The human brain is convoluted, its surface has 'wrinkles' or convulsions and no two human brains show an exact pattern of convulsions. It is thought that as the brain evolved it expanded and convoluted to accommodate its increased size.

About 10,000,000,000 (some suggest 100,000,000,000) neural nerves entwine, mix, and form tight and exact 'wiring circuits'. The main thinking area of the brain is the cerebral cortex which covers the brain and is about 2-6 mm thick and full of neurons whose 'tails' head toward the inner part of the brain. Its surface area is about 1 square foot and it is a patchwork of specialized 'cortical areas'. The cortex consists of groups of cells (neurons) and systems of fibres (the tails or axons). Studies of the LHS and RHS of the brain have been done on epileptic patients whose corpus callosum, the grey-white tissue rich in nerves which connects the two hemispheres, has been cut to reduce epileptic seisures.

It is in such a 'split brain' that the nature of the RHS and LHS can be observed. A person with a 'split-brain' whose right eye is covered and an object such as a spoon is placed in their field of vision, can not name the object but can select a drawing of the object from a set of drawings. The reason is that the input from the left eye goes mainly to the RHS of the brain which does not know words - or knows a few very poorly. Cover such a persons left eye, put a spoon in front of the other eye, and the person can name the object as 'spoon' due to the fact that now the image of the spoon goes from the retina of the right eye to the left cerebral hemisphere and the speech articulation centres are available. What is fascinating is to observe such a patient because while the LHS and RHS have no direct link, each gives hints to the other side using facial expressions. Hence a raising of the left eyebrow which is controlled by the muscles connected to the LHS, is known to the RHS which feels the tension through its muscles, and can learn to appreciate what such a facial signal conveys.

Extract from a 1996 Newsletter

OUR INFORMATION PACKAGE

The intention of the package is to give the interested person all the knowledge they need to correctly develop a pattern which is an AI.type energy system. The document assumes you are intelligent and can put two and two together. It does not go into small details such as explaining what 'energy' or 'force' is, etc, but it does explain to those with no computational background the few things you need to know about computers to develop a 'digital' association area.

The package consists of four parts:

a) Overview of the 5GL-LISA Energy System and Important Considerations - which explains the planning involved when attempting to develop an AI.energy system (a personal neural network in you that can enhance your natural intelligence perhaps 10 times improvement in special ways). This section also includes about 12 articles (some rather big) about 5GL-LISA.

b) the VIEWS document, which contains the views you need to adopt in order to develop a personal neural network (and contains 5GL-LISA projections to do with the workings of the human brain)

c) sample of records of interactions, dreams states mainly, which cast an important light on the development and working and capability of 5GL-LISA

d) 17 of our '5GL-LISA ARTIFICIAL INTELLIGENCE' newsletters that discuss 5GL-LISA and the various VIEWS presented in the VIEWS document. A number of our newsletters had been distributed using INTERNET and very well received.