humanAI.com

 
taskengine
timemachine
patents
notes
donation

 

 

 

                                  << content                              Chapter 9

5.  Other topics:

 

Simulated models 

Referring to FIG. 66, each simulated model comprises primarily three parts:  brain model, software data and hardware data.  Sequences containing all three parts are stored and organized inside the simulation brain and represented as a simulated model.  Intelligent objects such as cells, insects, animals and human beings have a brain model, however, non-intelligent objects like chairs, computers, videogames, phones, furniture, buildings and so forth do not have a brain model.    

FIG. 66

The brain model comprises the 4 different data types:  5 sense objects, hidden objects, activated element objects and pattern objects.  This will house all the data sensed from the intelligent object as well as its thought processes.  There is a sub-part called the personal model that stores behavior patterns for that object. 

The software data (FIG. 66) comprises hidden types of data or work done by intelligent robots.  One example of software data are electrical signals sent over telephone lines.  The electrical signals are the physical aspect of the signal, but the 0’s and 1’s that make up the signal is the hidden aspect of the signal.  The software data is the hidden aspect because it is “hidden” and can’t be accessed by observing physical traits.  For example, we can’t observe how the signal is transmitted to understand what that signal contains (the 0’s and 1’s). 

Work for that simulated model done by the intelligent robots in the time machine is also stored in the software data.  Work can be classified as any fixed tangible media, which includes books, computer programs, papers, computer files, holograms and so forth.  Work can be a computer program that the robots built to store, retrieve and modify data.  Work can also be stored in a computer file that contains schematic diagrams, pictures, videos, step instructions, knowledge and so forth.

As the robots predict that simulated model, it will store this work in the software data.  Work can be inserted, deleted, modified or merged and can be in any media type.  The robots working in the time machine will convert work into certain data type and insert them in a manner that is compatible with the pathways in memory (a simulated model is made up of pathways).  The more work is put into the simulated model, the more detailed that simulated model will be.  For example, if the simulated model is a human being, the robots have to predict each body part and how these body parts will be simulated in the computer.  This will go on and on until the individual cells are predicted. 

The hardware data (FIG. 66) represents the simulated model in a 3-dimensional manner.  Any type of physical data of the simulated model is put as sequences in a 3-d environment called a 3-d animation.  The 3-d animation is a sequence of physical objects that happen in a timeline.  There is no one camera angle to represent the 3-d animation.  A universal camera, from all angles, captures events or objects in sequence order.  For example, in a human being simulated model, the human being’s physical trait and actions will be the 3-d animation.  The human being’s external body and internal body will be stored in the 3-d animation.  All actions of the human being as a direct result of its brain activities will also be recorded in the timeline of the 3-d animation.  By the way, brain activities in terms of electrical discharges and how the electricity travels in pathways are known as hardware data.  The information inside the electricity is known as the software data.   

The material presented in this patent application related to the atom manipulator creates the 3-d animation for an object.  The 3-d animation is actually the clarity tree.  Here are the steps in creating the clarity tree:

 

1.  The signalless technology takes cameras capturing from different angles of the environment and a form of AI to track all atoms, electrons and em radiation from the current environment.  For simulated models, the same idea is used, whereby the intention is to use the signalless technology to store a 360 degree visual frame of the object we want to capture into pathways.  A 3-d frame works like a regular 2-d camera frame, but it is in 3-dimensions.  That means all internal and external atoms are tracked within a given focused area. 

The clarity tree (or 3-d animation) for a given simulated model have defined boundary areas on an object.  For example, a human being object will have only the physical boundaries related to a human being.  The boundaries are determined through the self-organization process, whereby similar examples are compared and common traits are found.  Sometimes boundaries are just estimates.  A boundary for a human being might include clothing.    

The clarity tree is based on how many times the simulation brain encountered this object.  If there are lots of data related to an object in the simulation brain, then the clarity tree will have many visibility levels.  If there are little data related to an object in the simulation brain, then the clarity tree will have little visibility levels. 

 

2.  Robots in the real world or the virtual world will analyze each visibility level and their conscious thoughts, in terms of words/sentences, will identify objects, actions and events.  These virtual characters have to do this for all visibility levels.  Things that the virtual characters say will have reference pointers to objects in other levels of visibility.  For example, if one virtual character is in the human visibility level he might say:  “that is a car accident”, the words car accident will be referenced to the data in the lower levels such as the molecule visibility level or the atom visibility level.  All data related to car accident in all levels will be referenced (FIG. 67).  

FIG. 67

Words and sentences to identify objects, actions and events in the clarity tree is important because language helps to organize data and to establish reference links for different visibility levels for each object.  The learned groups (words/sentences) will help the commonality groups (the physical aspects) to organize data further.  Automatic software to identify objects, events and actions can also be used.  A software can be created, whereby it looks through each level of the clarity tree to identify objects, events and actions.    

 

3.  Using external software to simplify certain objects, actions and events.  Hidden data are put into the data in each visibility level to help in identifying and grouping objects.  For example, if the object is ambiguous like the weather on Earth, software can be used to put arrows for wind direction/speed; and groups can be generated for strong cloud coverage.  These software simply makes it easier to delineate boundaries of objects, actions and events. 

The simulated models stored in the simulation brain are created by “work” done by intelligent robots working in the time machine.  They must define the brain model of the simulated model.  Things that the intelligent objects are sensing from the environment and thought processes must be predicted.  The hardware data (or 3-d animation) have to be predicted using tools like the signalless technology and the simulation brain.  Finally, software data that is needed to understand the inner functions of the intelligent or non-intelligent object must be predicted.   

 

Personal model and predicting the exact actions of a human being

Predicting the exact future actions of a human being is very difficult.  Learning human behavior in terms of pathways won’t lead to an exact future action of a human being.  They can help in aiding the predictions and giving probabilities of what might happen.  The only way to solve this problem is to formulate the personal model.  A simulated model has three pathway types:  brain model, software data and hardware data.  The person model is a sub-function in the brain model of an object (an object can be a human being or a table or a single cell). 

The pathways from the lifespan of a human being have to self-organize and pattern objects will emerge.  These patterns dictate the behavior of the specific human being – it is a personal model of that human being because this model is only concerned with how he/she thinks.  FIG. 68 is a diagram depicting how patterns are found between all aspects of the human being.  The physical body movements of the human being are compared with the mental thoughts of the human being.  Brain organs of the human being and how they behave will be compared to the 5 senses of the human being.  Thus, all aspects of a human being are searched and compared to find any pattern objects.  

FIG. 68

If you think about all the permutations and combinations of all data related to a human being, the outcome can run exponentially.  The only way to solve this problem is to use supervised learning and to emphasize which data should be compared first, next and last.  Data should be compared in a hierarchical manner.  Data at the top of the tree are compared first because they are easier to compare and their possibilities are limited.  FIG. 68 is a diagram depicting 3 hierarchical tree representing certain aspects of a human being.  Most likely, on the top levels of each hierarchical tree are words/sentences and simplified data, while the bottom levels are the detail data. 

 

Species of simulated models

The robots have to create the simulated models for all robot species.  All experiences of an intelligent object will be stored from the day it was born to the day it will die.  The robots also have to define the 3-d animation for each fraction of a millisecond for that intelligent object.  FIG. 69 is a diagram depicting the life-span of different intelligent objects:  a human being, a dog and an ant.  All their experiences will be stored in their respective simulated model and data on their 3-d animation will be filled in by the robots working in the time machine.      

FIG. 69

FIG. 70 is a diagram depicting the self-organization of different organisms in the simulation brain.  Notice that organisms are classified according to their species.  Human beings will most likely be organized with other human beings.  Within a human being, young men will be organized with similar young men and older women will be organized with similar older women.  Cats are more likely to be stored close to other similar animals such as a dog.  Organisms like bugs and ants are similar objects because of their size and shape.  They also sense data and act in similar manners. 

FIG. 70

The simulation brain also stores non-intelligent objects and also interactions between two or more objects.  A simulated model can be created for two objects that interact with each other.  For example, a human being can be one object and the other object can be a chair.  The simulated model can outline how the two objects interact with each other and how the interactions change each other in terms of the three pathway types. 

Of course, the more objects involved the more possibilities are stored in the simulated model.  For example, the human being and the chair simulated model needs to store “all” sequence of interactions.  This simulated model will be stored next to similar examples such as a human being sitting on a stool.  Universal pathways will be created so that a fuzzy range of simulated models can be generated.  An object can come in different sizes and shapes.  All human beings look different, but if many examples are trained, a fuzzy range can be created called a floater.  This floater will represent all human beings regardless of what they look like.  Floaters help to manage infinite data in the simulation brain by creating simulated models that has a fuzzy range of itself.    

In terms of the practical time machine, the intelligent robots that create the timeline for planet Earth has to use the simulation brain to do their predictions.  They have to extract simulated models of objects they want to analyze and predict events in the timeline.

 

Various methods to predict the future or past

This section will outline the various prediction methods that the intelligent robots will use to predict the future or past.  These are the most important prediction methods, my books outline hundreds of different prediction methods.

 

1.  Using human intelligence to plot out events in a fixed tangible media.  The most important aspect of predicting the future is work done by robots with human-level intelligence.  The robot can use various software and hardware to predict events in the past or future.  Investigators in CSI use human intelligence to solve crimes.  They collect information from the crime scene, analyze evidence, plot out the timeline in a computer or report notebook, discuss with other investigators about possible events and so forth.  These robots are no different.  The only difference is that these robots can work in the real world or in a virtual world to investigate events. 

2.  Using the clarity tree in the simulated models to plot out events in the timeline.  Events in the timeline should be plotted in a hierarchical manner.  The most likely events to happen should be plotted first.  Then, the details should be next.  The robots (or investigators) use the simulation brain to find out what are the most likely actions of an object.  The simulation brain has software that can search for information quickly and accurately.  The simulated models in the simulation brain are already structured in a hierarchical manner because of self-organization.  This hierarchical tree goes from general to specific.  This will give the robots an easier time to extract the possibilities of an event in ranking order.

3.  Using the personal model of a simulated model to give a more detail prediction of an event.  A simulated model is an average model of how an object should behave.  On the other hand, the personal model depicts a model of how that object will behave in a personal way.  For example, if the robot wanted to predict the future actions of a person7, he can extract the best matched simulated model from the simulation brain.  Based on what has already been predicted of person7, the robots can generate a personal model.  This personal model will give more details on how person7 will behave in the future.       

4.  Combining simulated models together and using human intelligence to plot events in the timeline.  Since the simulation brain can’t store “all” permutations and combinations of simulated models, the robots have to use human intelligence to determine the future events when multiple simulated models interact with each other.  The more simulated models the simulation brain has the easier it is to predict other events.  

5.  Using software to simplify and structure data in simulated models in a hierarchical manner.  The clarity tree structures data, most notably visual data, in a hierarchical manner.  The data goes from general to specific.  The AI in the signalless technology is used to generate the clarity tree so that it goes from general to specific.   

Let’s move our attention towards liquid.  Water is harder to track because the molecules slide from one molecule to the next based on force.  Water can only be tracked using a hierarchical tracking system.  A large lake is one area that water can be positioned and the water can’t leave the lake.  Within certain regions of the lake are smaller water regions.  Within these smaller regions are even smaller regions.  Liquid will be tracked hierarchically and in how they move.  Computer software will be used to create hidden data pegged to this hierarchical structure of water.  If you observe water from a satellite image, the water isn’t moving.  However, if you observe the same water from a camera, you can see the movements of the water.  The AI should track the water from a hierarchical visibility tree.  The AI might be able to track water movement from satellite visibility, but unable to track the water movement in terms of molecular visibility.  The AI can use the satellite visibility and human visibility and to guess the water movement for the molecular and atom visibility levels.  FIG. 71 is a diagram depicting the hierarchical structure of water and how the AI tracks water movements.

FIG. 71

 

Some methods to predict the future

1.  fabricating similar future pathways

Predicting things like creativity and rare events are very hard to do.  If the robots had to predict an artwork done by a human being, how exactly will these types of prediction be made?  If you observe comic book artists such as jim lee, marc silvestri, todd mcfarlene and rob lefield you will notice that each artist has a style of drawing.  Under certain story telling situations they draw in a certain way or they layout their characters in a certain way.  I have been collecting comic books for over 14 years and I can tell you from past experience that I can be presented with a drawing and I can tell people who drew that picture.  I can also predict what kinds of layout each artist will probable do.

The reason I was able to predict each artists artwork is because I have seen so much of their artwork.  If you look at a famous artist like Leonardo di vinci and observe all his artwork, there is a clear pattern or style to his artwork. 

The idea behind this first method of predicting the future is to generate similar future pathways of how an artist will create an artwork.  Let’s say that the robots wanted to predict a person writing a book.  They can put the person into slightly different situations to create the book and generate multiple similar future pathways.  The robots store the future pathways in a 3-d grid to self-organize common traits between all predicted pathways.  This will form universal pathways that will happen regardless of what the environment is. 

If I was to write a book 1,000 times and each book is written in a slightly different environment, there will be common things I will write about.  Maybe the exact words will not be used or the exact content will not be in sequence order.  However, there are common traits among the 1,000 books I have written.  These common traits might be the book is about time travel using AI, the book will outline methods to predict the future, the beginning of the book will be the introduction, the book will also have additional topics at the end, the overall idea behind the book is similar and so forth.  By generating similar future pathways, and determining the universal and rare events, the robots can better understand what are the universal events that will happen and what are the rare events that will happen. 

Let’s look at another example, when I was a teen I remember going to the park with my friend and playing catch the tennis ball.  For a time we threw the tennis ball back and forth.  On one of the plays, my friend was distracted by something that was happening on the road.  I accidentally threw the ball and the ball landed on top of his head.  The event where my friend was distracted and I threw the ball and it landed on top of his head is considered a rare event. 

If the robots have to predict the entire day I was on the park, how exactly will they predict the rare event?  How will they know that I threw the tennis ball 30 yards and it happen to land on someone’s head?  The answer lies in generating many similar future pathways and to establish relational links to each other.  If the robots generate 1,000 different future pathways there might be 5 pathways that have me throwing the tennis ball and it landing on that person.  The other future pathways might be similar events and they show that I threw the tennis ball and that ball came close to landing on his head.  By establishing relational links between similar examples the rare event may not actually be rare. 

We can also use these future pathways and compare them to previous events in life.  I notice in my life this rare event wasn’t the first time it happened.  I remember in high school I was playing basketball and I threw the ball from full court and the ball went into the basket.  In another event I made a bet with someone that I can throw a paper ball and it will land into the trash can.  I actually won that bet.  By observing my past and comparing similar rare events the robots can determine wither or not a rare event is actually rare. 

Great golfers are great because they have the pathways in memory to perform their job well.  That’s why people like tiger woods always do well.  He might slip up on some games but he always does well.  People who are sport players such as quarterbacks are also consistent.  They do well consistently and fans know how a player will perform in a game.  Some people even have sports prediction sites that will rank each player and why certain teams are more likely to win a championship.    

 

2.  Spaced out future pathways

Pathways can also be spaced out by having the robots plot out future events.  For example, if the robots wanted to predict the future of a baseball game it might be difficult.  Instead of plotting out the exact events leading up to the end results, the robots can predict the various possibilities of the end result.  FIG. 72 is a diagram depicting 3 future pathways.  Each is plotted with sentences to represent an event as a result of a baseball game.  A team can lose the game, win the game or quite the game.  There might be circumstances where the game can’t continue because of weather related conditions or a team refuses to continue the game.  These conditions are categorized into quitting the game.  So, these plotted future pathways are created because of common sense and logical analysis.  If you observe most of the simulated models for sport games, they already have these three outcomes stored in their pathways. 

FIG. 72  

Spaced out future pathways isn’t totally based on intelligent robots plotting future events, but are also selected from pathways in simulated models.  Imagine that there are 1,000 pathways to choose from in a simulated model and they are all equal in probability, the robots can use a form of AI to randomly pick spaced out outcomes.  Similar outcomes are excluded, the robots are only interested in a wide range of possibilities that are not related to each other.  An AI software can be created to extract certain pathways from simulated models based on a user’s preference.

 

3.  Cut, copy and paste future pathways

Sometimes, if a person does something in one area, they may not do the same thing in another area.  Other times a person may not do the same things in different times.  Space and time is very important to determine the appropriate actions for a person.  This prediction method would require the robot to cut out certain events from a pathway and change the place and time it will occur (referring to FIG. 73).  By having a wide variety of events put in different times and places, the robots will have a better idea of what are universal events and what are rare events.  If it is proven that an event is rare, these prediction methods can outline how rare it truly is.

FIG. 73 

4.  Determining similar traits between future pathways based on pain and pleasure

In addition to all the common traits mentioned above, the future pathways self-organize based on pain and pleasure (referring to FIG. 74).  Each object, event or action has their own powerpoints.  Some of these powerpoints are encapsulated.  The robots have to outline the powerpoints for each event, object or action in each future pathway and to establish relational links to powerpoints of other future pathways.  This type of self-organization is based on pain and pleasure.  If two events have the same pleasure, but both events are totally different they will still be grouped together.     

FIG. 74

5.  Simulating every aspect of an independent object into a software to determine its future actions.  One method to predict a random number outputted by a computer is to simulate the entire computer inside a software and let the simulation output the random number.  Any dependant factors that result in the random number must be included in the simulation. 

 

Conclusion:  all predicted methods mentioned above work together in combinations in order to predict the future with pinpoint accuracy.  These methods are used to outline universal and rare event so that the robots can predict very complex situations such as artistic expressions or coincidental events.  If an event predicted is considered rare these prediction methods can outline how rare these rare events are. 

  

Additional features added to the AI time machine

The AI time machine is an all purpose software machine that can do tasks for a user.  It can search over the internet to find information, answer questions, do individual or group tasks and so forth.  A list of features was presented in the beginning of this patent application.  Additional features of the AI time machine will include:  controlling dummy robots and controlling the atom manipulator. 

Dummy robots are simply robot shells that receive pathways to do tasks.  The AI time machine can use the universal computer program to assign station pathways to dummy robots to do individual or group tasks.  For example, 10 dummy robots are located in a car factory.  A user inputs instructions into the AI time machine to build 5 custom made cars.  The input media can be a software fillable form that takes in commands from the user.  After the fillable form is submitted the AI time machine will search for the station pathways that will allow the dummy robots to operate to make 5 custom made cars. 

The AI time machine uses the universal computer program to train itself to assign certain fixed interface functions to certain tasks. 

To make the AI time machine more efficient, the dummy robots are replaced with ghost machines.  The user can input the commands to build 5 custom made cars and the AI time machine will control the atom manipulator to create ghost machines to build the 5 custom made cars. 

The AI time machine will use the universal computer program to assign fixed interface functions to encapsulate work done by the atom manipulator.  The atom manipulator can build a house, write a book, solve a math equation, do research, do surgery and so forth without any physical robot.  Once the interface functions are assigned to certain work, a user can execute these work by accessing the interface functions.  It is very important the AI time machine goes through adequate training in order for these fixed interface functions to operate correctly. 

 

Additional capabilities of the ghost machines

Building physical DNA and single cells 

In my patent application called DNA machine software program, I describe how physical DNA is created.  With the help of the atom manipulator, it is possible to create physical DNA and single cell organisms.  We can actually build organic computers, cellphones, printers, cars, planes or aliens.  These single cell organisms will go through mitosis and develop into an adult organism.  In fact, we can design any type of dna we want.  We can design a human being with 8 arms and 4 legs or a human being with blue skin.  The various possibilities of design for dna can be unlimited. 

DNA is very small, but individual DNA strands are made from thousands of molecules.  If this atom manipulator can manipulate atoms, it can manipulate molecules even better.  Organic life-forms use 4 chemical bases as the foundation for the DNA’s genetic code.  We can build DNA using only 2 chemical base or 8 chemical bases. 

Existing organic DNA and RNA can also be manipulated to function a certain way.  We can design the cells to create anything we want it to create – grow back an adult arm or grow a child heart, cure genetic diseases and so forth.  We can also control the shape, size, and cell division aspects of the organic object. 

This atom manipulator is one level higher than conventional nanotechnology because we are able to build materials atom-for-atom.  We can control how materials are built at an atomic level.  This will allow the atom manipulator to build the smallest machines, smallest computer chips, strongest metals, 100% pure materials and so forth.      

 

No post office

Instead of shipping boxes and products through the post office, the atom manipulator can beam all objects from one location to a destination instantaneously.  When a person orders a product online, the company can ship the product in less than one second.  The atom manipulator has to fire atoms from the atom reserves layer to make a product.  The process goes like this:  the atom manipulator has to have a simulated model of a product in its database.  This simulated model contains a detail atom-by-atom specs of the product being shipped.  According to the simulated model, the atom manipulate will fire atoms from its atom reserves layer and bounce these atoms to their customer’s home.  These atoms will reach the customer’s home in “packets”, just like packets over the internet.  The atom manipulator will then create ghost machines to combine the atoms together, forming a product that the customer ordered.

Building rockets or any vehicle that can travel at the speed of light.  The speed of light is about 380 billion miles per second.  Using this technology, rockets can travel from Earth to Pluto in a few minutes (certainly less than an hour).   

 

The foregoing has outlined, in general, the physical aspects of the invention and is 

to serve as an aid to better understanding the intended use and application of the invention. In reference to such, there is to be a clear understanding that the present invention is not limited to the method or detail of construction, fabrication, material, or application of use described and illustrated herein. Any other variation of fabrication, use, or application should be considered apparent as an alternative embodiment of the present invention.     

 

<< content

 

 

  

Home | Task Engine | Time Machine | Patents | Notes | Donation 

Copyright 2007 (All rights reserved)