Saturday, October 3, 2009
This Blog is Moving to a New Location
Hi, and thanks for stopping by! In order to save time and consolidate our creative efforts, we're moving all future postings to our web site at:
Thursday, May 7, 2009
Practical Mental Relativity - A Phone Conversation - Part One
mp3 of a phone conversation between me and Juniko Moody regarding how to apply Mental Relativity psychology theory in a practical way in every day life. Click to listen or download.
Origins of Mental Relativity
Many people have asked how we came up with the Mental Relativity Theory. Well, it didn't happen all at once. In fact, it all began with our attempt to understand the structure of fictional stories. That was a three year full-time effort, 8 hours a day. And the theory we now grew out of quite different concepts those we have now refined.
In going through my archives, I just discovered four hours of recordings we made in 1991 to document our very first attempt at a "complete" theory - kind of a unified field theory of story. Here are all four hours of audio in mp3 formt, divided into 8 parts. Enjoy!
Saturday, May 2, 2009
Proposed Search Engine Algorithm from 1997
Back in '97 I came up with a way to apply Mental Relativity psychology theory to interactive search engine algorithms which would not only learn the interests of a user and predict what else they would want, but also predict things they were going to want that they did not YET want.
This idea would cause search engines to lead the user, rather than simply parse data and follow. Further, it could guide users to new areas of interst they had never before considered and of which they were completely unaware.
Here 'tis:
A Mental Relativity Approach
to Search Engines
Proposed by Melanie Anne Phillips
The caliber of a search engine is judged by the quality of its results. When it returns the expected information it is deemed accurate. When it brings information of greater detail or ferrets out information that is more obscure, it is deemed powerful. But if a search engine could anticipate a user's wants and return information not yet specifically requested - that would seem to be nothing short of magic.
Currently, search engines are rather unsophisticated. They look for keywords and phrases in document titles, keyword lists, and perhaps even in the document body text. The expected next generation of search engines will pay more attention to context, looking for word position in phrases, sentences, headers, and perhaps even looking for patterns of words that do not fall within single phrases.
What I am proposing goes beyond either of these approaches by taking a sidestep into a new perspective on the whole process, provided by Mental Relativity theory. What I propose is a search engine with the ability to forget.
The human mind gathers data by observation, much as a search engine encounters information. The human mind pays attention to word position - how close is the word to the beginning OR ending of a phrase, sentence, paragraph, or page. In fact, words at the exact middle of any grammatical unit carry much less weight in memory than those at the beginning or end, which carry about the same weight.
Further, the human mind works much like the electron shells in an atom. Repeated observation charges electrons up to a higher level of excitement or energy. But, those electrons in the highest energy shells also have the shortest half-lives. That is to say, the most highly charged electrons are more likely to spontaneously emit a photon and drop to a more stable level than less highly charged electrons.
In the mind, what we repeatedly observe the most becomes more highly charged. When internal thought - creative thought - occurs, our familiarity with the item most observed diminishes as the energy is changed into a conscious concept or thoughton which now runs through our brains as a packet of awareness. We then look for patterns in our environment that resemble that contained in the thoughton. When we encounter such a pattern, the new observation is out of phase with the internal thoughton and the energies from the two cancel.
In this manner, we seek out that with which we are most familiar, but when we find it, lower our motivation to seek more of the same. Instead, the peripheral information in the new observation is added to our mental database. New words or patterns are now the most highly charged, and the process begins again in search of different kinds of observations.
So, the first step in creating a first generation Mental Relativity search engine would be to keep a dynamic database of weighted words and/or phrases based on frequency and position within grammatical phrases. The next step would be to deduct one exposure to the word or phrase for every new instance encountered. In this manner contextual data patterns focus and rise to the surface as creative searches rather than simple pattern duplication.
Although the above search engine would be much more intuitive than any existing tool, it hardly scratches the surface of the potential sophistication of future versions. For example, as outlined, there is no accommodation for user control or input. The user simply follows the search engine wherever it leads. When the Mental Relativity perspective is applied more vigorously, however, the degree of sophistication increases progressively.
A most basic refinement would be to give user control to the number of words contained in the dynamic database. The human mind will eventually forget what it has not encountered in a long time. By cutting off the size of the word list, words or phrases least often encountered will be dropped completely from the list after a period of time in favor of those more commonly observed. This control would dictate the degree of inertia carried by an observation, or in other words, how much of a tendency there would be for the search engine to fall into a the same old ruts.
Another refinement would be in regard to the aforementioned plan of having only the most weighted word or phrase dictate the scope of the search. Because data at different levels of exposure have different half-lives, they will fire eventually, even if higher levels are firing more frequently. In the search engine, this means that the words and phrases in any given search parameter should include not only the most weighted, but also those that have waited the most time. So, some obscure word or phrase observed a long time ago and infrequently, might be included in a search when its half-life dictated a firing. In this manner, the serendipitous nature of creative thought shows up in the search routines.
Finally, the mind often runs a search on itself, turning completely inward in self-consideration and discovering new meaning, even without new observation. To reflect this attribute, the Mental Relativity search engine should keep a record of past databases. These would simply be the collections of weighted words and phrases from earlier searches, and would not be overly large in size, especially when parameterized.
Each time a normal search is performed, an additional search would be performed of the database of previous databases. This procedure would turn up new free associations of words and even phrases not ever encountered directly in observation. These words and phrases would then be added to the search parameters for the next external search.
This control mimics the nature of male or female mental sex. A user could determine whether the search engine would first look to the internal database and then the external observations or vice-versa. An external first search engine would be the equivalent of male mental sex and an internal first search engine would be female mental sex.
In addition, the frequency with which the search engine looks to the internal or external could be controlled. Just as we have found that in humans, we are all biased toward time (represented by the history of previous databases) or space (represented by the external observations sought by the search engine), we are not all biased to the same degree. By adjusting the frequency of internal compared to external searching, we adjust the degree to which the search engine's mind is biased to the temporal vs. the spatial.
So, a user could control the mental sex of the search engine and also the degree to which it is interested in linear vs. holistic issues. This could be useful both for those who want the search engine to mimic their manner of thinking and also for those who would like input from an opposite mind set as a counterpoint to their own. In fact, controls could be provided to allow for both spatial and temporal searches so that results could be compared or combined to provide the most expected and serendipitous results simultaneously.
You will note that four separate algorithms have been outlined above:
1. The hunt for the most weighted item, which mimics the conscious mind.
2. The limit to the size of the word list, which mimics memory.
3. The spontaneous searching based on expired half lives, which mimics the subconscious
4. The ratio between internal and external searching, which mimics the preconscious.
All four aspects of the mind have been represented in this overall matrix of algorithms. Because they are created in a binary sense in software, they are not truly self-aware, which would require the equivalent of a biochemical environment. But, a truly self-ware search engine would be useless, as it would search what it wanted, not what the user wanted. For purposes of a search engine, then, this proposal outlines the most human-like and creative approach possible while still being under the complete control and guidance of the user.
In conclusion, as long as search engines do not possess the ability to forget, they will be little more than binary processing machines. But by applying Mental Relativity theory to the problem, a truly intuitive search engine can be created which will continually surprise users by anticipating the direction of their personal growth and bringing them just what they wanted before they knew they wanted it.
It is my belief that the power of a search engine built to these specifications would revolutionize data search and retrieval in all areas, most notably in the internet and intranet markets. Because of its unique approach and algorithms, the software implementation of such a search engine is almost certainly patentable.
This combination of power and patentability makes the Mental Relativity search engine a compelling, licensable, product with a world market, including business, science, and education, not to mention sales to existing search facilities such as Yahoo, Infoseek, and Excite, as well as uses in "push" technology, and as the driver of sophisticated help functions in other software.
This idea would cause search engines to lead the user, rather than simply parse data and follow. Further, it could guide users to new areas of interst they had never before considered and of which they were completely unaware.
Here 'tis:
A Mental Relativity Approach
to Search Engines
Proposed by Melanie Anne Phillips
The caliber of a search engine is judged by the quality of its results. When it returns the expected information it is deemed accurate. When it brings information of greater detail or ferrets out information that is more obscure, it is deemed powerful. But if a search engine could anticipate a user's wants and return information not yet specifically requested - that would seem to be nothing short of magic.
Currently, search engines are rather unsophisticated. They look for keywords and phrases in document titles, keyword lists, and perhaps even in the document body text. The expected next generation of search engines will pay more attention to context, looking for word position in phrases, sentences, headers, and perhaps even looking for patterns of words that do not fall within single phrases.
What I am proposing goes beyond either of these approaches by taking a sidestep into a new perspective on the whole process, provided by Mental Relativity theory. What I propose is a search engine with the ability to forget.
The human mind gathers data by observation, much as a search engine encounters information. The human mind pays attention to word position - how close is the word to the beginning OR ending of a phrase, sentence, paragraph, or page. In fact, words at the exact middle of any grammatical unit carry much less weight in memory than those at the beginning or end, which carry about the same weight.
Further, the human mind works much like the electron shells in an atom. Repeated observation charges electrons up to a higher level of excitement or energy. But, those electrons in the highest energy shells also have the shortest half-lives. That is to say, the most highly charged electrons are more likely to spontaneously emit a photon and drop to a more stable level than less highly charged electrons.
In the mind, what we repeatedly observe the most becomes more highly charged. When internal thought - creative thought - occurs, our familiarity with the item most observed diminishes as the energy is changed into a conscious concept or thoughton which now runs through our brains as a packet of awareness. We then look for patterns in our environment that resemble that contained in the thoughton. When we encounter such a pattern, the new observation is out of phase with the internal thoughton and the energies from the two cancel.
In this manner, we seek out that with which we are most familiar, but when we find it, lower our motivation to seek more of the same. Instead, the peripheral information in the new observation is added to our mental database. New words or patterns are now the most highly charged, and the process begins again in search of different kinds of observations.
So, the first step in creating a first generation Mental Relativity search engine would be to keep a dynamic database of weighted words and/or phrases based on frequency and position within grammatical phrases. The next step would be to deduct one exposure to the word or phrase for every new instance encountered. In this manner contextual data patterns focus and rise to the surface as creative searches rather than simple pattern duplication.
Although the above search engine would be much more intuitive than any existing tool, it hardly scratches the surface of the potential sophistication of future versions. For example, as outlined, there is no accommodation for user control or input. The user simply follows the search engine wherever it leads. When the Mental Relativity perspective is applied more vigorously, however, the degree of sophistication increases progressively.
A most basic refinement would be to give user control to the number of words contained in the dynamic database. The human mind will eventually forget what it has not encountered in a long time. By cutting off the size of the word list, words or phrases least often encountered will be dropped completely from the list after a period of time in favor of those more commonly observed. This control would dictate the degree of inertia carried by an observation, or in other words, how much of a tendency there would be for the search engine to fall into a the same old ruts.
Another refinement would be in regard to the aforementioned plan of having only the most weighted word or phrase dictate the scope of the search. Because data at different levels of exposure have different half-lives, they will fire eventually, even if higher levels are firing more frequently. In the search engine, this means that the words and phrases in any given search parameter should include not only the most weighted, but also those that have waited the most time. So, some obscure word or phrase observed a long time ago and infrequently, might be included in a search when its half-life dictated a firing. In this manner, the serendipitous nature of creative thought shows up in the search routines.
Finally, the mind often runs a search on itself, turning completely inward in self-consideration and discovering new meaning, even without new observation. To reflect this attribute, the Mental Relativity search engine should keep a record of past databases. These would simply be the collections of weighted words and phrases from earlier searches, and would not be overly large in size, especially when parameterized.
Each time a normal search is performed, an additional search would be performed of the database of previous databases. This procedure would turn up new free associations of words and even phrases not ever encountered directly in observation. These words and phrases would then be added to the search parameters for the next external search.
This control mimics the nature of male or female mental sex. A user could determine whether the search engine would first look to the internal database and then the external observations or vice-versa. An external first search engine would be the equivalent of male mental sex and an internal first search engine would be female mental sex.
In addition, the frequency with which the search engine looks to the internal or external could be controlled. Just as we have found that in humans, we are all biased toward time (represented by the history of previous databases) or space (represented by the external observations sought by the search engine), we are not all biased to the same degree. By adjusting the frequency of internal compared to external searching, we adjust the degree to which the search engine's mind is biased to the temporal vs. the spatial.
So, a user could control the mental sex of the search engine and also the degree to which it is interested in linear vs. holistic issues. This could be useful both for those who want the search engine to mimic their manner of thinking and also for those who would like input from an opposite mind set as a counterpoint to their own. In fact, controls could be provided to allow for both spatial and temporal searches so that results could be compared or combined to provide the most expected and serendipitous results simultaneously.
You will note that four separate algorithms have been outlined above:
1. The hunt for the most weighted item, which mimics the conscious mind.
2. The limit to the size of the word list, which mimics memory.
3. The spontaneous searching based on expired half lives, which mimics the subconscious
4. The ratio between internal and external searching, which mimics the preconscious.
All four aspects of the mind have been represented in this overall matrix of algorithms. Because they are created in a binary sense in software, they are not truly self-aware, which would require the equivalent of a biochemical environment. But, a truly self-ware search engine would be useless, as it would search what it wanted, not what the user wanted. For purposes of a search engine, then, this proposal outlines the most human-like and creative approach possible while still being under the complete control and guidance of the user.
In conclusion, as long as search engines do not possess the ability to forget, they will be little more than binary processing machines. But by applying Mental Relativity theory to the problem, a truly intuitive search engine can be created which will continually surprise users by anticipating the direction of their personal growth and bringing them just what they wanted before they knew they wanted it.
It is my belief that the power of a search engine built to these specifications would revolutionize data search and retrieval in all areas, most notably in the internet and intranet markets. Because of its unique approach and algorithms, the software implementation of such a search engine is almost certainly patentable.
This combination of power and patentability makes the Mental Relativity search engine a compelling, licensable, product with a world market, including business, science, and education, not to mention sales to existing search facilities such as Yahoo, Infoseek, and Excite, as well as uses in "push" technology, and as the driver of sophisticated help functions in other software.
Friday, May 1, 2009
Op Quads
Dramatica software's model of Mental Relativity psychology, for which we received a patent for its implementation in the Story Engine, is based on a series of operations - dynamics that are applied to the dramatica structural model to twist and turn it like a Rubik's Cube of psychology. Like the structure, the operations fall into Quads (groups of four). The following picture shows some example Op Quads that are applied to the StoryEngine, and what they do (in short-hand symbols).
Thursday, April 30, 2009
How Women & Men Approach Life.
Here are some generalizations I worked up during our early development of the Mental Relativity Theory.
Subscribe to:
Posts (Atom)
