Decision Making and Computational Creativity

When I think about where automated/computer decision making exists in this stack, it is in level 6 – computational creativity. You see, level 5 – artificial intelligence analyzes the world around it, and level 6 synthesizes things from its understanding. One of the things that can be synthesized is decisions. Level1&2:data. Level3&4:information. Level5&6:understanding. Level7&8:reproduction.

When building endemes by hand you are making decisions, so this form of building is a creative action. Endemes systemetize the construction of information, so endeme building can systemetize creativity.

Advantages and Disadvantages of Endemes

Endemes are useful for some things and not others.

Not to make light of these disadvantages because they are significant, why do I like them?
1. Endeme sets can play well with each other and a great variety of
information can be stored as endemes
– Things of different types can be compared and even related.
– It breaks down the walls that strict logic creates between similar data structures
2. It is oriented toward storing information much more than data
THEREFORE: Endemes contain a tiny piece of intelligence/information in a theoretically scalable form
– they might be used to build an artificial brain
3. Folk programming keeps creating codes. Endemes allow us to manage them.
4. There are ‘no wrong answers’ in a properly built endeme set.
This means endemes can be used for fault tolerant genetic algorithms,
and graceful degradation in systems.

The main problem with the endeme information structure is that
1. it welds multiple fields into a less machine-accessable format which means:
– it can’t be directly used as an index
– you have to ‘pack’ it when you want to create one
– you have to ‘unpack’ it when you want to get at the information back out.
2. Data is lost when you move it from a numeric to a relative strength format.
3. You have to build an interpreter to make sense of it
4. You have to cram fields together until you can cover everything in scope in 22 characteristics or less

– Endemes are not for precise data: The obstacles mean that endemes are not for everything. For eaxample if you want to store temperature, you might not want to use an endeme for the data, although you might want to use it for the field.

Action Verbs to Think Up Level 3 Project Ideas

This list covers action verbs to think about information in order to create new project ideas. Note that ‘negate’ gets used with other action verbs like ‘hurt’; note that these verbs may be combined.

  • Advocate: sell/promote
  • Benefit: help/improve
  • Create: Build/Develop/Construct
  • Decision: Decision Making/Choice/Action
  • Effect: Event/Result/Output/end/finish
  • Future: planning/prediction
  • Govern: manage/handle/guide/control
  • Hurt: damage/problem/hazard
  • Individualize: Customize/User preference
  • Joined: Relationships/interaction
  • Keep: settle/fix/be
  • Location: physical world
  • Modify: change/alter
  • Negate X: avoidance/nullification/Negation
  • Optimize: Balancing/Resources
  • Presentation: Present/Format/Arrange/Layout
  • Quanta/data: Data/text/numbers/bits/files/documents/measurements/values
  • Receive: Get
  • Send: Put/publish
  • Technology: Integration/Technology/Interfaces/UIs
  • Understand: Learning/Understanding/Knowledge building
  • View: Vision/Recognition/spotting/finding/searching/detection

Examples of information when used inside an application

Here are some examples of capabilities that using information provides for a program when used inside an application:

  • Allocation: Balancing/distribution/Balancing/resource allocation
  • Building AI: Knowledge representation/AI/Data–>Information–>Ontologies->AI.
  • Context/metadata: Metadata/Context/Richer metadata.
  • Defined data: characterized data/data context/semantic data/soft coded context.
  • Evolving systems: Evolution/Genetic Algorithms/Genetic algorithms.
  • Fuzzy information: Anti-aliasing data and information.
  • Generators: Creation/Creating new things programmatically.
  • Handling 0nf: automation/Conflated information handling (0NF).
  • Importance/order: Ordering/Importance management/Prioritization.
  • Judgment: Action/decision/function/Decision automation.
  • Knowledge/Information: Core information functionality above the level of data.
  • Languages: Natural language/sub-natural languages.
  • Mutual languages for users/humans/computers, HCI creative partnering.
  • Networking: Mutual languages for data networking.
  • Overlooked possibilities: Thorough coverage/Overlooked possibility identification.
  • Information proxies: Information data equivalent/Proxy creation and analysis.
  • Query/search: Filtering/search/matching/Filtering, searching, and matching.
  • Recognition: Viewing/recognizing/profiles, signature/pattern recognition.
  • State management: Status/State/Situation information, complex state management
  • Type assimilation: Kinds/bits/assimilating database complexity.
  • Using the results of data mining/information persistence/Information storage.
  • Viewing systems: Condensing data/sparse data or information UIs.

Creativity HCI

Human computer communication through creativity: Endemes provide an excellent way to merge user desired characteristics and randomness (aritificial creativity). An endeme is a permuted list of characteristics that can be combined in any way and in any order. If the computer comes up with some of them and the user comes up with some of them, then they merge well.

Happy medium: Endemes are designed to be simple enough for a computer to understand, and abstract enough for a person to understand and to be useful for a person.

Support Creativity: Endemes support creativity of the user.

Exposing an endeme: An endeme can be exposed to the user even if it is fairly technical because endemes deal with information and users think in terms of information.

Building a Sparse Grid Through Attention Management

I have seen grids in UIs that have dozens of columns, and the user has to scroll right a page or two to see all of the items.

We can build UI grids that show only the columns a user wants or needs to see. Columns could be added and removed in the grid as desired by the user. We could build an attention management endeme or two to indicate which columns to show. Each column would have an endeme and a user would have an endeme to identify which columns they want to see. Using endeme matching the desired columns would be shown. The organization could also have an endeme identifying which columns are important in the business situation for each grid. In addition, columns with data changes, unusual values or critical values could be moved into those columns shown. Users could add and remove columns, changing their attention management endeme if they so desire when they do this.

The advantage of this is increased productivity by the user because their attention is managed toward what they need to see, i.e. cost savings. Another advantage of this is that many more columns could be included as possibilities in a grid. Instead of having a few dozen, the grid could handle a few hundred columns. More data when managed thorugh attention management endemes means a user that is smarter in their decisions resulting from seeing the data, i.e. cost savings and added business opportunity.

Design Endeme Sets to Model Information

How to Design Endeme Sets

Designed to Combine: Each characteristic in an endeme should be designed to combine with any of the others resulting in a large number of possible emergent properties as they combine.

I sometimes takes years: It sometimes takes years to develop a good endeme set. It has to be both compact in its length and perfect in its application and excellent in its ability to combine its elements. By perfect, I mean its combinations cover what it needs to cover and its combinations do not generate bunches of ‘invalid’ results.

You have to start abstracting: Once you have gone beyond 22 items.

To test an endeme set: you can test/develop an endeme set by thinking of something it should describe and seeing if the endeme set can produce an endeme covering it.

What about two endeme sets?: Sometimes you will need more than one endeme set to describe something, but at least one of them has to be really good and essentially capture the essence of the thing being described. Creating two at the same time to work together or to parcel out the domain space does not work well. Focus on one until you have it well done, then go and make another if there is more domain space to cover. For one thing, it’s hard to do functions for situations if you have multiple endeme sets completing a situation.