Skip to content

Engelbart Hlamt 1962

Outline for Doug Engelbart's "Augmenting Human Intellect: A Conceptual Framework"

These notes are typed into an outline from paper notes taken while reading this paper.

Version: 0.02

Future:
  1. Try to get the notes structure to match the table of contents.
  2. Find another copy to verify against my printed copy, make sure no info was lost.
  3. Try to get at least an outline of part 3A together.
  4. Part 3B is long and not very helpful but maybe some richer definitions can be pulled from it.
  5. Part 4 was research recommendations. Needs summarized to some degree but is very old and may not be useful.
Key References:
  • (1) Kenneth J.L., Putt, G.H. - " Administration of Research in a Research Corporation", RAND Corporation Report P-847 (20 April 1956)
  • (20) Fein, Louis - "The Computer Related Science (Synnoetics) at a University in the Year 1975"
  • What is the Whorfian Hypothesis? Neo-Whorfian Hypothesis? See language section in Part II.
Outcomes:
  • This paper generates language for approaching the H-LAM/T system. Does it give methodology or artifacts?
  • Is it my job to keep generating language and start building artifacts and methodology to continue amplifying our intelligence?
  • Are computers sufficient? Do we write computer programs for 500 years now, progressively generating language and automating more metholology as we spike off as more abstract beings?
  • This paper wants to "automate external symbol manipulation" - in other words it wants to build giant trees of language which are used to uncover new language. See section on language.

Augmenting Human Intellect:

Abstract:

"This is an initial summary report of a project taking a new and systematic approach to improving the intellectual effectiveness of the individual human being. A detailed conceptual framekwork explores the nature of the system composed of the individual and the tools, concepts, and methods that match his basic capabilities to his problems. One of the tools that shows the greatest immediate promise is the computer, when it can be harnessed for direct on-line assistance, integrated with new concepts and methods.

Part I - Introduction

  1. Part 1A - General ( "Terms and Definitions" )

    1. Augmenting Human Intellect = "Increasing a human's ability to..."
      1. Approach a complex problem situation
      2. Gain comprehension to suit his needs
      3. Derive solutions to problem
    2. Increased Capabilities
      • "Faster, better, beyond original capabilities"
    3. Complex Situations = "Professional Problems"
      • Problem scale is 20 minutes to 20 years
    4. Means
      • extensions of native sensory, motor, and mental capabilities
    5. Improvement Space
      • Visualize this extending from/with the H-LAM/T Capability Space
      • Will cover in section 2

  2. Part 1B - Objective of the Study

    1. Find: factors that limit effectiveness of the individual's basic information-handling capabilities in meeting society's needs for problem solving in its most general sense
    2. Develop: Techniques, Procedures, Systems - "which better match these basic capabilities to the needs, problems, and progress of society"
    3. Specifications:
      1. Both long- and short-term research goals in augmentation
      2. Specific actions in:
        1. Work Environment
        2. Thinking
        3. Skills
        4. Methods of working
      3. Framework is a basis for:
        1. evaluating possible relevance of work/knowledge from existing fields
        2. assimilate relevant things from those fields
          • "Not be in a vacuum of study"
    4. Framework is a basis for choosing starting points
      • reveal areas where research can be done
      • develop methodologies for that research
    5. H-LAM/T Conceptual Framework is a "rudimentary framework"
      • a place to start at and evolve from

Part II - Conceptual Framework

  1. Human has:
    • Motor channels: Transmit information
    • Sensory channels: Receive information
  2. Human processing:
    • Conscious Processing
    • Unconscious Processing
  3. Augmentation Means:
    1. Artifacts
    2. Language
    3. Methodology
      • Organization by methods, procedures, and strategies
    4. Training
  4. "Process Hierarchies"
    • Structure / Organization of each tiny step
  5. "Augmentation Means"
    • Serve to break down complex problems to tiny steps
    • "tiny step" = "sub-process"
  6. There is no way to "find the bottom" of the subprocess tree
    • It is not necessary to start at the bottom
    • A "simple enough" bottom layer will do
      • Does Engelbart mean simple enough to describe wholly or pragmatically? May not matter.
  7. Begin from basic capabilities:
    1. Sensory-Mental-Motor Process Capabilities
    2. Add process capabilities of Artifacts
  8. Each individual has their own repertoire of process capabilities which are selected and adapted for execution on a process
  9. A person and his/her artifacts can be said to have a "toolkit" called a repertoire hierarchy which extends from his basic capabilities
  10. 3 categories of process capability:
    1. Explicit-Human
    2. Explicit-Artifact
    3. Composite (Human to Artifact Interface)
  11. H-LAM/T System - Definition
    • Human using
    • Language, Artifacts, and Methodology in which
    • he is Trained
  12. H/LAM-T System - Terms
    1. has the capability and
    2. uses the repertoire to
    3. perform the process
  13. Executive processes embody all methodology of H/LAM-T System.
    • Habit
    • Strategy
    • Rules of Thumb
    • Prejudice
    • Learned Method - NOTE: implies training is contained in methodology?
    • Intuition
    • Unconscious Dictates
  14. NOTE: Methodology is enumerated above, which leaves only artifacts and language.
    • Is this collection (and any similar extensions) along with artifacts, language enough to embody all interactions a human has with his substrate?
    • Language evolves around systems. Uncovering new systems will resolve new language.
    • What is training then? It seems to be a subset of method but is included in the acronym.
    • Maybe training is intended to be an extension of methods but as a special and critical process for onboarding process capabilities in humans (viewing a human as an artifact as well).
    • There is an analogy Engelbart uses between humans and artifacts, both containing repertoire hierarchies of process capabilities. An artifact is generally designed to execute a process capability. A human can be trained to execute a process capability.
  15. A new innovation ripples/propagates both up and down the capability hierarchy.
  16. "Synthesis"
    • New capabilities from old capabiltiies. Calls for a systems engineering approach.
    • "Synergy and Synthesis" - Use metal allow tensile strength analogy to understand how organizing some capabilities can create a radical new "synthetic capability"
      • NOTE: This is cribbed BY Buckminster Fuller's "Operating Manual for Spaceship Earth". Engelbart was first unless it can be found in an earlier work. Regardless, the idea of alloys having complex new properties is potent. The world still has magic and it is materials engineering.
      • NOTE: It would be nice to have analogies other than materials engineering otherwise we should all just become materials engineers...
  17. capabilities are "changeable parts"
    • When changed, capabilities explose latent "synthetic capabilities" in that "repertoire hierarchy"
    • His idea: by changing interchangeable parts, "side reactions" or new capabilities can be unearthed.
    • An internal definition for what we call inventing by benevolent mistakes in experiment.
  18. Consider the whole H-LAM/T System as an ongoing experiment.
    • NOTE: All humanity is in a PERMANENT, UNIFIED experiment. All action is data.
    • Gather data as such (on the interactions in the system)
    • some data targets: capabilities, effectiveness, etc.
  19. As humans have added tools, our common/shared repertoire hierarchies have grown and formed around us. Radical change is the norm.
    • Can we accelerate radical change with guided research?
  20. "Intelligence Amplifier"
    • Does not imply increasing raw intelligence.
    • An amplifier makes louder what is already there.
    • We seek to amplify intelligence by restructuring the H-LAM/T System.
      • NOTE: He doesn't say repertoire hierarchies here, he says restructure the whole system.
      • NOTE: His hooks for restructure are Language, Artifacts, Methods (including Training).
      • NOTE: So what does he focus on?
        • We could do:
          • automation - move more and more trained methods into artifacts
          • science - create and apply new models and language (math?) to describe the universe
          • applications engineering - human methods seems like the weakest part of H-LAM/T... get a human to do more complex techniques
            • Unless we are already down this rabbit hole and people increasingly program complex capabilities into dynamic artifacts right now
  21. NOTE: Often H/LAM-T will exist natively in a poorly built and/or poorly maintained state. Retraining may be 80% of the value of amplification.... sometimes.
  22. LAM-T Augmentation means represent the amplifier:
    • Simplified using Doug's definitions: "The means by which we augment LAM-T are the amplifier"
    • Simplified more: "The means are the amplifier"
  23. Man-Artifact Interface:
    • NOTE: Human-Computer Interaction is the field today.
    • This term is derived from industrial revolution term "man-machine interface"
    • Two Domains:
      1. Human
      2. Artifact
    • Explicit-Human processes often exchange across an interface to Explicit-Artifact processes
    • Composite Processes - often designed solely for a functional match, "coupling", between the explicit processes.
  24. Language evolves to a situation
    • Language can force its users down a way of thinking. Language does exert a force on its own evolution.
    • NOTE: Language evolves to describe new concepts, which is a form of mental modeling of phenomena which can reveal new capabilities and thus contribute to novel processes.
    • NOTE: Human-Language aided thought generates new ways of thinking, new methods, and new processes. These can later be trained or incorporated in artifacts.
    • Capabilities drive actions and drive language evolution
    • "We offer the following hypothesis, which is related to the Whorfian Hypothesis: Both the language used by a culture, and the capability for effective intellectual activity, are directly affected during their evolution by the means by which individuals control the external manipulation of symbols."
    • "Automated external symbol manipulation" - Synmbols which humans represents concepts he is manipulating can be rearranged before his eyes. Essentially giant trees of words and concepts that can uncover new language automatically by processing the base concepts of current language.
    • Note that in the next paragraph, Doug predicts the action of using google on a smartphone to search a term in real time. To some degree we are already living in a much more fluid-language world.
    • Question: How would having a complete, fast dictionary impact language evolution?
      • NOTE: See encyclopedia dramatica, urban dictionary. People try to redefine things to what they want them to mean all the time, today. Terminology enters the zeitgeist very quickly but fades as quickly and terms often have different meanings to different clades. See: "Safe Space", 2016 representations and attacks. Organizations have to maintain internal definitions to be able to communicate.
  25. Two points in designing new repertoire hierarchies:
    1. Materials - fundamental capabilities from which all others must be constructed.
      1. Decomposition Search
        • Decompose capabilities to discover new materials (see: synthetic capabilities)
      2. De Novo - e.g. Materials Engineering (other analogies?)
    2. Principles - Structures, explicitly synergetic structures
      • "synergism" used by biologists and physicists for a siimilar concept:
        • Webster's Unabridged Dictionary (quote from paper): "...cooperative action of discrete agencies such that the total effect is greater than the sum of the two effects taken independently..."
      • Synergy is important because it is the threshold beyond which the means become greater than the sum of their materials.
      • Five tentative types of structure:
        1. Mental
        2. Concept
        3. Symbol
        4. Process
        5. Physical
    3. Human contributes many types of capability... "wears many hats"
    4. A complex capability has two classes of capability (NOTE: manager vs employee)
      1. Executive class
      2. Direct contributive class
    5. Repertoire Hierarchy (sic)
      • A mountain of white collar talent that sits atop and controls the talens of the "workers"
      • Goal: Maximize productivity, Gain comprehension, Solve problems.
      • Engelbart sees the production work as a flat base layer with a pyramid of white collar talent on top
      • "Theorem" - We could expect significant gains from improving symbol and process structuring.
    6. Finding: Humans switch frequently between one symbol structure and another continuously.
    7. "Human Augmentation Means" will require:
      • Training
      • Mental Tricks
      • Improved Language
      • New Methodology
      • "Unconscious Processes may be important"
    8. A human should be able to draw on explicit artifact capabilities at many levels in the process hierarchy