3 Lessons of Ancient & Modern Philosophy for Creative People-Centered System Design


Published on

Human-centered computational systems are increasingly important, but their design requires a delicate combination of qualitative and quantitative methods that is unfamiliar to most engineers

Published in: Technology
  • Be the first to comment

No Downloads
Total Views
On Slideshare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide

3 Lessons of Ancient & Modern Philosophy for Creative People-Centered System Design

  1. 1. Three Lessons of Ancient & Modern Philosophy for Creative People-Centered System Design David E. Goldberg Industrial & Enterprise Systems Engineering University of Illinois at Urbana-Champaign Urbana, Illinois 61801 [email_address]
  2. 2. 2007 & Creativity <ul><li>Globalization shaking confidence. </li></ul><ul><li>Cheap, effective talent in Shanghai and Bangalore can be hired and managed at a distance. </li></ul><ul><li>Routine kaizen (continual improvement) outsourced to the countries that generate a million engineers per year (China & India). </li></ul><ul><li>Creativity is not a nicety, it is not an option. </li></ul><ul><li>It is about economic survival. </li></ul><ul><li>Category enhancement is not enough. </li></ul>
  3. 3. Roadmap <ul><li>3 things I won’t be discussing. </li></ul><ul><li>Cold war mindset in an Internet world. </li></ul><ul><li>How I changed my mind. </li></ul><ul><li>3 lessons from modern & ancient philosophy. </li></ul><ul><ul><li>Searle and the construction of social reality. </li></ul></ul><ul><ul><li>Socrates/Plato and the importance of dialectic. </li></ul></ul><ul><ul><li>Aristotelian data mining and its application to social networks. </li></ul></ul><ul><li>People-centered design at the center of this creativity revolution. </li></ul>
  4. 4. 3 Things I Won’t Talk About <ul><li>I absolutely positively won’t talk about the following three things: </li></ul><ul><ul><li>GA scalability </li></ul></ul><ul><ul><li>Human & computer agency and the 4-quads </li></ul></ul><ul><ul><li>DISCUS, SAiNT & all that. </li></ul></ul>
  5. 5. GA Scalability & Efficiency <ul><li>GAs don’t work. No proofs. </li></ul><ul><li>Wrong!!! </li></ul><ul><li>GAs were shown to scale well in 1993. </li></ul><ul><li>Wrote up the story of GA scalability in 2002. </li></ul><ul><li>Plays on two levels: </li></ul><ul><ul><li>GAs as cool technology. </li></ul></ul><ul><ul><li>GAs as models of effective inventive/innovative processes. </li></ul></ul><ul><li>Understand at both levels. Beware 1 st gen GA/EC. </li></ul>
  6. 6. Billion-Variable Noisy Problem
  7. 7. H-C Agency in 4 Quadrants (Kosorukoff & Goldberg, 2002) Selective agent Inventive agent human computational human computational Standard Genetic Algorithms Computer Aided Design (CAD) Interactive Genetic Algorithms Human Based Genetic Algorithms
  8. 8. DISCUS
  9. 9. Semantic Reflection: Keygraphs http://www-discus.ge.uiuc.edu High frequency Low frequency Keyword
  10. 10. Social Reflection: IDM IDM = Influence Diffusion Method
  11. 11. Cold War Institutions, Internet World <ul><li>Large, centralized corporations, governments, and institutions (including universities). </li></ul><ul><li>Revolutions in 20 th century in transportation & communications give us radically different world because of transaction costs & network returns. </li></ul><ul><li>Neo-people-centered systems shaped by these forces. </li></ul>Ronald H. Coase (b. 1910)
  12. 12. How I Changed My Mind <ul><li>Wrote Change in Engineering Education (1993). </li></ul><ul><li>Started DISCUS research 2003. </li></ul><ul><li>Took Nextumi Chief Scientist slot 2004. </li></ul><ul><li>Presented Postmodern Systems Engineering (2004). </li></ul><ul><li>Recent influences, Pink’s A Whole New Mind and Price’s tech visionary (TV) research. </li></ul>
  13. 13. Cold-War View of Humans <ul><li>During the Cold War, humans were an obstacle to the proper functioning of a system. </li></ul><ul><li>Tom Wolfe’s, The Right Stuff, plot: tension between pilots and techies who would eliminate them. </li></ul><ul><li>Cold War view: Humans are error in the loop, and error is to be eliminated. </li></ul>
  14. 14. Postmodern Systems Engineering <ul><li>In internet world, human beings are integral part of the system. </li></ul><ul><li>The major innovations of our time involve people in the loop as major actors: </li></ul><ul><ul><li>Google: Search as human preference machine. </li></ul></ul><ul><ul><li>MSOffice: PPT, Word, Excel as human productivity engine. </li></ul></ul><ul><li>Postmodern view: Not error in the loop. Humans ARE the loop. </li></ul><ul><li>Brute facts of physics not dominant in postmodern systems. </li></ul><ul><li>Examples: </li></ul><ul><ul><li>What are the “physics” for Ebay? </li></ul></ul><ul><ul><li>What equations of motion govern Google? </li></ul></ul><ul><ul><li>What constitutive relations for MSOffice. </li></ul></ul>
  15. 15. Chief Scientist in a Web Startup <ul><li>Interested in commercial potential of genetic algorithms. </li></ul><ul><li>Asked to evaluate an idea inspired by my book on genetic algorithms. </li></ul><ul><li>Ended up as co-founder and chief scientist. </li></ul><ul><li>Unsettling to design product that has not existed. </li></ul><ul><li>Surprised by what I spend time thinking about. </li></ul><ul><li>Philosophical reflection most helpful tool. </li></ul><ul><li>This inspired idea for course teaching creative modeling techniques </li></ul>
  16. 16. Category Creators v. Enhancers <ul><li>Premium is on category creators —those who creates new categories of product and service. </li></ul><ul><li>This requires different skill set. </li></ul><ul><li>Right-brained thinking: integrative, creative, intuitive. </li></ul><ul><li>MFA + Engineer vs. MBA + Engineer. </li></ul>
  17. 17. Course: Creative Modeling for TVs <ul><li>Introduction </li></ul><ul><li>Models of creativity </li></ul><ul><li>Brainstorming </li></ul><ul><li>What is a model? What is a TV? </li></ul><ul><li>Construction of engineering reality. </li></ul><ul><li>2 techniques from Athens </li></ul><ul><li>Visualization and napkintalk </li></ul><ul><li>Canonical models </li></ul><ul><li>Facebook </li></ul><ul><li>Qual-quant shift </li></ul><ul><li>Little models </li></ul><ul><li>Tales from the trenches. </li></ul><ul><li>Squeezing little models. </li></ul><ul><li>Mixed, patched, and meta-models </li></ul>
  18. 18. 3 Lessons from Philosophy <ul><li>Course covers qualitative and quantitative modeling appropriate to new category creation: </li></ul><ul><ul><li>Dialectic in creative modeling. </li></ul></ul><ul><ul><li>Aristotelian data mining in creative modeling. </li></ul></ul><ul><ul><li>The construction of engineering reality. </li></ul></ul><ul><li>Will mention approach of little models to get a flavor of first-quant move from qual. </li></ul>
  19. 19. Tabula Rasa: Curse & Blessing of Category Creator <ul><li>How do we design when we don’t know how to talk about what we are designing? </li></ul><ul><li>Let’s start at the human beginnings of conceptual clarity. </li></ul><ul><li>Let’s start at the beginning of formal philosophy. </li></ul><ul><li>Let’s start with two key techniques from Athens. </li></ul>
  20. 20. What Examples of New Thought? <ul><li>Clearest examples are from philosophy. </li></ul><ul><li>Presocratic  Socrates  Plato  Aristotle. </li></ul><ul><li>Mechanisms of the new thought: </li></ul><ul><ul><li>Socratic dialectic </li></ul></ul><ul><ul><li>Aristotelian data mining </li></ul></ul>
  21. 21. Socrates and Dialectic <ul><li>Socrates was a pain in the neck. </li></ul><ul><li>Walked around Athens asking everyone impossible questions. </li></ul><ul><li>Then proved their answers were wrong, but rarely gave an answer himself. </li></ul><ul><li>Nonetheless, Socrates’s method was useful. </li></ul><ul><li>Conversation trying to probe what things really are (or might be). </li></ul><ul><li>Questions were the rights ones. Whitehead’s famous remark. </li></ul>Socrates (470-399 BCE)
  22. 22. The Probing of Dialectic <ul><li>Questions directed at the essence of things. </li></ul><ul><li>What is the meaning of a common phrase? “What is virtue?” </li></ul><ul><li>Answers often betray our lack of knowledge and understanding. </li></ul><ul><li>Examine answers critically, often with more questions. </li></ul><ul><li>Ask penetrating questions about the answers. </li></ul>
  23. 23. What’s This Got to Do with Products? <ul><li>Questions & conversation is at roots of all new products. </li></ul><ul><li>Research on tech visionaries shows that problem finding is the main activity of successful TVs </li></ul><ul><li>Spark of insight may come as flash, but dialectic necessary in new product creation. </li></ul><ul><li>Three roles of questions: </li></ul><ul><ul><li>Probe customers. </li></ul></ul><ul><ul><li>Probe organizational hurdles. </li></ul></ul><ul><ul><li>Probe product developers. </li></ul></ul>
  24. 24. Probing Customers <ul><li>Focus groups are immensely powerful. </li></ul><ul><li>Can be informal conversations with potential customers. </li></ul><ul><li>Can be formal focus groups behind the one-way mirror or over web. </li></ul><ul><li>The surprise of Nextumi: Search not solved problem. </li></ul>
  25. 25. Tactics of Dialectic <ul><li>Critical: </li></ul><ul><ul><li>Equivocation: use of term in different senses. </li></ul></ul><ul><ul><li>Question begging: assuming the conclusion. </li></ul></ul><ul><ul><li>Infinite regress: infinite sequence implying incoherence. </li></ul></ul><ul><ul><li>Loss of contrast & emptiness: distinction with little or no difference. </li></ul></ul><ul><li>Creative: </li></ul><ul><ul><li>Definition: Seek essential distinctions. </li></ul></ul><ul><ul><li>Analogies: Certain dialectical similarities. </li></ul></ul><ul><ul><li>Thought experiments: Hypothetical w/ true premises that does not follow. </li></ul></ul>http://www.amazon.com/Practice-Philosophy-Handbook-Beginners-3rd/dp/0132308487
  26. 26. Aristotelian Data Mining <ul><li>Called The Philosopher by some. </li></ul><ul><li>Amazing range and scope of work. </li></ul><ul><li>Created many of basic categories of college curriculum. </li></ul><ul><li>Founded a school the Lyceum. </li></ul><ul><li>We have 1/3 his output (2000 pages in 30 books). </li></ul><ul><li>Categories and Metaphysics. </li></ul><ul><li>Method very modern: </li></ul><ul><ul><li>Empirical search for data. </li></ul></ul><ul><ul><li>Considered attributes, which he named. </li></ul></ul><ul><ul><li>Classified data according to his attributes. </li></ul></ul><ul><li>Can we break this down? </li></ul>Aristotle (384-322 BCE)
  27. 27. A Hierarchy of Things <ul><li>In Aristotelian data mining we seek list of essential qualities that delineate products or services. </li></ul><ul><li>Notion of genus and species comes from Aristotle. </li></ul><ul><li>Things divided into groups of like kinds. </li></ul><ul><li>Can be subdivided further. </li></ul><ul><li>At the bottom are particulars or substances. </li></ul><ul><li>10 categories: substance, quantity, quality, relation, place, time, position, state, action, and passion. </li></ul>
  28. 28. Aristotelian Product Spaces <ul><li>Consider space of existing products that are related or similar. </li></ul><ul><li>Look for different exemplars that represent different types of products. </li></ul><ul><li>Taking viewpoint of the customer here. </li></ul><ul><li>May need separate decomposition for design. </li></ul><ul><li>Consider, for example, social networking space. </li></ul>
  29. 29. MySpace
  30. 30. Facebook
  31. 31. Flickr
  32. 32. Yahoo
  33. 33. What Common? What Different? <ul><li>Can we make a list of attributes that separate the space? </li></ul><ul><li>Can use J. S. Mills methods or methods of modern data mining. </li></ul><ul><li>Attributes: </li></ul><ul><ul><li>Mode of communication (email-IM-mobile-wall) </li></ul></ul><ul><ul><li>Sense of community (ind-group-friends) </li></ul></ul><ul><ul><li>Gate to community (edu filter-anybody) </li></ul></ul><ul><ul><li>What shared (text-docs-photos) </li></ul></ul>
  34. 34. Some Techniques with Attributes <ul><li>Dimensionalization of spaces very helpful to high-level thought. </li></ul><ul><li>Listen to discussions and try to dimensionalize quickly. </li></ul><ul><li>3 techniques that come to mind: </li></ul><ul><ul><li>Cartesian products </li></ul></ul><ul><ul><li>Expert systems </li></ul></ul><ul><ul><li>Hypertrophy of the dimensions </li></ul></ul>
  35. 35. Dimension Hypertrophy <ul><li>Are there missing categories on the dimensions? </li></ul><ul><li>Did we capture all the modes of things shared? </li></ul><ul><li>Example, text-photos-docs: What about videos? </li></ul><ul><li>Is that a good idea? </li></ul>
  36. 36. YouTube & $1.65 Billion Buyout
  37. 37. Canonical vs. Creative Analysis <ul><li>Example: Porter’s five forces. </li></ul><ul><li>NPD models. </li></ul><ul><li>Design models. </li></ul><ul><li>Need ability to do qualitative analysis on the fly. </li></ul><ul><li>Need engineers who can do this routinely. </li></ul>http://www.berg-marketing.dk/GIF/five_forces.gif
  38. 38. Construction of Engineering Reality <ul><li>Mill Prof of Philosophy of Berkeley. </li></ul><ul><li>Philosopher of language and mind. </li></ul><ul><li>Early work took off from Austin’s work on speech acts. </li></ul><ul><li>What does language have to do with it? </li></ul><ul><li>His book, The Construction of Social Reality (Free Press, 1995) , critical to our study. </li></ul><ul><li>Helps us understand social and institutional facts, separate physics from the social. </li></ul>John R. Searle (b. 1932)
  39. 39. Brute vs. Institutional Facts <ul><li>There are objects in the world that don’t depend on observers, brute facts, e.g. mountains, trees, atoms. </li></ul><ul><li>There are other objects that depend entirely on people and their interactions, institutional facts, money, chess, football . </li></ul><ul><li>Engineers study a lot about physical world and world of brute facts. </li></ul><ul><li>Much design about getting brute facts right, e.g. a screwdriver. </li></ul><ul><li>Are engineered objects mere brute facts? </li></ul>Screwdriver
  40. 40. Epistemology versus Ontology <ul><li>Epistemology: The study of knowledge. </li></ul><ul><li>Ontology: The study of what exists. </li></ul><ul><li>Our ontology: </li></ul><ul><ul><li>Live in world of physical particles in fields of force. </li></ul></ul><ul><ul><li>Some of these living systems with consciousness evolved. </li></ul></ul><ul><ul><li>These systems are intentional: Think and act toward other objects. </li></ul></ul>
  41. 41. Objectivity versus Subjectivity <ul><li>Have existence versus knowing, as well as objective versus subjective. </li></ul><ul><li>Examples: </li></ul><ul><ul><li>Mountain: existence  objective </li></ul></ul><ul><ul><li>Pain in toe: existence  subjective </li></ul></ul><ul><ul><li>Pain in toe: knowledge  objective </li></ul></ul><ul><li>Ontological subjectivity does not prevent epistemological objectivity. </li></ul>
  42. 42. Intrinsic versus Observer-Relative <ul><li>Intrinsic feature: Something that exists independent of observer (observer independent). </li></ul><ul><li>Example: A mountain. </li></ul><ul><li>Observer-relative feature: Something exists only relative to an observer. </li></ul><ul><li>Example: A screwdriver. </li></ul><ul><ul><li>Physical existence is intrinsic. </li></ul></ul><ul><ul><li>Screwdriverness is observer-relative. </li></ul></ul><ul><li>Engineered objects are at least partially observer-relative. </li></ul><ul><li>Why spend all time on the physics in curriculum? </li></ul>Fujiyama
  43. 43. Structure of Social Universe <ul><li>Mind creates an objective social reality. </li></ul><ul><li>Example, money: </li></ul><ul><ul><li>Trivial physics: money not money because of material existence. </li></ul></ul><ul><ul><li>Money, money because of our intentions. </li></ul></ul><ul><li>Other examples: language, government, universities. </li></ul><ul><li>Object fits description because we think it does. </li></ul><ul><li>What is ontology of the social and the institutional? </li></ul>
  44. 44. Building Blocks of Social Reality <ul><li>Need 3 new elements: </li></ul><ul><ul><li>Collective intentionality. </li></ul></ul><ul><ul><li>Assignment of function. </li></ul></ul><ul><ul><li>Constitutive rules. </li></ul></ul>
  45. 45. Collective Intentionality <ul><li>Need the notion of “we intend together.” </li></ul><ul><li>Attempts to reduce to individual intention are complex. </li></ul><ul><li>Existence of biological organisms with collective intentionality suggests CI is a primitive. </li></ul><ul><li>Social insects as an example: ants or termites for example. </li></ul>
  46. 46. Social vs. Institutional Facts <ul><li>Social fact is any fact involving 2 or more agents w/ collective intentionality. </li></ul><ul><li>In Searle’s terms, ants create social facts. </li></ul><ul><li>Institutional facts go far beyond social facts as we shall soon see. </li></ul>Australian ant hill
  47. 47. Assignment of Function <ul><li>Use of objects as tools: </li></ul><ul><ul><li>Monkey uses stick to get banana. </li></ul></ul><ul><ul><li>Man sits on rock. </li></ul></ul><ul><li>Physical existence facilitates function, but function is observer relative. </li></ul><ul><li>All function assignment is observer relative. </li></ul><ul><li>Aside: What does this say about engineering design? </li></ul>
  48. 48. Constitutive Rules <ul><li>How to distinguish between brute facts and institutional facts. </li></ul><ul><li>Types of rules: </li></ul><ul><ul><li>Some rules regulate: “Drive on rleft side of road.” </li></ul></ul><ul><ul><li>Some rules regulate and constitute: Rules of chess both regulate conduct of game and create it. </li></ul></ul><ul><li>Constitutive rules form: X counts as Y in C. </li></ul><ul><li>“ Move two and over one” counts as a knight’s move in chess.” </li></ul>
  49. 49. Web Life: Institutional Complexity <ul><li>Go on Google, search for online book seller, sign in to Amazon.com using account ID, order a book, using a credit card, get recommendations from recommender system & order some of those books, too. </li></ul><ul><li>Get confirmation message via e-mail account, and books delivered by FedEx. </li></ul>
  50. 50. Institutional Facts Don’t Wear Out <ul><li>They get better with use. </li></ul><ul><li>Use reinforces collective intentionality. </li></ul><ul><li>Types of Institutional fact creation: </li></ul><ul><ul><li>Emergence: early examples, money, marriage, property, emerged. </li></ul></ul><ul><ul><li>Designed: TCP/IP, Google, Ebay, other web companies are designed and go viral with widespread adoption.. </li></ul></ul>
  51. 51. Some Engineering Artifacts Institutional <ul><li>All engineering artifacts social. </li></ul><ul><li>Which objects have the big 3: collective intentionality, assigned function, X counts as Y in Z. </li></ul><ul><li>Example TCP/IP, 4-byte number counts as an address in TCP/IP network. </li></ul><ul><li>Generally, standards and protocols are designed institutional facts. </li></ul>
  52. 52. Institutional/Physical Landscape
  53. 53. Matters to People-Centered Design <ul><li>The web has changed a lot. </li></ul><ul><li>Interconnected, software reconfigurable systems enable brave new world of institutional facts: </li></ul><ul><ul><li>Easily constructed. </li></ul></ul><ul><ul><li>Easily propagated (viral marketing). </li></ul></ul><ul><ul><li>Easily iterated (systems upon systems). </li></ul></ul><ul><li>Has led to need for discipline of “postmodern systems engineering.” </li></ul>
  54. 54. Postmodern Design Principles <ul><li>Systems thinking is key. </li></ul><ul><li>Remember the big 3: </li></ul><ul><ul><li>Collective intentionality: viralness. </li></ul></ul><ul><ul><li>Assignment of function: designed. </li></ul></ul><ul><ul><li>Constitutive rules: X as Y in C, creating games. </li></ul></ul><ul><li>Consider 4 types of status: </li></ul><ul><ul><li>Symbolic: take care in use of words. </li></ul></ul><ul><ul><li>Deontic: e.g. rights and obligations of a user. </li></ul></ul><ul><ul><li>Honor: can be meaningful. </li></ul></ul><ul><ul><li>Procedural steps: systems view often requires interlocking steps. </li></ul></ul><ul><li>Inventiveness all the way up. </li></ul>
  55. 55. Bottom Line <ul><li>Creativity is now imperative for advanced economies. </li></ul><ul><li>More category creation, less enhancement. </li></ul><ul><li>Curse & blessing of the tabula rasa. </li></ul><ul><li>Must recall how to think new thoughts & get epistemology and ontology of world right. </li></ul><ul><li>Leads to a new discipline of postmodern systems design. </li></ul><ul><li>People-centered systems at the heart of this revolution. </li></ul>
  56. 56. More Information <ul><li>TEE, the book. http://eu.wiley.com/WileyCDA/WileyTitle/productCd-0470007230.html </li></ul><ul><li>TEE, the blog. www.entrepreneurialengineer.blogspot.com </li></ul><ul><li>TEE, the course. http://online.engr.uiuc.edu/webcourses/ge498tee/index.html </li></ul><ul><li>MTV, the course. http://online.engr.uiuc.edu/webcourses/ge498tv/index.html </li></ul><ul><li>Engineering and Technology Studies at Illinois (ETSI) http://www-illigal.ge.uiuc.edu/ETSI ). </li></ul><ul><li>2007 Workshop on Philosophy & Engineering (WPE) http://www- illigal.ge.uiuc.edu/wpe </li></ul><ul><li>Illinois Genetic Algorithms Lab http://www-illigal.ge.uiuc.edu/ </li></ul>
  57. 57. Optimal Teams: Deciding & Doing <ul><li>Dichotomy, in Simon’s Administrative Behavior. </li></ul><ul><li>Non-dimensional form: </li></ul>