Phil 7.8.18

Scott Klemmer Keynote 2

  • What are interesting things that we can do with computers and teaching – 2011
  • Objective truth <-> Contextual truth
  • Design is in the middle, between objective and subjective truth
  • The act of assessing work is a good way to improve understanding
  • Problem finding as opposed to problem solving
  • “A negotiation around the valuation criteria” Jeff Nicholson
  • Negotiations also happen between the creators and the users, particularly in software design. The initial design is the starting point of that journey
  • What counts as preferred shifts over time
  • Talkabout – The subway model. Pick a time that you’re going to show up, and we’ll put you in a group. Small groups discuss topics.
  • Assigning to globally diverse discussion groups increase grades by greater amounts than more local, less diverse groups. Open-ended questions
  • DSCN0348DSCN0349DSCN0350DSCN0351DSCN0352

Participated in the panel on innovation in crowds (invited). There is a video, so I can figure out who to add:

  • Christopher Tucci,
  • Gianluigi Viscusi (GG)
  • Rosy Mondardini
  • Thomas Malone
  • Joel Chan
  • Philip Feldman

Eszter Hargitti – U of Zurich

  • Awareness of what is possible
  • The ability to create and share content
  • Wikigroan?
  • DSCN0353DSCN0354DSCN0355

When Ties Bind And When Ties Divide: The Effects Of Communication Networks On Group Processes And Performance DSCN0356.JPG_1DSCN0357.JPG_2

  • Network structural variance

Enhancing Collective Intelligence of Human-Machine Teams DSCN0358DSCN0359

  • Cognitive and ethnic diversity predict collective intelligence
  • Group structure, high level communication and equality of communication
  • It’s the quality of the individuals and the quality of the connections
  • Coordination technologies – connect humans

Implicit Coordination in Peer Production Networks DSCN0360DSCN0361DSCN0362DSCN0363

Collective Intelligence Systems for Analogical Search (must read! Joel Chan is at UMD)

  • Really interesting, worth reading. Purpose and mechanism may be related to belief spaces. Definitely trainable using NN to find purpose mechanism

Rational Collective Learning in the Laboratory

  • Groupthink. as a failure of design
  • Randomy constructed groups can make good design choices given failing parts with a history.