Upgrade to Pro — share decks privately, control downloads, hide ads and more …

Embracing Constraints

Embracing Constraints

(Presented at Codemotion Berlin 2017: https://berlin2017.codemotionworld.com)

This talk will explore the relationship between the problems we try to solve, the constraints we encounter, and the solutions those constraints not only allow, but enable us to find. From architecture to accessibility, from computational linguistics to cartography, examples abound: limitations on time, data, and other resources can have an undeniably positive impact on our creativity and problem-solving abilities. We’ll investigate how we can recognize, value, and comply with the restrictions that can help us, yet reject those that are holding us back.

654527a5cff1756177ef0b1bb0af7aa3?s=128

Anjana Sofia Vakil

October 13, 2017
Tweet

Transcript

  1. Embracing Constraints @AnjanaVakil | Codemotion Berlin | 13 October 2017

  2. Hallo! I’m @AnjanaVakil The Recurse Center

  3. @AnjanaVakil Speaking of Twitter...

  4. @AnjanaVakil Speaking of Twitter...

  5. @AnjanaVakil Constraints: How do they affect our work? Are they

    all created equal? Which help, which hinder?
  6. @AnjanaVakil Constraints for Concentration

  7. @AnjanaVakil A11y Tech for everyone

  8. @AnjanaVakil A11y Challenge for devs Accessibility is hard :( The

    A11y Project a11yproject.com/about.html “
  9. @AnjanaVakil A11y Challenge for devs Web Content Accessibility Guidelines (WCAG)

    2.0 w3.org/TR/2008/REC-WCAG20-20081211
  10. @AnjanaVakil A11y Challenge for devs Web Content Accessibility Guidelines (WCAG)

    2.0 w3.org/TR/2008/REC-WCAG20-20081211
  11. @AnjanaVakil A11y Clarity for products Web Content Accessibility Guidelines (WCAG)

    2.0
  12. @AnjanaVakil A11y Clarity for products Content vs. Presentation

  13. @AnjanaVakil Elm elm-lang.org

  14. @AnjanaVakil Elm The joy of constraints Elm restricts the way

    you program, resulting in maintainable code no matter what. There are no runtime exceptions so debugging is way less of an issue. Ossi Hanhinen, Futurice “How Elm made our work better” futurice.com/blog/elm-in-the-real-world “
  15. @AnjanaVakil Elm Types vs. Typos

  16. @AnjanaVakil Elm Types vs. Typos

  17. @AnjanaVakil Elm Types vs. Typos

  18. @AnjanaVakil Elm The joy of constraints Constraints: • Limit design

    space • Limit choices • Give guarantees Ilias van Peer “Elm - The Freedom of Constraints” PartialConf 2017 speakerdeck.com/zwilias/elm-the-freedom-of-constraints “
  19. “ @AnjanaVakil [There are] barriers that lead to breakthroughs. Patricia

    D. Stokes Creativity from Constraints 2006: Springer publishing, New York. p. 7
  20. @AnjanaVakil Constraints for Creativity

  21. “The power of constraints” Phil Hansen youtu.be/BgoAFS3xu74 @AnjanaVakil

  22. “ @AnjanaVakil Free to do anything, most of us do

    what's worked best, what has succeeded most o ten in the past. Patricia D. Stokes Creativity from Constraints 2006: Springer publishing, New York. p. xii
  23. @AnjanaVakil Speech to Text

  24. @AnjanaVakil Speech to Text /h e ɪ ˈs ɪ r

    i/ Acoustic model
  25. @AnjanaVakil Speech to Text Trained on 15,000 hours selected from

    30 years(!!) of speech Olga Kapralova et al., “A big data approach to acoustic model training corpus selection”, INTERSPEECH 2014 193.6.4.39/~czap/letoltes/IS14/IS2014/PDF/AUTHOR/IS140948.PDF
  26. @AnjanaVakil Speech to Text Microsoft msdn.microsoft.com/en-us/library/hh378476 < 30 Apple apple.com/ios/feature-availability/#siri

    < 40 Google cloud.google.com/speech/docs/languages < 70 Humans .ethnologue.com/guides/how-many-languages > 7000 Languages Supported
  27. @AnjanaVakil Speech to Text constrained Innovations in machine learning Siri

    Team, “Improving Neural Network Acoustic Models by Cross-bandwidth and Cross-lingual Initialization”, Apple Machine Learning Journal, Vol. 1, Issue 2, August 2017. machinelearning.apple.com/2017/08/01/cross-initialization.html
  28. @AnjanaVakil Speech to Text constrained Innovations in machine learning Siri

    Team, “Improving Neural Network Acoustic Models by Cross-bandwidth and Cross-lingual Initialization”, Apple Machine Learning Journal, Vol. 1, Issue 2, August 2017. machinelearning.apple.com/2017/08/01/cross-initialization.html the hidden layers learn feature transformations that are less language-specific and instead generalize between languages “
  29. @AnjanaVakil Map Data

  30. @AnjanaVakil Map Data constrained mapbox.com/ telemetry

  31. @AnjanaVakil Map Data constrained Map Data constrained Innovations in data

    collection @AnjanaVakil
  32. @AnjanaVakil Map Data constrained Map Data constrained Innovations in data

    collection @AnjanaVakil
  33. @AnjanaVakil Map Data constrained Map Data constrained Innovations in location

    services mapbox.com/telemetry @AnjanaVakil
  34. “ @AnjanaVakil Creativity thrives best when constrained. Marissa Mayer bloomberg.com/news/articles/2006-02-12/creativity-loves-constraints

  35. “ @AnjanaVakil But constraints must be balanced with a healthy

    disregard for the impossible. Marissa Mayer bloomberg.com/news/articles/2006-02-12/creativity-loves-constraints
  36. “ @AnjanaVakil [Some constraints] preclude the surprising & promote the

    expected Patricia D. Stokes Creativity from Constraints 2006: Springer publishing, New York. p. xii
  37. @AnjanaVakil Constraints for Conformity

  38. @AnjanaVakil “Java” Script the language must “look like Java” “

    Netscape to Brendan Eich 1995 brendaneich.com/2008/04/popularity
  39. @AnjanaVakil “Java” Script The Java influences, especially y2k Date bugs

    but also the primitive vs. object distinction (e.g., string vs. String), were unfortunate. “ Brendan Eich 2008 brendaneich.com/2008/04/popularity
  40. @AnjanaVakil Pipe Lines True diversity means better teams, better financial

    returns, better companies and a better, more innovative world. “ Project Include projectinclude.org
  41. @AnjanaVakil Pipe Lines

  42. @AnjanaVakil Pipe Lines

  43. @AnjanaVakil Pipe Lines NPR Morning Edition, “When women stopped coding”,

    21 October 2014 npr.org/sections/money/2014/10/21/357629765/when-women-stopped-coding
  44. @AnjanaVakil Pipe Lines

  45. @AnjanaVakil Pipe Lines The worst kind of group for an

    organization that wants to be innovative and creative is one in which everyone is alike and gets along too well. “ Professor Margaret A. Neale Stanford Graduate School of Business gsb.stanford.edu/insights/diverse-backgrounds-personalities-can-strengthen-groups
  46. @AnjanaVakil Consider your constraints

  47. Question constraints @AnjanaVakil

  48. Reject constraints for conformity @AnjanaVakil

  49. Embrace constraints for concentration @AnjanaVakil

  50. Embrace constraints for creativity @AnjanaVakil

  51. @AnjanaVakil Let constraints work for you

  52. Vielen Dank! Thanks to: Mozilla TechSpeakers Havi Hoffman & Denise

    Graveline Mapbox Young Hahn & Lauren Budorick Codemotion Berlin organizers SlidesCarnival.com @AnjanaVakil anjana.vakil@mapbox.com