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.

Anjana Sofia Vakil

October 13, 2017
Tweet

More Decks by Anjana Sofia Vakil

Other Decks in Technology

Transcript

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

    all created equal? Which help, which hinder?
  2. @AnjanaVakil A11y Challenge for devs Accessibility is hard :( The

    A11y Project a11yproject.com/about.html “
  3. @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 “
  4. @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 “
  5. “ @AnjanaVakil [There are] barriers that lead to breakthroughs. Patricia

    D. Stokes Creativity from Constraints 2006: Springer publishing, New York. p. 7
  6. “ @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
  7. @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
  8. @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
  9. @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
  10. @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 “
  11. “ @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
  12. “ @AnjanaVakil [Some constraints] preclude the surprising & promote the

    expected Patricia D. Stokes Creativity from Constraints 2006: Springer publishing, New York. p. xii
  13. @AnjanaVakil “Java” Script the language must “look like Java” “

    Netscape to Brendan Eich 1995 brendaneich.com/2008/04/popularity
  14. @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
  15. @AnjanaVakil Pipe Lines True diversity means better teams, better financial

    returns, better companies and a better, more innovative world. “ Project Include projectinclude.org
  16. @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
  17. @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
  18. Vielen Dank! Thanks to: Mozilla TechSpeakers Havi Hoffman & Denise

    Graveline Mapbox Young Hahn & Lauren Budorick Codemotion Berlin organizers SlidesCarnival.com @AnjanaVakil [email protected]