Artificial Intelligence

Artificial Intelligence

"AI is in the new electricity" (Andrew Ng) and its worth billions because recent real world examples have made it's value proposition clear for companies. Whether you are in finance, education, healthcare, manufacturing or any industry, AI can be applied to a wide variety of problems such as new techniques in supply optimization to adapting and improving credit fraud identification to earlier disease diagnosis and individualized treatment.

There is a race to incorporate AI into products and services as users become more accustom to the increased unique experiences that these tools can offer. There is also a race to make AI and its related tools easier to understand and integrate. To overcome the need for PhDs and an army of data scientists, efforts are in progress to simplify and repurpose proven AI models. For example, retailers are using established computer vision models vs building them from scratch to visually search on product images.

Corporate priorities will highly influence the direction of AI research. We will explore the hype and the reality of AI's impact as well as how it is being democratized to meet growing demand.

2168aa4564112d3ba88869ca3cc994b3?s=128

Melanie Warrick

October 10, 2017
Tweet

Transcript

  1. Artificial Intelligence Melanie Warrick | @nyghtowl

  2. @nyghtowl “AI is the new electricity” - Andrew Ng ‘17

    AI = energy source
  3. @nyghtowl

  4. @nyghtowl

  5. @nyghtowl Machine Learning

  6. @nyghtowl

  7. @nyghtowl

  8. Where we are now ...

  9. @nyghtowl “Science and engineering of making intelligent machines...” - John

    McCarthy ‘55
  10. @nyghtowl

  11. @nyghtowl

  12. @nyghtowl Narrow General Super

  13. @nyghtowl

  14. new data ~ 10B MB/sec

  15. @nyghtowl Narrow => Math & Research Fields ➔ Machine Learning

    ➔ Statistics & Probability Models ➔ Generative & Adversarial Models ➔ Graph Theory ➔ Quantum Computing
  16. @nyghtowl Artificial Intelligence Machine Learning Deep Learning

  17. None
  18. @nyghtowl

  19. “AI-first world” Android Apps Gmail Maps Photos Speech Search Translation

    YouTube and many others ... Used across products: 2012 2013 2014 2015
  20. Example Use Cases Search Search ranking Speech recognition Android Keyboard

    & speech input Gmail Smart reply Spam classification Drive Intelligence in Apps Chrome Search by image Assistant Smart connections across products Maps Parsing local search Translate Text, graphic and speech translation Cardboard Smart stitching Photos Photos search
  21. @nyghtowl 40% savings

  22. @nyghtowl Machine Learning

  23. None
  24. @nyghtowl Hurdles ... - Complex problem-solving - Emotional & social

    IQ - Rule change & new domain rapid adoption - ...
  25. Where we are going ...

  26. @nyghtowl General Intelligence

  27. @nyghtowl Democratization ~ Transfer Learning Cloud Speech API Cloud Vision

    API Cloud Natural Language API Cloud Translate API Cloud Video Intelligence Cloud Jobs API
  28. @nyghtowl Augmentation

  29. @nyghtowl 10 years => 1 Day

  30. @nyghtowl Behavior Analysis

  31. @nyghtowl

  32. @nyghtowl 2017 ~ 8.4B Devices

  33. @nyghtowl 2017 ~ 1.2T Searches/Yr

  34. @nyghtowl Intelligence Augmentation - Medical & Genomics - Sales &

    Marketing - Security - Operations & Utilities - IoT - Smartphones & Search - ...
  35. @nyghtowl Commonplace its assumed

  36. Getting there from here ...

  37. @nyghtowl

  38. @nyghtowl

  39. @nyghtowl Defined Problem Data Systems Experts

  40. @nyghtowl Challenges ➔ Data & Experts ➔ Tuning ➔ Generalization

    ➔ Anthropomorphism ➔ Bias
  41. @nyghtowl New tech is like magic Change => fear |

    excitement
  42. @nyghtowl New tech is like magic Change => fear |

    excitement
  43. @nyghtowl Last Points AI => Narrow Data is Queen Democratization

    => Transfer Learning Commonplace = Electricity | Wheel
  44. @nyghtowl References: Images • iStock.com/4x-image • iStock.com/hometowncd • iStock.com/diego_cervo •

    iStock.com/teekid • iStock.com/laremenko • iStock.com/alexialex • iStock.com/ sarah5 • iStock.com/Grzegorz Petrykowski • iStock.com/stellalevi • iStock.com/adventtr • iStock.com/d1sk • iStock.com/vectortatu • iStock/smoothgroover22 • iStock/phive2015 • https://www.explainxkcd.com/wiki/images/a/a7/timeline_of_bicycle_design.png | XKCD • https://commons.wikimedia.org/wiki/File:Cancer_cells_(1).jpg • https://deepmind.com/blog/deepmind-ai-reduces-google-data-centre-cooling-bill-40/ • http://maxpixel.freegreatpicture.com/Treat-Variety-Chocolate-Decorated-Doughnuts-Donuts-1368342
  45. @nyghtowl References: Images • https://commons.wikimedia.org/wiki/File:Internet_of_Things.jpg | McGeddon • https://blog.openai.com/dota-2/ |

    Valve Corporation • https://eugene.kaspersky.com/2016/09/09/the-artificial-artificial-intelligence-bubble-and-the-future-of-cybersecurit y/ | Austin Powers | Makeameme.org • https://commons.wikimedia.org/wiki/File:Duomo_Lucca_cathedrale_Lucques_labyrinthe.jpg | Myrabella • http://www.funnyjunk.com/funny_pictures/631996/Armageddon/ | bittercrackpot • https://www.flickr.com/photos/photoverulam/3567531681 | Richard Gillin • https://www.flickr.com/photos/moyogo/4884992 | Denis Jacquerye • https://www.flickr.com/photos/rosenfeldmedia/6949089460 • Copyright and disclaimer notice: https://creativecommons.org/licenses/by/2.0/ • License notice: https://creativecommons.org/licenses/by/2.0/legalcode Additional Resources: • https://analyticsweek.com/content/big-data-facts/ • https://www.techemergence.com/machine-learning-in-genomics-applications/ • https://www.technologyreview.com/s/609048/the-seven-deadly-sins-of-ai-predictions/
  46. @nyghtowl Artificial Intelligence Melanie Warrick @nyghtowl