Droidcon Berlin 2018 - Day One
  • Day One at Droidcon Berlin 2018
  • Schedule
  • Speed Dating
  • Android Telephone
  • Feeback
  • Panels
    • Rx Must Die
    • Women in Tech
    • Cross Platform on Android
    • Rise of the Machines
  • Session Ideas
    • Session Ideas
    • Impact of Google's Policies
    • Androids Play Mario Kart Together
    • Navigation Clean
    • Android Community Platform
    • AI / Machine Learning
    • Swagger API
    • AR in Android
    • React Native
    • Accessibility
    • How to learn Android
    • Build Speed
    • Gradle Multi Modules
    • Kotlin DSL Live coding
    • Kotlin Problem
    • Paging like a Pro
    • TDD in Android
    • Work together
    • Fun in Enterprise apps
  • Workshops
    • Sketchnoting
Powered by GitBook
On this page
  • Rise of the Machines
  • Examples
  • Biases
  1. Panels

Rise of the Machines

Rise of the Machines

The Future of Programming - How AI/ML will change our lives

A panel discussion

  • Fernando Cejas, IBM

  • Juergen Fey, Androidian

  • Meghan Kane, Novoda

  • Dr.Roman Belter, APPSfactory

  • Hoi Lam, Google

Nowadays, machines are powerful enough to run local

Examples

  • Image recognition

  • Tone recognition in comments

  • Smart advice, suggestions in communications

Interpretability, understanding way AI makes a decision is still in active research.

Know your data! AI can be a good application for your data.

Important to work together between UX and engineers to overcome uncertainty. Important to generalize enough when model is used outside the original use case.

Opportunitites: Differnt simple to use libraries

IBM watson Other IBM tools to connect data source Azure machine learning Google AI Cloud service

Analyze if you need a model.

Biases

  • AI is like a mirror of humanity

  • Users of AI need to be aware of the biases

  • Google looking into minority groups to have the product less biased

Higher quality data is more important than quantity.

AI teams are still detached from the test of the company. AI should be something horizontally.

AI will solve many of daily issue like bug fixing, etc. in the future. Management is still to understand how to deal with AI.

Steps for AI in development into the future: 1. Identify problems in current code 1. Suggest better solution 1. Generate own solutions

PreviousCross Platform on AndroidNextSession Ideas

Last updated 7 years ago

Research papers are quite accessible for the general puplic. E.g. , Standford computer vision course

https://paperswithcode.com/