Social Implications of Bias in Machine Learning

Explore the social implications of bias in machine learning through this 38-minute conference talk by Fiona Coath at Devoxx. Gain insights into how datasets and algorithms can perpetuate historical social biases, potentially leading to inaccurate results and exacerbating existing discrimination. Discover real-world examples and potential solutions to address these challenges. Learn how to assess the impact of biased machine learning models and transform this risk into a powerful tool for positive social change. Understand the importance of responsible AI development and the role of all professionals involved in creating ML tools in shaping a more equitable future. Delve into topics such as prejudice, statistical bias, the human condition, and the concept of "garbage in, garbage out" in machine learning. Explore case studies involving orchestras, Google Translate, and the algorithmic justice league. Examine the challenges of black box algorithms and the right to explanation. Discover tools and techniques for reducing bias, including the Open Data Institute Canvas, LIME, and Google's "What If" tool. Reflect on the responsibilities of AI developers and the potential for rapid improvements in AI systems to address societal issues.

Introduction
Social Bias
Prejudice
Statistical Bias
The Human Condition
What is Machine Learning
Overview
Garbage in Garbage out
Open Data Institute Canvas
Summary
Orchestras
Protected Attributes
Reducing Data
Adding Missing Data
Raising Machines
Google Translate
algorithmic justice league
the stem
diversity
The Forgotten
People Plus AI Research
Google
The Black Box
The Right to an Explanation
Open the Black Box
Simplifying Algorithms
Fixed Bugs
Lime
Tree Frog
What If
Feedback loops
Facebook
Compass
Silver Lining
Governance Usage
Amazon
Whos responsible
Pharmaceutical example
OpenAI example
We can quickly patch
We can demand better products
We can create updates very quickly
We realized that data is biased
We have the power
We want to go
Lets learn
We have an immense power
Change and shape society