Building an Ethical Data Science Practice

Explore the journey of building an ethical data science practice in this 41-minute talk by Cal Al-Dhubaib from Open Data Science. Discover the challenges and solutions in operationalizing ethical AI, including having difficult conversations about bias and risk, transforming ethics from a value to a virtue, and implementing practical tools and processes. Learn how to make a compelling business case for ethical data science, move beyond platitudes to real-world implementation, and incorporate ethics into talent acquisition and development strategies. Gain insights on fostering proactive risk management, building representative data science teams, and improving retention rates while addressing critical issues such as bias in AI, data complexity, and the importance of diversity in the field.

Introduction
What is Ethical Data Science
Examples of AI gone wrong
Data complexity
Sentiment analysis
The virtuous cycle
Diversity in data science
Open invitation
Connect with Cal
How can data science have an active role in combating inequity
How are other organizations responding to ethical challenges
How can we prevent models from becoming biased
How would you systematically test for biases
segregation between population
trading complexity for accuracy
outro