Ghost in the Machine: The Consequences of Bias in AI

No ratings

Presented at DevOpsDaysMinneapolis 2019 by

Nivia Henry

youtu.be

What if the code you’ve written alienated some of its users? Or what if it caused a group of people real-life harm? If you’re a responsible engineer, these should be among your biggest concerns.Artificial Intelligence and Machine Learning has introduced a new paradigm wherein code can do actual harm outside of its contained environment. While some of the effects may be comical (think: bad art made by AI), others can be deeply unfair to large groups of your users.Come with Nivia, an engineering manager at Spotify, on a journey to explore the world of machine learning. We’ll explore how algorithms are created, trained, and implemented, and more importantly, discuss practical steps to mitigate and eliminate such biases.