The Intersection of Social Impact and Machine Learning

Online abuse and harassment increasingly stop people from engaging in conversation introducing a monumental challenge: how do you reduce toxicity and establish a safe environment online? Jigsaw’s (a unit within Alphabet) Conversation-AI team is finding new ways to combat this problem with machine learning. This intersection of social impact and machine learning creates a very interesting working dynamic for a product manager and presents unique challenges! The nature of the work is rapid pace and very experimental, and when looking to tackle complex issues with unconventional approaches, you are encouraged to fail fast and move on to the next potential solution. It is important to have strict guidelines to mitigate bias and remain ethical in your practices to ensure positive change and avoid inadvertently perpetuating prejudice. Furthermore, success with impact-driven work is less defined and not focused on easily identifiable output metrics like revenue. Rather, usage in a less popular region or influencing industry action that make your team’s solution a commonplace can be successful outcomes.

Key learnings include how to be an effective product manager in an impact-driven context and how to successfully leverage technologies like machine learning in the social impact space.