Jett Oristaglio, Data Science and Product Lead of Trusted AI at DataRobot, joins us to talk about Humble AI and to answer the question: “What is everything you need to trust an AI system with your life? ” In this podcast, he focuses on one critical dimension of Trusted AI: humility.

As AI becomes ubiquitous, more high-stakes decisions will be made automatically by machine learning models. But as the world changes, an AI system is faced with new examples that it hasn’t seen before. Without guardrails, automated decisions can turn into catastrophic failures if left unchecked and can destroy trust in AI. It is now critical that AI systems are built to be humble — just like humans, AI should know when it doesn’t know the right answer.

In this discussion, Jett reviews the fundamentals of making your AI system humble, including:

• Understanding the limitations of your model and when human intervention is needed
• Defining Humble AI triggers that protect from common failures of model overconfidence
• Monitoring your model over time for new errors and sources of overconfidence

Listen

Or Watch