How to Reduce AI Risk While Improving Productivity With Ovetta Sampson

AI can produce impressive results, but responsibility for those results still belongs to people.

In this episode, Adi Klevit interviews Ovetta Sampson about responsible AI implementation and the increasingly important role of human judgment. Ovetta explains the distinction between traditional machine learning and generative AI. With generative AI, determining whether an output is appropriate or useful often requires subjective human judgment, making it critical for businesses to establish clear standards around how these tools are used.

Ovetta and Adi also discuss some of the risks that come with widespread AI adoption, including copyright concerns, privacy, unreliable information, and the possibility of inaccurate information being repeated until it is perceived as true. Generative AI can produce an answer that sounds authoritative without actually determining whether that answer is accurate. That makes human review, verification, and critical thinking essential parts of any process that incorporates AI.

The conversation then turns to how businesses can use AI more effectively. Ovetta emphasizes a deceptively simple principle: remember that you're working with a machine. Once businesses understand how the technology processes information, they can create better systems for interacting with it by providing clear context, objectives, scope, guardrails, and examples. Rather than allowing everyone to experiment independently, companies can create repeatable processes for using AI while keeping people accountable for the final result.