AI Activations - Testing Where It Matters: Building AI That Holds Up in the Wild

As AI moves from experimentation to production, trust is no longer optional. It is a prerequisite. In this conversation, Cobus Greyling, AI Evangelist, interviews Tiarne Hawkins, CEO of Optica Labs, to explore how AI systems behave outside controlled environments. Together, they will examine the disconnect between lab-based testing and real-world usage, why many promising AI systems break after deployment, and how teams can rethink testing to uncover failure modes early. The discussion will also look ahead at what founders and enterprise leaders must prioritize now to ensure their AI systems remain reliable, resilient, and trustworthy over time.