Trust and Advancements in AI: Reinforcement Learning and Human-Centric Data

Generative AI has revolutionized how we write, code, and create—but it still hallucinates. In this episode of Today in Tech, host Keith Shaw sits down with Byron Cook, Distinguished Scientist and VP at AWS, to break down why AI continues to make things up—and what we can do to stop it.

They explore:

What AI hallucinations really are—and why they’re not always bad

How automated reasoning can serve as a “logic cop” for generative AI

The challenge of defining truth in business contexts

Agentic AI: Why it needs stronger safeguards before widespread adoption

How Amazon is using formal methods to improve accuracy and reduce risk

When hallucinations are harmless—and when they can cost you your job, money, or reputation

If you’re an enterprise leader, developer, or tech enthusiast trying to understand the next phase of trustworthy AI, this is the episode for you.

:bell: Don’t forget to like, comment, and subscribe for more tech insights every week!