Are we doomed, or on the cusp of an AI-powered golden age? Can Claude’s Constitution, Microsoft’s Humanist AI Code of Conduct, or government regulation actually rein in AI, and should they?

Directions on Microsoft’s Barry Briggs is joined by industry analysts Peter O’Kelly and George Gilbert to debate AI doom versus AI breakthrough, and whether the guardrails being built today can keep us safe.

They discuss whether self-preservation is an emergent property of AI agents, why reward hacking may be a bigger risk than literal-minded AI, and how Anthropic’s and Microsoft’s approaches to AI governance differ. They also look at what these shifts mean for Microsoft customers weighing Copilot, Fabric IQ, and competing platforms, how AI could reshape jobs and organizations, and whether we’re headed for dystopia, paradise, or somewhere in between.