Curated playlist · 03
Where can the human say no?
Agents, predictions, shadow use, and nuclear decision systems test whether “human in the loop” has any teeth.
The question beneath the question
Who can inspect, interrupt, overrule, and revoke the system when the answer matters?
Human oversight is not a person clicking approve. It is evidence, time, authority, recourse, and a practiced ability to stop the system.
“My approach was: how do I not just unblock him, but teach him to unblock himself?”
Dan Botero
The listening order
Each stop changes the question.
Babak Hodjat
Deep-Dive Into Agentic Workflows, w/ Cognizant’s Head of AI
Translate governance principles into scopes, thresholds, sources, and escalation.
Dan Botero
The Wild Story of “Octavius Fabrius,” the World’s First AI Agent to (Kind of) Land a Job, w/ Dan Botero
Follow a coached agent across identity, platforms, permissions, and an alarming boundary failure.
Carissa Véliz
When AI Forecasts Become Self-Fulfilling (and Who This Hurts), w/ Carissa Véliz
Make contestability concrete when a prediction affects a person’s opportunity.
Rick Caccia
Why “Shadow AI” is the Biggest Business AI Story No One is Talking About, w/ Rick Caccia
Confront the ordinary reality of useful tools appearing before formal approval.
Ankit Panda & Andrew Reddie
The Real Risks of AI and Nuclear War, w/ Ankit Panda & Andrew Reddie
Stress-test human authority where time pressure and consequence are extreme.
Carry it forward
Design the override before the emergency. If nobody can reconstruct, challenge, and stop the system, the human is decoration.
Continue with the related deep dives, evidence notes, or full conversations.