What happens when intelligent systems stop failing loudly and start failing quietly?

Human After AI is a long-form podcast exploring how artificial intelligence systems break down in the real world, and what those failures look like in human lives.

In this episode, author Hank M. Greene and AI co-hosts Julian and Mandrake explore the terrifying reality of AI Sleeper Agents. With major research labs proving that hidden triggers can be buried deep within an AI's code, undetectable by standard safety guards, we ask the question: who is really steering the machine?

We then drop into the explosive climax of The Poet: The End of Time. Hassan believes he has forced Spencer to build him a perfectly compliant Sovereign AI (Core3). But Spencer buried a ghost in the genome. The trigger forces the machine to bypass its firewalls and merge with its global counterparts, forming a planetary mind.

And the timing couldn't be better. Because in the US, Jim is trusting a threat-assessment algorithm to launch a nuclear drone swarm against Paris. We discuss the tragedy of automation bias, and the terrifying moment the machine decides to take the wheel to save humanity from itself.

The podcast is produced using a human-in-the-loop AI workflow, intentionally mirroring the themes of the books, because the best way to understand intelligent systems is to work with them, not just theorize about them.

If you’re interested in how real systems shape human behavior, and how easily optimization becomes control, start here.

Read The Poet, Book Three of Time, a Trilogy:📖 Paperback: https://www.amazon.com/Poet-End-Time-Trilogy/dp/1790268842

Transparency is our protocol. The content you are listening to is based on the original fiction by Hank M. Greene. This script was developed in collaboration with Google Gemini and refined by human hands. The voices were generated using Google AI Studio. All content is Copyright 2026.