When the Machines Read Your Mind | a Reflect Podcast by Ed Fassio

What happens when artificial intelligence starts guessing what we’re thinking… and occasionally gets it right? In this episode, we dive into MIT’s groundbreaking research on theory of mind in human-AI collaboration—the science of how machines model our beliefs, intentions and blind spots, and how we’re (sometimes unsuccessfully) trying to model theirs.

From search-and-rescue robots that politely correct human teammates, to brain-wave studies showing what really happens when you let ChatGPT “help” with your writing, this conversation pulls back the curtain on a future where humans and algorithms learn to think together.

And somewhere between the science and the slightly uncomfortable self-reflection, I share a bit about my own work on K2A—my ongoing attempt to drag knowledge out of the abstract and bolt it into tools that actually work in the real world. If theory of mind is the next big frontier for intelligent collaboration, then K2A is my little construction project on that frontier.

Expect a mix of mentorship, dry humor, and straight talk about the messy, exhilarating collaboration unfolding between our neurons and the machines we built to “help” us think. Because ready or not, we’re coevolving with our technology… and the music is just beginning.

Tune in to explore:
• How MIT is teaching AI to infer human intentions
• Why human-AI teams sometimes flop
• What AI tools might be doing to your brain activity
• The philosophical debate over whether AI truly “understands” anything
• How K2A fits into the bigger picture of cognitive collaboration

If you’ve ever wondered whether AI can read your mind—or whether you’re still using all of yours—this episode is your stop.

About Ed Fassio

Ed Fassio has spent his career sitting at the crossroads of human insight and technological possibility, helping organizations turn scattered knowledge into focused, practical action. With a background that zigzags through strategy, innovation, and leadership development, Ed has built a reputation for cutting through noise with that trademark blend of straight talk, dry humor and “let’s-get-this-working” pragmatism. His current work on K2A (Knowledge to Agent) pushes that mission even further, designing systems that help humans and AI think with each other—not past each other. If there’s a better way to turn ideas into impact, Ed is usually in the middle of it, sleeves rolled up, asking the questions everyone else tiptoes around.