Merlin Donald
Foreword
By Edo Segal
In the winter of 2025, I spent thirty days building something that should have taken months. An AI-powered concierge kiosk that could hold conversations in multiple languages, generate custom music, and learn from every interaction. When it came alive on the CES showfloor, I felt the vertigo of standing at a threshold I hadn't quite prepared for.
That vertigo is why Merlin Donald's framework matters now.
Most of the AI discourse operates in what Donald would call the theoretic layer—the realm of algorithms, outputs, and productivity metrics. We measure tokens generated, tasks completed, efficiency gains. We debate whether AI thinks or merely predicts. We worry about jobs displaced and capabilities expanded. All of this matters, but it misses something essential about what's actually happening to human cognition in