AI is supposed to reduce technical debt, but what if it’s actually making the problem worse?

In this episode of Today in Tech, host Keith Shaw sits down with Gary Hoberman, Co-Founder of Unqork, and David Ferrucci, CTO of Unqork and former IBM Watson leader, to unpack how generative AI, low-code platforms, and “vibe coding” can quickly multiply hidden risk instead of eliminating complexity.

They break down why digital transformation hasn’t solved tech debt, how AI-generated code can speed up architectural mistakes, and why governance, component reuse, and disciplined system design matter more than ever. Drawing on Gary’s experience managing global engineering organizations and Dave’s work building Watson for Jeopardy!, this conversation reveals what enterprise leaders must understand if they want to use AI without creating the next generation of legacy problems.

Key topics include

* Why tech debt keeps growing after modernization efforts

* How AI coding tools can accidentally amplify bad architecture

* The limits of low-code, no-code, and “citizen developer” platforms

* Governance and guardrails for safe enterprise AI adoption

* What the future holds for software development and AI-assisted teams