AI writes code faster than ever. But what does that speed actually cost? Two findings from this year’s research are hard to ignore: AI token spend is turning into a material, unpredictable line item, and technical debt is shifting from the code level to the architecture level, right as AI generates code faster than teams can review it. That’s what we get into in this episode of the SIGNAL podcast, with Luc Brandts, CEO of Software Improvement Group. This conversation draws on our upcoming State of Software 2026 report, our annual look at the hidden costs and risks of software, this year measured against the full force of AI adoption. In this episode, Werner Heijstek sits down with Luc to talk about: - the real cost of AI-generated code, from token spend to review capacity, - why architecture, not the code itself, is where technical debt gets expensive, and Chapters (00:00) Intro(02:14) Agentic software engineering(03:30) AI output(05:45) How tokens work(07:17) AI Tokenomics(10:42) Pricing confusion(15:19) Architectural debt(18:46) Deterministic and AI(20:48) Metrics and KPIs that resonate Download the State of Software 2026 report here: https://www.softwareimprovementgroup.com/publications/state-of-software-2026/ The SIGNAL podcast, powered by Software Improvement Group, is a monthly show where we turn complex IT topics into business clarity. Hosted by Werner Heijstek, Senior Director at Software Improvement Group. Subscribe to never miss an episode and visit our website for more information: https://www.softwareimprovementgroup.com/ #podcast #ai #artificialintelligence #technicaldebt #aitokens #ceo #cto #cio #businesspodcast #technology #businessstrategy #softwareengineering _________ Gartner, Magic Quadrant for Technical Debt Management Tools, By Tigran Egiazarov et. al, 20 May 2026. Gartner, Market Guide for Technical Debt Management Tools, By Tigran Egiazarov et. al, 13 April 2026. GARTNER is a trademark of Gartner, Inc. and/or its affiliates. Magic Quadrant is a registered trademark of Gartner, Inc. and/or its affiliates and is used herein with permission. All rights reserved. Gartner does not endorse any company, vendor, product or service depicted in its publications, and does not advise technology users to select only those vendors with the highest ratings or other designation. Gartner publications consist of the opinions of Gartner’s business and technology insights organization and should not be construed as statements of fact. Gartner disclaims all warranties, expressed or implied, with respect to this publication, including any warranties of merchantability or fitness for a particular purpose.
