Humans & AI — Episode 43

Matthijs De Vries — Fixing AI Hallucinations with Verifiable, Traceable Data

Nov 13, 2025·26 min·by Flor LaorgaHumans & AI Show
NOW PLAYING · Humans & AI · Episode 43 26:24
Matthijs De Vries — Fixing AI Hallucinations with Verifiable, Traceable Data
0:0026:24

Matthijs De Vries — Fixing AI Hallucinations with Verifiable, Traceable Data

Nov 13, 2025

Matthijs De Vries — Fixing AI Hallucinations with Verifiable, Traceable Data

Matthijs de Vries is the founder of Nuklai, a company focused on building reliable data infrastructure to improve how AI interacts with information. With a background that started early—he began coding at 11 and was a professional developer by 15—Matthijs has since launched several ventures, including Nexera, which works on tokenization infrastructure. His passion lies in combining blockchain technology, data, and AI to create tools that deal with some of today’s biggest challenges in the tech world.

In this episode, Matthijs shares his journey from coding prodigy to serial entrepreneur and dives into how Nuklai tackles AI hallucinations—those moments when AI generates inaccurate or misleading responses. He explains how their Nexus query engine creates clear audit trails and connects AI outputs back to their original data sources, making AI’s “thought process” much easier to verify. Along the way, he touches on the importance of responsible data use, privacy concerns in generative AI, and why the decentralization of AI still faces big hurdles. Matthijs also reflects on the lessons learned through failure and iterating on product-market fit, highlighting the need to focus on real problems backed by user feedback. Looking forward, he envisions a future where AI helps people focus less on routine tasks and more on meaningful life questions.

Related

Chen Amit — Why Finance Can't Tolerate AI Hallucinations
Chen Amit — Why Finance Can't Tolerate AI Hallucinations
Chen Amit of Tipalti on why finance can't tolerate AI hallucinations, and how to bring AI to the office of the CFO with control.
Rachel Kirkham — AI for Audit: Full-Population Analytics, Risk Scoring & Trust
Rachel Kirkham — AI for Audit: Full-Population Analytics, Risk Scoring & Trust
Rachel Kirkham, CTO at Mindbridge, on AI for audit: full-population analytics, anomaly detection, trust through transparency, and leading AI teams.
Kevin Surace — AI-First Leadership, QA Automation, and the Future of Work
Kevin Surace — AI-First Leadership, QA Automation, and the Future of Work
Kevin Surace is the CEO of Appvance and a longtime Silicon Valley builder known for moving easily between hard engineering problems and big-picture thinking about what comes next. Often described as the father of the virtual assistant, he’s led work that spans early cellular data smartphones, human-like AI assistants, and newer applications of generative AI in software testing. He also holds 94 worldwide patents and has been recognized across tech and innovation circles, including honors from Inc. Magazine, CNBC, and the World Economic Forum. What makes his background especially distinctive is that it isn’t limited to software: he’s also worked in entertainment, which gives him a practical view of how creativity and technology can feed each other rather than compete.
Anya Cheng: Why “Ideas Are Cheap” and Unique Data Is the Real Moat
Anya Cheng: Why “Ideas Are Cheap” and Unique Data Is the Real Moat
Anya Cheng is the co-founder and CEO of Taelor, an AI-powered men’s clothing subscription service built around sustainability and convenience.

Copyright © 2026 AI Frontier Network | Privacy Policy | Terms of Use