Why 95% of All AI Projects Fail - Boris Scharinger | Podcast #174 @JousefM
Why 95% of All AI Projects Fail - Boris Scharinger | Podcast #174  @JousefM
Uploaded September 2026 | Updated November 2026, 2 days ago
πŸ”— Connect with Boris Scharinger on LinkedIn: de.linkedin.com/in/borisscharinger

Boris Scharinger is an Industrial AI Strategist at Siemens and the author of Industrial AI: From Pilot to Profit. He has spent years running AI strategy across hundreds of use cases inside one of the world's largest industrial companies, and he has watched most of them die somewhere between the proof of concept and the production line.

In this episode, we sit down to discuss why 95% of AI proofs of concept never get deployed, the 5-10x cost multiplier between a successful pilot and a shipped solution that almost nobody budgets for, and how to tell a commodity use case apart from a genuinely differentiating one.

We also get into the harder questions: why industrial AI does not scale the way consumer AI does, what the "harness" around a model really costs to build, how to convince a CFO to fund foundational data capability before a single use case has proven ROI, and why engineers instinctively resist AI in the first place.

If you are running an AI initiative in an industrial or engineering organization, or you are trying to figure out why your pilots keep stalling, you will find this episode useful. Tune in!

ONLINE PRESENCE

🌍 My website - jousefmurad.com
πŸ’Œ My weekly science newsletter - jousef.substack.com
πŸ“Έ Instagram - / jousefmrd
🐦 Twitter - https://x.com/Jousefm2
CONTACT:
β€”β€”β€”β€”β€”β€”β€”β€”
If you need help or have any questions or want to collaborate feel free to reach out to me via email: support@jousefmurad.com

#IndustrialAI #Siemens #AgenticAI
Why 95% of All AI Projects Fail - Boris Scharinger | Podcast #174Simulating Reality with Simscape: Multi-Body Dynamics & AI – Steve Miller | Podcast #152How to Train a Better AI Model Without More DataHuman Verification Of AI OutputGeometric Deep Learning - Altairs PhysicsAI - Eamon Whalen & Jonathan Ollar | Podcast #142Why MATLAB Online Has 2.8M UsersWhy Fast AI Prototypes Beat Perfect CodeAgenticAI - Changing The Fundementals Of Engineering WorkflowsThe Future of Engineering & Additive Manufacturing  - Michael Robinson | Podcast #135Why E-Motors Fail Without Temperature InsightGenerative AI for Protein Engineering - Franziska Geiger  | Podcast #131Connected Design & Visualization in Autodesk Alias - Barry Kimball and Jakob Lohse | Deep Dive
Jousef Murad | Deep Dive |

Why 95% of All AI Projects Fail - Boris Scharinger | Podcast #174

SHARE TO X SHARE TO REDDIT SHARE TO FACEBOOK WALLPAPER