In the AI era, the fastest winners are not the best idea generators — they’re the best idea killers. The AI boom has unleashed an explosion of startup ideas, demos, and hype. Yet most AI startups that look impressive today will quietly disappear tomorrow. Why? Because “cool AI” is often mistaken for "valuable AI".
In this candid, contrarian talk, Mahdi Fanidisfani draws from over a decade of hands-on leadership across startups, scale-ups, and mission-critical institutions to unpack what actually separates signal from noise in today’s AI startup landscape.
Speaker:
Mahdi Fanidisfani — Tech Director at MR MARVIS | Founder & CEO | AI & Digital Transformation Advisor
What you’ll learn:
- Why the AI boom produces more ideas than great companies
- The most common failure patterns in AI startups (solution-first thinking, weak data moats, demo-driven products)
- Why killing ideas early is a competitive advantage — and how to do it systematically
- Practical frameworks to stress-test ideas fast: market pull, defensibility, and data reality checks
- The real difference between “cool AI” and AI that creates lasting value
- Lessons from watching ideas fail, pivot, and survive in fast-moving AI markets
Mahdi shares the exact stress tests and decision frameworks he uses when advising founders — shaped by real scar tissue from building companies, leading large engineering teams, and saying “no” to many seductive ideas.
If you’ve ever wondered why so many AI startups shine at demo day but vanish six months later, this talk is for you.
