The State of AI: Elon’s $1T Package, Apple’s $600B for Trump & How Startups Win w/ Dave, AWG & Blitzy Founders Brian Elliott & Sid Pardeshi | EP #193
Timestamps are as accurate as they can be but may be slightly off. We encourage you to listen to the full context.
Podcast Summary
In this episode, Peter Diamandis and his co-hosts explore the trillion-dollar landscape reshaping entrepreneurship, featuring Blitzy founders Brian Elliott and Sid Pardeshi who just announced a groundbreaking achievement. The West Point alum and NVIDIA veteran reveal how their enterprise AI platform crushed the industry-standard SWE bench with an unprecedented (34:23) 86.8% score—shattering previous records while processing up to 100 million lines of legacy code. From automating decades-old COBOL systems (23:58) to delivering 5x development velocity gains in enterprises, Blitzy exemplifies how nimble startups can ride the wave of trillion-dollar AI investments rather than be crushed by them—proving that understanding deep enterprise problems and building on the shoulders of giants creates unstoppable competitive advantage.
Speakers
Peter Diamandis (Host)
Entrepreneur, bestselling author, and founder of the XPRIZE Foundation and Singularity University. He hosts the Moonshots podcast and focuses on exponential technologies that create abundance.
Dave Blunden
Venture capitalist at Link Ventures, investing in early-stage AI companies. His portfolio includes unicorns like Mercor and breakthrough companies like Blitzy, focusing on entrepreneurs who understand deep domain problems.
Alex Wiesner Gross
AI researcher and physicist with expertise in complex systems and artificial intelligence. He provides technical analysis on breakthrough AI developments and the implications of exponential technology growth.
Brian Elliott
Co-founder and CEO of Blitzy, West Point graduate and Harvard Business School alum. Former Army Ranger who served in Syria, now building enterprise-grade autonomous software development platforms that tackle millions of lines of legacy code.
Sid Pardeschi
Co-founder and CTO of Blitzy, BITS graduate (India's MIT equivalent) and NVIDIA veteran who witnessed the company's growth from billions to trillion-dollar valuation. Holds 27 patents and architected Blitzy's breakthrough 86.8% SWE-bench verified score.
Key Takeaways
Surf the Trillion-Dollar Capital Waves
When massive CapEx gets deployed (like the trillion-dollar investments flooding AI), become a power customer, not a competitor. Use those resources to make your product exponentially better—Blitzy thrives because every new model release strengthens their platform. (44:10)
Go Deep on Domain Expertise
While 19-year-olds build photo apps, tackle the unglamorous trillion-dollar markets that require specialized knowledge. COBOL-to-Java migrations, actuarial risk systems, semiconductor manufacturing—these deep domains create unassailable moats. (30:56)
Screen for Mission-Driven Ambition First
Before evaluating technical skills, identify people who see their work as "one of the most important things they will do in their lifetime." This creates the 997 work culture that crushes benchmarks and attracts top talent organically. (83:48)
Build for Multi-Week Inference Cycles
While competitors get stuck in 5-7 minute "no man's land," architect for extended runs that deliver substantial results. Blitzy runs 12-hour to multi-week processes, tackling 60 million lines of code where others fail at scale. (32:24)
Obsess Over Quality at Any Compute Cost
The other side of an AI-generated pull request is exponentially more expensive human labor. Invest heavily in validation, compilation, and testing systems—high-quality code justifies premium pricing and creates sustainable competitive advantage. (23:16)
Statistics & Facts
- 86.8% achievement on SWE Bench verified, representing the highest score ever recorded on the benchmark, surpassing previous records by over 11 percentage points. This is described as "completely unprecedented" for AI software engineering capabilities. (33:54)
- Mercor achieved a $10 billion valuation with a $500 million revenue run rate after just two years, starting from a $30 million valuation when Dave invested. This represents one of the fastest growth trajectories from startup to near-decacorn status in venture capital history. (12:15)
- Blitzy successfully handles up to 60 million lines of code, with typical enterprise clients seeing 20 million line codebases processed. The system has ingested up to 2 billion lines of code overall when counting updates and transformations. (38:59)
