The Twenty Minute VC (20VC): Venture Capital | Startup Funding | The Pitch 20VC: Raising $500M To Compete in the Race for AGI | Will Scaling Laws Continue: Is Access to Compute Everything | Will Nvidia Continue To Dominate | The Biggest Bottlenecks in the Race for AGI with Eiso Kant, CTO @ Poolside
102 snips
Oct 7, 2024 
Eiso Kant, Co-founder and CEO of Poolside.ai, dives into the competitive world of Artificial General Intelligence (AGI) funding, sharing insights on their recent $500M raise. He discusses how Poolside differentiates itself from other AI models and the challenges of competing with giants like Nvidia. The conversation touches on scaling laws, the future of model performance, and whether $600M is sufficient to remain relevant in the rapidly evolving AI landscape. Kant also explores the dynamics of AI infrastructure and the emotional journey of entrepreneurs.
AI Snips
Chapters
Books
Transcript
Episode notes
Cost of Entering the AI Race
- $100 billion might be the entry price for becoming a hyperscaler in the AI space, not for innovating in AGI.
- Dollars are critical for compute, but not a guarantee of success in the AGI race.
Data Center Evolution
- Model evolution outpaces data center development; frontier models won’t be trained on the same data center twice.
- Data center requirements change significantly due to scaling model size and the need for interconnected servers during training.
Nvidia's Blackwell Chip
- Nvidia's Blackwell chip offers potential performance gains, especially for inference, not training.
- Upgrading to new generations of chips isn’t always necessary, even with performance improvements.
