All-In with Chamath, Jason, Sacks & Friedberg

Arvind Krishna & Nikesh Arora: Quantum Threat, Open vs Frontier AI, AI Cyber Warfare & SaaS Collapse

40 min

In this episode, Arvind Krishna (IBM CEO) and Nikesh Arora (Palo Alto Networks CEO) join the All-In podcast to discuss the imminent quantum computing threat to cryptography, the economics of open vs frontier AI models, and the transformation of cybersecurity into AI-driven warfare. They argue for immediate adoption of post-quantum cryptography, predict that open weight models will dominate enterprise AI usage, and forecast massive consolidation in the SaaS industry as AI makes thousands of independent software products obsolete.

Summarized by Podsumo

🎧 Listen 🎙️ Ask about this episode

✨ Key Takeaways

1

Arvind Krishna warns that quantum computers could break today's cryptography by 2030, urging immediate migration to post-quantum cryptography.
2

Nikesh Arora argues that open weight models (e.g., Chinese models) are 40x cheaper than frontier models and will dominate enterprise use for 95% of tasks, with frontier models reserved for high-stakes scenarios like active cyber breaches.
3

The discussion frames cybersecurity as modern warfare, emphasizing alliances (shared vulnerability databases), asymmetry (cheap defense), and technological superiority (AI-generated code and automated patching).
4

Both guests predict massive consolidation in SaaS: the average enterprise uses 500–5,000 vendors, but AI will enable replacement with 5–10 major platforms, disrupting the software industry.
5

Vertical integration (like IBM mainframe or NVIDIA) is seen as the winning strategy for AI stacks, locking in customers through memory, context, and end-to-end control.

💬 Notable Quotes

"If you pick the middle point, let's call it 2030, well, it will break today's cryptography. That's not a supposition, that's just a reality."

— Arvind Krishna
"You can take a Chinese open weight model, run your own GPUs, and drop the cost to about a dollar to two dollars for a million tokens. Which enterprise will not accept a 40 times cheaper answer for 95% of the work?"

— Nikesh Arora
"The average enterprise uses anywhere from 500 to 5,000 distinct software vendors. ... Could I replace 490 of them with minor AI capabilities? ... That implies massive consolidation."

— Arvind Krishna