Nikesh Arora on The Future of Token Costs | Memory Becoming the Moat & Why Enterprise AI Isn't Ready

AI is fundamentally reshaping enterprise workflows, moving from passive SaaS containers to intelligent, opinionated applications that require a radical rethink of business processes. While consumer models prioritize breadth, enterprise success depends on depth—specifically, the ability to integrate proprietary context and edge-case training to enable agentic decision-making. Cybersecurity serves as a critical accelerant in this transition; as AI tools weaponize vulnerability discovery, enterprises are forced to accelerate their security posture. Rather than "AI washing" products, companies must adopt an incremental, Tesla-like approach to autonomy, focusing on proprietary data and memory to build defensible moats. Ultimately, the future of enterprise value will accrue to those who successfully bridge the gap between frontier model capabilities and the deep, context-specific intelligence required for autonomous business operations.