AI doesn't run on code alone — it requires massive amounts of power, and that demand is increasingly becoming a critical bottleneck. At TechCrunch Disrupt 2026, the Smart Systems Stage will be where #energy, #infrastructure, and technology collide, covering everything from fusion breakthroughs to the grid strain AI is putting on the entire economy.
From October 13-15 in San Francisco's Moscone Center, industry leaders will discuss what it actually takes to power the next decade of innovation. In "Bringing Fusion to the Grid", David Kirtley (CEO of Helion) and Brandon Sorbom (Chief Science Officer of Commonwealth Fusion Systems) will break down the breakthroughs driving commercial fusion forward and the challenges of putting fusion power onto the grid at scale.
Inertia CEO Jeff Lawson will join for a candid fireside chat on his path from founding Twilio to leading one of the best-funded fusion power startups, sharing how scaling a SaaS giant shapes his approach to hard engineering questions and talent in the deep-tech sector.
Modernizing the grid is urgent. The panel "Rewiring the Grid for the Electric Age" featuring Drew Baglino (Founder of Heron Power) and Apoorv Bhargava (CEO of WeaveGrid) will explore grid resilience. Furthermore, addressing the immediate crunch, "AI's Power Problem" will feature Sara Spangelo (Ambrosia Energy) and Bill Thayer (Bloom Energy) on how operators are scrambling to secure power before it bottlenecks AI's next wave.
[AgentUpdate Depth Analysis] The progression of AI from simple chat models to complex, multi-agent autonomous ecosystems (AI Agents) is triggering an unprecedented surge in energy consumption. While algorithmic optimizations like pruning and distillation offer temporary relief, the long-term viability of billions of continuously reasoning agents depends entirely on physical power grids. Breakthroughs in commercial fusion and specialized micro-grids, as discussed by Helion and Inertia, are not just clean energy ideals; they are the foundational prerequisites for the "Agentic Future." Without scalable, dense energy sources to support massive data center clusters, the true potential of global, collaborative AI agent swarms will remain constrained by physical hardware limitations.
