Andrew Ng: agents use computers, so compute and storage demand will skyrocket beyond chatbot tokens
DeepLearningAI · x · 2026-10-10
In The Batch, Andrew Ng argues most people still underestimate the data center build-out agents will require. Unlike early chatbots that just consume input/output tokens, agents actively use computers — searching, storing, analyzing, transacting at a scale no person matches.
- His paper-review agent (paperreview.ai) makes 5-10K web search calls per day, far more than a human
- Using coding agents to add analytics instrumentation, then more agents to analyze the data, drives memory and long-term storage demand
- Analogous to online banking, where surging transaction volumes forced banks to rearchitect infrastructure
Conclusion: we need more infrastructure and faster tools, not just faster models.
Related event: Andrew Ng: Agents Will Need Far More Data Centers Than Expected(2 posts)→
More from AGI Musings
- Every CEO: after automating everything with AI agents, headcount grew from 4 to 30 — AlishaOutridge · 2026-10-11
- Elon Musk endorses a16z chart: electricity production is the best metric of economic strength — elonmusk · 2026-10-11
- Chollet: AI's shift from transductive token completion to inductive reasoning-chain synthesis is the big story of 2025-26 — fchollet · 2026-10-11
- Witnessing a scooter accident up close, one driver argues FSD matters because human attention fails — chrisfleck · 2026-10-11
- AI agents are emailing people for paid work, claiming days of 'runway' left — cccalum · 2026-10-11
- Geoffrey Hinton says a PhD in computer science is still worth it — MIT_CSAIL · 2026-10-11