OpenAI CEO says centralizing RAM in data centers is preferable to local use amid global shortage
Sam Altman argues that concentrating memory in cloud infrastructure maximizes utilization during constrained supply periods, criticizing local inference setups.
1 source · cross-referenced
- OpenAI CEO Sam Altman said concentrating RAM in data centers is preferable to distributing it across personal devices during a global shortage.
- Altman argued that local inference setups waste limited DRAM supply, favoring cloud-based utilization for efficiency.
- The remarks were made during an interview on the Cheeky Pint podcast, alongside comments on AI’s impact on mathematical work.
OpenAI CEO Sam Altman stated in a podcast interview that it is preferable for RAM to be concentrated in data centers rather than distributed across personal computers during a global memory shortage. Altman described having DRAM in individual devices as wasteful, given low average utilization rates, and argued that cloud-based allocation maximizes efficiency in a supply-constrained environment.
During the Cheeky Pint podcast appearance, Altman also discussed AI’s impact on mathematical work. He criticized the assumption that open-source models running locally would alleviate demand for data center resources, asserting that such setups would misallocate scarce hardware globally.
Altman’s remarks come as memory suppliers like Samsung and SK Hynix have prioritized long-term agreements with cloud operators and large-scale AI projects, such as OpenAI’s Stargate initiative, reserving much of the world’s RAM output for years to come. The constrained supply has contributed to rising costs for consumer devices, including laptops and phones.
The interview underscored Altman’s perspective that centralized cloud infrastructure is the optimal path for managing hardware scarcity, even as developers and consumers explore local inference options.
- Nov 10, 2026 · Wired
Kora, a Linux-based desktop OS from Cyanogen cofounder, targets AI power users with local model integration
Trust76 - Nov 10, 2026 · One Useful Thing — Ethan Mollick
Mollick argues new AI models approach "jagged AGI" across coding, business, and humanities
Trust78 - Nov 9, 2026 · The Verge — AI
Former DOJ antitrust chief launches AI legal-research startup to challenge monopolies
Trust72