Elon Musk has responded to comments by OpenAI chief executive Sam Altman about how personal computers use memory, saying that people deserve access to their own computing resources as the discussion over global memory supply and AI infrastructure gains attention.
Altman’s remarks centred on the allocation of dynamic random-access memory, or DRAM, which is used by personal computers and other computing devices. He argued that, from a global resource perspective, it is inefficient for large amounts of memory to remain in individual computers when they are used only occasionally.
Altman said that many personal computers hold substantial amounts of the world’s limited DRAM supply while using that memory for only a small portion of the time. He suggested that cloud computing could make more efficient use of available memory by allowing computing resources to be shared across users and workloads.
The comments were made during an interview with YouTuber Cheeky Pint while Altman was discussing constraints on computing resources in the AI industry.
Musk’s response, reported by India Today on November 2, came amid online discussion about whether consumers should retain local computing capacity or increasingly depend on cloud-based AI services.
The exchange highlights a broader debate over how computing resources should be distributed as demand for AI systems grows. Large AI models require significant computing infrastructure, including memory, processors and data centres. At the same time, personal computers use RAM for everyday applications, gaming, productivity software and, increasingly, AI workloads that run directly on users’ devices.
Local AI processing can allow users to run certain models without sending every request to a remote service. Cloud-based systems, by contrast, can pool computing capacity across many users and offer access to larger models, depending on the service and workload.
Altman’s comments drew criticism online from users who argued that consumers should not be expected to give up personal computing resources to address supply constraints faced by technology companies.
The debate reflects competing priorities: improving overall utilisation of scarce computing resources while preserving users’ ability to own and control the hardware in their personal devices.
Neither the discussion nor the public reaction establishes that personal computer users are responsible for shortages in memory supply. Instead, the exchange has brought renewed attention to the growing demand for computing infrastructure as AI adoption expands.
