EXPLORINEER
en

US — The Verge

Sam Altman addresses RAM comments: people should ‘put as much as they want in their computers’

AI1 min read

OpenAI CEO Sam Altman says he wants everyone to have lots of computer memory. His comments came after a podcast remark that seemed to say the opposite. The story is about the fight over RAM, the memory chips computers use.

What happened

Sam Altman, the head of OpenAI, spent election night answering critics. He changed his name on X to "Ram Altman." He said he wants people to put as much memory as they want in their computers. He said his words were taken out of context. This came after a podcast interview where he said it is bad for everyone to keep a lot of the world's limited RAM in their own computers, where it is used only a small part of the time.

Key facts

  • Altman said personal computers use RAM about 5% of the time.
  • Intel expects the RAM shortage to last through at least next year.
  • SK Hynix predicted the shortage could last until 2030.
  • Samsung expects the shortage to last until at least 2028.

Why RAM is in short supply

RAM is the fast memory a computer uses to run programs. AI data centers, which are huge buildings full of computers, have been buying up much of the world's RAM. This has made supply tight. As a result, makers of computers and devices have raised prices.

Why it matters

Many people find it hard to buy the memory they want because of higher prices. Altman's remarks show a clash between AI companies that need huge amounts of memory and everyday users who want more for their own devices. The shortage may last for years.

What to watch

  • Whether memory makers build more RAM production, as Altman has urged.
  • Whether RAM prices fall or stay high in the coming years.
  • How AI data centers and home users compete for the same memory chips.

Glossary

RAM
Short-term memory inside a computer. It holds the data a computer is using right now.
OpenAI
An AI company that makes ChatGPT and other AI tools.
AI data centers
Large buildings full of powerful computers that run AI programs.
Intel
A company that makes computer chips, including processors.
SK Hynix
A South Korean company that makes memory chips.

Ad

Add 150 words. Learn 10 a day.

Dicter word books send only the words you learn to your review notes.

More in this topic

AIVnExpress

What can Vietnamese-language Siri AI do on iPhone?

Siri AI in the iOS 27.3 Beta lets users interact with it by voice in Vietnamese, carry out tasks on the device, or ask questions like other AI chatbots. The article explains what the Vietnamese-language Siri AI can do on iPhone.

AITechNode

Tencent launches Hy4 with vision and computer-use capabilities

Tencent launched Hy4 on November 4, adding visual understanding and computer-use capabilities to its flagship AI model. The 770-billion-parameter model can interpret charts and documents, reason across multiple images, and operate command-line tools and graphical interfaces. Tencent plans to release its weights and a system card shortly, and says a lower-cost Hy4 flash model is also planned.

AITechNode

ByteDance reportedly tightens access to Claude and Codex, even for employees paying out of pocket

According to recent reports, ByteDance has since early this month restricted employees from using personally purchased accounts for third-party AI tools such as Claude and Codex to access internal resources, covering all L4-level code repositories and Feishu documents. ByteDance vice president of technology Hong Dingkun said the move is driven mainly by data security concerns, and denied that it aims to force employees onto the company's own products. The restrictions are being rolled out in phases and are expected to affect thousands of technical employees.

AIITmedia

GitHub Releases "Copilot Fusion" Generally Available, Selecting the Best AI Model for Each Task to Balance Performance and Cost

GitHub announced the general availability of GitHub Copilot Fusion at GitHub Universe 2026, a multi-model feature that uses one or more AI models depending on each task's characteristics. It offers three patterns, Single, Cascade, and Critique, along with load levels (Standard, High, Max, and Fast) to balance quality, cost, and latency. A public preview also lets users run tasks locally on Fusion Standard with MAI Code 1.1 Flash or Qwen3.8 Flash Next.