🔍 Read the full analysis: Why Altman’s PC Memory Remark Matters To The AI Industry on ThorstenMeyerAI.com
Get smart everyday buys delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
TL;DR
OpenAI CEO Sam Altman said on a podcast that it is “quite terrible” for limited memory to sit in personal computers that he estimated are used only 5% of the time. Days later, he met Samsung executives in South Korea, but no new memory supply agreement was announced. The episode highlights a growing competition between AI data centers and consumer devices for memory chips, while the exact outcome of the meeting remains undisclosed.
OpenAI CEO Sam Altman said October 29 that it was “quite terrible” for scarce memory to sit in personal computers that he estimated are used only 5% of the time, arguing that shared data centers could use it more fully. Two days later, he met about 50 Samsung executives in South Korea as concern grows over memory supplies for AI infrastructure; Samsung confirmed the meeting, but no new supply deal was announced.
Altman made the remark on Cheeky Pint, a podcast hosted by Stripe co-founder John Collison. In the discussion, Collison asked why high memory prices had not yet brought more supply. Altman replied that “the invisible hand is failing right now” and criticized the allocation of limited memory to personal computers that are not in constant use. His 5% figure is an estimate he gave during the conversation, not a reported industry measurement.
Samsung said the October 31 gathering took place at its Human Resources Development Center in Yongin and brought together about 50 senior executives from its electronics, finance and biotech affiliates. It described the session as leadership training on AI and said participants discussed the technology’s direction, its effect on management and decision-making, and new markets. Samsung’s account did not identify a new chip contract or specify supply volumes, prices or commitments.
Other details have been reported but not confirmed by both companies. Bloomberg reported that the visit was intended to deepen cooperation on memory chips and AI infrastructure. The Seoul Economic Daily reported that OpenAI hardware executive Richard Ho accompanied Altman and that OpenAI’s planned accelerator could use Samsung’s next-generation HBM4 memory. Those points have not been confirmed by OpenAI or Samsung. Yonhap also reported, citing industry sources, that Altman met SK Group chairman Chey Tae-won and SK hynix executives in Seoul; the discussion’s subject was not disclosed.
Infographic 01 · Confirmed, reported, not announced
What we actually know about Altman’s Samsung meeting
Samsung confirmed the meeting and framed it as AI leadership training. The memory angle comes from reporting. No deal was announced.
- Meeting with about 50 presidents and executivesConfirmed by Samsung
Electronics, finance and biotech affiliates; Samsung Human Resources Development Center. Theme: “The Next 10 Years Shaped by AI: The Enterprise’s Choice”.
- Altman: the edge comes from how you use AI, not from adopting itConfirmed by Samsung
As relayed by Samsung via Yonhap; quoted similarly by the Seoul Economic Daily.
- Aim: deeper cooperation on memory chips and AI infrastructureReported
Bloomberg, “as global shortages intensify”.
- OpenAI’s chip-hardware chief traveled along; HBM4 linkOne outlet
Seoul Economic Daily, citing industry sources. Not confirmed by either company.
- SK chairman Chey Tae-won met Altman the same dayOne outlet
Yonhap, citing industry sources. Content not disclosed.
- Volumes, prices, a new contractNot announced
Nothing announced. No statement from OpenAI on the visit.
How Samsung and OpenAI got here
- 1 Oct 2025Letter of intent: memory for Stargate, data centers, floating data centersSamsung and OpenAI announcement
- Jun 2026Samsung rolls out ChatGPT Enterprise and Codex to employeesOpenAI, as reported by the Korea Herald
- Jul 2026Samsung chairman Lee Jae-yong visits OpenAI in San FranciscoKorea Herald
- 29 Oct 2026Samsung’s record quarter; Altman’s “quite terrible” DRAM remarkSamsung results · Cheeky Pint podcast
- 31 Oct 2026Altman meets about 50 Samsung executives in YonginSamsung via Yonhap · Bloomberg
“Confirmed by Samsung” means Samsung’s own account, as relayed by Yonhap and the Korea Herald. A letter of intent is a statement of plans, not a purchase order.
Sources: Samsung via Yonhap, 31 Oct 2026, Korea Herald, 31 Oct 2026, Bloomberg via The Straits Times, 31 Oct 2026, Seoul Economic Daily, 1 Nov 2026, Samsung and OpenAI letter of intent, 1 Oct 2025.
Powered by Thorsten Meyer AI
AI Data Centers Compete for Memory
The dispute Altman’s comment points to is about who gets access to limited memory, not whether consumer PCs are useful. AI systems rely on large amounts of memory, including high-bandwidth memory used alongside advanced processors in data centers. If suppliers direct more capacity toward AI servers, PC makers may face tighter availability or higher costs for the memory used in ordinary computers.
That could affect buyers even if retail shelves remain stocked. The source article cites TrendForce’s forecast that conventional DRAM contract prices would rise 10% to 15% in the fourth quarter; it also reports TrendForce warned that PC memory supply could decline in 2027 as more capacity shifts to servers. These are forecasts, not confirmed outcomes. Possible effects include price increases or manufacturers choosing lower memory and storage configurations in some mainstream devices.
Altman’s argument also reflects OpenAI’s commercial interest: the company operates cloud AI services that run in data centers. His view that shared infrastructure can put hardware to more use is an argument for centralizing computing, not an independent finding that consumer-owned memory is wasted. The allocation question matters to businesses weighing whether to buy equipment or rent AI capacity, as well as to device makers competing with data-center customers for components.
Products Worth Considering
Apple Mac mini Desktop Computer with M4 chip, 10‑core CPU and GPU, 24GB Unified Memory, 512GB SSD, Gigabit Ethernet
Apple MacBook Pro 14 2021 with M1 Max, 64GB RAM, 2TB SSD, Silver
Apple MacBook Air 15-inch (2026) with M5 chip, 16GB memory, 512GB SSD, Silver
As an affiliate, we earn on qualifying purchases.
Samsung’s Existing OpenAI Ties
The Yongin meeting took place amid an existing relationship. In October 2025, OpenAI and four Samsung companies signed a letter of intent. Samsung said at the time that OpenAI’s memory demand was projected to reach “up to 900,000 DRAM wafers per month.” That was a projection cited in connection with the letter, not evidence that Samsung had committed to deliver that volume or that it is being supplied today.
In June 2026, OpenAI said Samsung Electronics would introduce ChatGPT Enterprise and Codex to employees in Korea and its global device division. The Korea Herald reported that Samsung executive chairman Lee Jae-yong visited OpenAI in San Francisco in July. The source article also reports that Samsung’s chip division recorded 106.5 trillion won of the company’s 107.4 trillion won in operating profit in the third quarter. Those results show the division’s importance to Samsung, but do not establish that the Altman meeting produced a supply arrangement.
Samsung characterized the October 31 event as an AI leadership program, while Bloomberg connected the visit to memory and infrastructure cooperation. The distinction matters: a confirmed meeting and a reported business purpose are not the same as a confirmed deal.
““From a global allocation perspective, it’s quite terrible” that so much limited memory sits in personal computers “being used 5% of the time.””
— Sam Altman, speaking on John Collison’s Cheeky Pint podcast
No Public Chip Deal Yet
Samsung has not announced a new memory supply contract with OpenAI from the October 31 meeting. The companies have not disclosed whether they discussed volumes, prices, delivery schedules or production capacity. Samsung also did not publish a full attendee list.
The reported participation of Richard Ho, possible use of Samsung HBM4 in an OpenAI accelerator, and the meeting’s specific focus on memory cooperation remain unconfirmed by the companies. The 5% PC-use figure is Altman’s estimate, and the source material provides no independent measurement or broader baseline for it. It is also unclear how much, if at all, new AI demand will reduce memory available to PC makers or raise prices for consumers.
Watch Supply and Pricing Signals
The clearest near-term indicators will be company announcements and supplier forecasts. A confirmed contract would be expected to clarify whether Samsung will provide memory to OpenAI, what kind of chips are involved and when deliveries could begin. Until then, the October meeting confirms contact, not a deal.
For device buyers and businesses, the next signals are memory prices and PC supply. TrendForce’s cited fourth-quarter price forecast and its warning about potential 2027 PC supply declines describe risks to monitor, not settled results. Manufacturers’ product specifications and pricing may show whether tighter supply is being passed on through higher costs or lower configurations.
Key Questions
What did Altman say about memory in personal computers?
On the Cheeky Pint podcast, Altman said it was “quite terrible” from a global allocation perspective for scarce memory to sit in personal computers he estimated are used 5% of the time. That percentage was his estimate, not a verified measurement.
Did Samsung and OpenAI announce a new memory deal?
No. Samsung confirmed that Altman met about 50 executives in Yongin on October 31, but neither company announced a new supply contract, volumes or prices.
Why does AI demand affect PC memory?
AI data centers use large quantities of memory, and increased demand for server components can compete with other customers for supply. The source material cites a forecast of higher conventional DRAM prices and a warning that PC memory supply could decline in 2027, but those outcomes are not certain.
Is OpenAI’s accelerator confirmed to use Samsung HBM4?
No. The Seoul Economic Daily reported that possibility, citing industry sources. OpenAI and Samsung have not confirmed it.
What should readers watch next?
Watch for a formal supply announcement, further company statements about OpenAI’s hardware plans, and updates to memory-price and PC-supply forecasts. So far, the meeting is confirmed, but its commercial outcome is not public.
Source: ThorstenMeyerAI.com
