OpenAI chief executive Sam Altman “met with Samsung Groupโs top executives in South Korea on Saturday to deepen cooperation on memory chips and AI infrastructure as global shortages intensify”, Bloomberg reported on 31 October. Samsung hosted the session for “about 50 executives” in Yongin, south of Seoul.
Two days earlier, Samsung reported third-quarter capital expenditure of KRW 29 trillion, “an increase of KRW 12.2 trillion quarter-on-quarter (QoQ) and KRW 19.8 trillion year-on-year (YoY)”. By our arithmetic that puts a year-earlier quarter at KRW 9.2 trillion, so Samsung is now spending about 3.2 times as much per quarter.
Business Pill · OPERATING LEVERAGE
A short explainer on operating leverage, why profits swing more than revenue when costs are largely fixed, as in memory chipmaking. It teaches the general idea only and says nothing about this meeting.
The key insight: Samsung reported KRW 29 trillion of third-quarter capex, about 3.2 times a year earlier by our arithmetic, two days before Altman met its chip head and about 50 executives in Yongin.
Bloomberg’s report, by Yoolim Lee, ran at 06:40 UTC under the headline “OpenAIโs Altman Meets Samsung Executives as Chip Crunch Deepens”. Only its first two paragraphs were readable to us; the rest is behind the paywall.
Citing a Samsung statement, Bloomberg wrote that the executives came from Samsung’s “electronics, financial and biotechnology affiliates” and that “The discussions focused on how artificial intelligence can reshape industries, corporate decision-making and business opportunities over the next decade”.
Yonhap described “a series of leadership seminars designed to advance the company’s drive for AI transition”, held at a Samsung human resources center. It reported that Altman “highlighted that gaining a competitive edge does not hinge on simply adopting the AI technology”.
Samsung Electronics capital expenditure by quarter, KRW trillion, from its Q3 2026 results release (29 October 2026): KRW 29 trillion in Q3 2026, of which KRW 27.2 trillion in Device Solutions. Q2 2026 (KRW 16.8 trillion) and Q3 2025 (KRW 9.2 trillion) and the 3.2x ratio are our arithmetic from the release’s quarter-on-quarter and year-on-year changes; Q4 2025 and Q1 2026 are not shown.
What Was Posted on X
OpenAIโs Sam Altman reportedly made a top-secret visit to South Korea and met with Samsungโs top executives.
(According to local Korean media, he also met with SK hynix.)
Altman visited South Korea exactly one year ago, right when this whole surge in RAM prices began. Now, a year later, heโs backโand heโs met with the two memory giants.
Jukan (@jukan05) posted that Sam Altman reportedly made a top-secret visit to South Korea and met Samsung’s top executives, adding that, according to local Korean media, he also met SK hynix. Jukan says Altman visited exactly one year ago, when, in Jukan’s words, the surge in RAM prices began; that timing is the poster’s claim.
Who Was in the Room
SBS reported that Altman arrived on “a sudden two-day visit” and named the OpenAI team: “Richard Ho, Vice President of Hardware”, Kiran Mani, Andreas Hartl and Arvind KC. On Samsung’s side it listed “Jun Young-hyun, Vice Chairman and Head of Samsung Electronics’ DS (Device Solutions) Division”, Roh Tae-moon of the DX division, and the heads of Samsung SDI and Samsung Electro-Mechanics.
SBS said a meeting with chairman Lee Jae-yong “was reportedly not on today’s schedule” and that Altman left the same afternoon. It added that Altman gave the executives a “Codex Micro Keyboard” for OpenAI’s coding agent. SBS notes that its English article “was translated by AI and may contain errors”.
SBS also recalled that Samsung Electronics and OpenAI signed a letter of intent “in October of last year to cooperate on building global AI infrastructure, pushing forward with the supply of memory semiconductors required for OpenAI’s “Stargate” project”.
The Business Lever: Memory Is the Input in Short Supply
Samsung’s results release says its memory business set a record “amid capacity constraints” and that “Strong demand momentum for servers led to higher market prices across all applications”. It says Samsung “significantly scaled up sales of HBM4”.
Of the KRW 29 trillion, the release puts KRW 27.2 trillion in the Device Solutions (chip) division, about 94% by our arithmetic. For the full year Samsung expects about KRW 84.4 trillion, including KRW 76.8 trillion for that division; with KRW 52.8 trillion spent there in January to September, KRW 24.0 trillion is left for the fourth quarter, by our arithmetic.
As we read the attendee list SBS published, OpenAI brought its vice president of hardware and Samsung brought the head of its chip division, which fits Bloomberg’s description of a meeting about memory chips and AI infrastructure.
As we read Samsung’s release, about 94% of its third-quarter capex went to the chip division, by our arithmetic, while its memory business reported record results amid capacity constraints.
As we read the reports together, none of them gives a supply volume, price or contract from the meeting; the link between the visit and memory supply is Bloomberg’s framing and the 2025 letter of intent that SBS recalls.
Sam Altman, as quoted by SBS News (AI-translated), Yongin, 31 October 2026
“Simply using AI will no longer guarantee a sustainable competitive advantage moving forward. Competitive advantage will depend on how you leverage AI to move faster, build better products, and stay ahead of your competitors.”
Three Implications
FOR OPENAI SBS lists OpenAI’s vice president of hardware among the attendees; Bloomberg says the talks aimed to deepen cooperation on memory chips and AI infrastructure.
FOR SAMSUNG Its release puts KRW 27.2 trillion of KRW 29 trillion third-quarter capex in the chip division and expects about KRW 84.4 trillion for the full year.
FOR MEMORY BUYERS Samsung says server demand led to higher market prices across all applications; it leaves KRW 24.0 trillion of chip-division capex for the fourth quarter, by our arithmetic.
The framework says “Memory is becoming the new compute. HBM growth is outpacing accelerator demand, making memory the rate-limiting factor in the AI scaling equation.”
As we read it, an AI lab sending its hardware chief to a memory maker’s chip head is the kind of meeting that thesis anticipates. The framework predates this visit and does not discuss it.
What happened next (31 Oct): SK’s chairman hosts Altman too
SK Group chairman Chey Tae-won held a lunch meeting in Seoul with Sam Altman, with senior executives from SK Hynix and SK Telecom attending, Bloomingbit reported, citing the Korea Economic Daily. The paper said the two sides likely discussed HBM supply-chain cooperation and next-generation AI data centres; the specific contents were not disclosed and no new agreement or figure was reported. The meeting builds on the October 2025 SK–OpenAI partnership (a memory-supply letter of intent and an AI data-centre MOU for southwestern Korea).
What Is Not Established
None of the readable reports gives a supply volume, price, contract or investment figure from the meeting, and Samsung’s statement as quoted describes a discussion of AI strategy. Bloomberg’s framing of memory cooperation comes from its first paragraph; the rest of its report was not readable to us. We did not find a statement from OpenAI.
This Week In The AI Supercycle
Execution Fault Lines, Concentrated Demand, and Geopolitical Firewalls
This week’s issue: the AI build is not short of demand for its parts. It is short of cheap money, and its end revenue rests on a narrow base of large, often venture-funded buyers.
Altman met about 50 Samsung executives in Yongin on 31 October, with OpenAI’s hardware chief and Samsung’s chip head in the room, according to Bloomberg and SBS. Samsung’s KRW 29 trillion of third-quarter capex is about 3.2 times a year earlier, by our arithmetic.
A note on sourcing. The meeting is reported by Bloomberg, whose article was readable to us only in its first two paragraphs, and by Yonhap and SBS; SBS says its English text was translated by AI. Capex figures come from Samsung Electronics’ third-quarter results release. We did not contact OpenAI or Samsung. Figures labelled our arithmetic are our own.
Gennaro Cuofano is a CRO and tech executive who has worked in AI since late 2015, starting with bringing NLP, natural-language generation, voice and chatbot products to the marketing industry. His background is in law and finance: he holds a Master's degree in Law and an International MBA with an emphasis on corporate finance (LUISS Business School and the University of San Diego, 2012), and worked as a financial analyst at a real-estate investment firm in San Diego and as an assistant controller. His work focuses on business model strategy, business engineering and, more broadly, structural analysis: how companies make money, read from their own filings. He created FourWeekMBA and leads research there and at The Business Engineer, his newsletter on AI and business strategy, with over 95,000 subscribers and more than 1,000 published analyses. His writing has also appeared on Entrepreneur (for example, How to Digitize a Family Business in Europe) and HackerNoon. Find him on LinkedIn and Substack. See how we source.
