TechFlow Selected published a new edition of “TechFlow Intelligence Bureau” at 11:45:32 on June 18, 2026. The column said its AI Agent had completed a daily patrol across more than 200 global information sources, spanning crypto, AI and technology, filtering out 99% of the noise and retaining the signals it considered necessary. TechFlow also described its own editorial focus as in-depth Web3 reporting and reading the direction of the current. This edition brought together developments in AI costs, on-chain stock trading, hardware security, technology companies, public markets and geopolitics.
AI costs, governance disputes and product risks
In the AI and large-model section, Anthropic CEO Dario Amodei spoke openly in a recent interview about why he left OpenAI, saying he “could not trust a person whose behavioral pattern was disturbing and dishonest.” TechFlow framed the comment as a rare public statement on internal conflicts within Silicon Valley’s AI circles, and noted that community discussion treated it as validation of rumors about OpenAI’s governance disorder. The same section reported that OpenAI generated $13 billion in revenue in 2025 but recorded losses as high as $38.5 billion. Even with surging revenue, uncontrolled costs left its commercialization path under heavy pressure. TechFlow’s own sharp comment said OpenAI lost the equivalent of three and a half Twitter acquisition prices in one year, while Sam Altman’s idea that “AGI needs money-burning” is being stress-tested by the market.
Microsoft Copilot was also reported to have a “SearchLeak” vulnerability. Security researchers warned that the flaw could be used for data theft and could expose user emails and two-factor authentication codes, while Microsoft had not yet issued a formal response. DeepSeek was mentioned in two contexts: the United States added more than 100 companies to a security-risk list, but DeepSeek was not included for the time being; at the same time, users discussed clear changes in the chain of thought shown by the web version’s “fast mode.” Midjourney launched a Medical vertical application, entering the area of medical image generation. The product drew disputes over compliance and accuracy, and Hacker News recorded 523 comments questioning the ethical boundary of medical AI generation.
Solana dominates a record day for tokenized stocks
The crypto and Web3 section focused on Solana’s role as the main venue for on-chain stock trading. According to the roundup, tokenized stocks set a new record for the largest single-day trading volume, reaching $142 million yesterday, with 97% of that activity taking place on Solana. SpaceX ($SPCX) recorded $3.1 billion in cumulative trading volume across the nine days before and after its IPO, while Hyperliquid was described as becoming a new venue for pre-IPO price discovery. The cited sources for this item included Twitter/Buffalu and Twitter/Grayscale, placing the development at the intersection of on-chain assets and traditional equity narratives.
Other Web3 items included OKX announcing spot trading for the RE token and converting pre-market futures into standard perpetual contracts. South Korean exchange Upbit simultaneously classified AERGO as a trading warning project. TechFlow cited Twitter/6551News for this update. Together with the record tokenized-stock trading activity on Solana, these items formed the main crypto-related signals in the edition.
Chips, memory, OLED production and security features
The chips and hardware section said AMD had quietly removed a memory encryption feature from consumer-grade Ryzen CPUs in a new AGESA firmware release. TechFlow wrote that the security feature was silently deleted, leaving users unaware and exposed to vulnerability risks, while AMD engineers remained silent. On AI memory, SK Hynix began supplying samples of its next-generation HBM4E 12-layer stacked AI memory. The roundup described the AI chip arms race as moving into memory, with TSMC capacity shortages pushing a surge in demand for Samsung foundry services. Industry discussion cited in the item said DRAM supply-chain tightness would continue until 2027. In China’s display industry, the country’s first 8.6-generation AMOLED production line entered mass production in Chengdu with total investment of 63 billion yuan. TechFlow described it as one of the world’s first production launches for that generation of line and as a marker for China’s panel industry entering the high-generation OLED era.
Technology-company news continued the themes of AI cost and security. Apple CEO Tim Cook said RAM costs were “unsustainable” and that iPhone prices would rise. The report said AI functions were increasing memory demand, and that Apple had publicly stated it would pass costs on to consumers, a rare advance warning of price increases in recent years. In a Xiaomi-related AI rumor case, Xi’an police solved a criminal case involving false information generated by AI and arrested four people. TechFlow called it the first publicly reported criminal case in China involving AI-generated rumors. The related Zhihu discussion reached 580,000 views, with users debating the legal boundary of AI-generated misinformation. Ars Technica also reported a large-scale data leak exposing thousands of sensitive network credentials across a broad range of targets, including multiple critical-infrastructure networks.
NASA, SpaceX, Trump and energy-market geopolitics
In the U.S. stocks and finance section, NASA selected Eric Schmidt’s rocket company for a Mars mission, placing it in competition with SpaceX. TechFlow said the former Google CEO’s aerospace company had received a NASA order and that SpaceX’s monopoly position was facing a direct challenge for the first time. The same section noted that Sequoia partner Roelof Botha had joined SpaceX’s board. On politics and energy, Donald Trump defended a U.S.-Iran agreement by saying he did not want to see an economic disaster and did not want to be Hoover, who caused the 1929 crash. The United States and Iran reached a memorandum of understanding under which Iran agreed to dilute its enriched uranium stockpile in exchange for sanctions relief. The first shipment of Iranian oil had already broken through the U.S. blockade and moved to market. Republican Senator Bill Cassidy criticized the deal as “the worst diplomatic blunder in decades,” while market discussion centered on changes around the Strait of Hormuz and the global energy structure.
Geopolitical-risk items also included Ukraine carrying out a second airstrike on a Moscow refinery within one week, and Germany sending warships to the Red Sea. TechFlow said crude oil supply chains were again under pressure. Germany’s defense minister said two warships had been dispatched in response to military action related to the Strait of Hormuz. The column’s own sharp comment said Trump did not want to be Hoover, but markets had already started pricing him: from “Make America Great Again” to “Don’t let the economy crash again,” the speed of narrative downgrade at this White House was compared with DeepSeek’s chain of thought.
New products, digital inheritance and the day’s hidden line
The new products and trends section recorded three additional developments. Lore, an open-source version-control system designed for scalability, was released. It received 1,150 votes and 611 comments on Hacker News and was described as a potential challenger to Git. In China’s virtual-property field, a mother won court support in a lawsuit to inherit 87 game accounts from her deceased son, and the platform’s standard contract terms were ruled invalid. TechFlow called it China’s first successful publicly reported case involving inheritance of virtual property. The Zhihu discussion reached 8.64 million views as users debated digital inheritance rights. The Chinese indie game The Scroll of Taiwu released its full 1.0 version after eight years of development, while a Zhihu discussion with 1.28 million views focused on the completeness of the finished product.
TechFlow’s “hidden line of the day” connected Apple’s AI-driven price increase, OpenAI’s money-burning losses, the Copilot vulnerability and AMD’s removal of a security feature. The column argued that the cost of the AI boom was moving from capital markets to consumers and to the security baseline. At the same time, record on-chain stock trading and Trump’s compromise with Iran to avoid economic collapse pointed to the same anxiety: when AI infrastructure costs run out of control and geopolitical risks rise, who pays the bill for the bet? The original answer was consumers and taxpayers. At the end, TechFlow invited readers to join its official community and listed its Telegram subscription group at https://t.me/TechFlowDaily, its official Twitter account at https://x.com/TechFlowPost, and its English Twitter account at https://x.com/BlockFlow_News.