MarsBit

BlockBeats

2026-09-26 08:04:09

BlockBeats apologizes for misreport on fomo App, says no paid smear campaign was involved

BlockBeats has issued a public explanation over an erroneous security report published on Sept. 23, 2026, at 19:38, when it pushed a story titled "fomo official iOS app contains crypto-stealing malware." The outlet said the app in question was actually called FomoPeek, not the widely used trading app fomo, and that its wording and push notification caused many readers to believe fomo users faced a theft risk. According to the statement, the mistake quickly spread across social media and triggered speculation that the article may have been a paid attack from a competitor, especially as overseas KOLs were at the same time criticizing trading losses on fomo. BlockBeats said it conducted an internal review and found no evidence of any paid smear arrangement. Instead, it concluded that the duty editor failed to verify the product name before republishing a security alert that had already been covered on Sept. 19. The outlet said it removed the incorrect article, updated the malware name to FomoPeek, stopped distribution of the wrong version across its homepage, search, and other channels, and disciplined the editor involved. It also said future reports involving security issues and specific product names will require a second round of verification before publication.

260

MarsBit

2026-09-19 01:33:10

MarsBit weekly picks spotlight Fed rate hike, tokenized stocks and Ethereum gas changes

MarsBit’s Weekly Editor’s Picks for Sept. 12 to Sept. 18 rounds up a broad set of crypto and macro stories that the outlet sees as worth deeper attention. The selection starts with the U.S. Federal Reserve’s unanimous decision on Sept. 16 to raise its benchmark rate by 25 basis points to a 3.75% to 4.00% range, alongside a dot plot that pointed to at least one more hike this year. MarsBit said the hawkish signal pressured U.S. stocks and gold, lifted the dollar to a one-month high, and sharply bear-flattened the yield curve. The roundup also highlights X’s Cashtag Partner Program in the U.S., which adds a Trade entry for supported stock, ETF and crypto cashtags and connects users to brokers including Interactive Brokers, Moomoo, Gemini, Kraken and Coinbase. In policy, MarsBit says the U.S. Securities and Exchange Commission used an “innovation exemption” to open a path for tokenized stock trading after the CLARITY bill failed in the Senate. Other featured topics include the delisting path of U.K.-listed BTC treasury company Satsuma, the implications of EIP-8141 Frame Transactions for paying Ethereum gas without holding ETH in a wallet, ARC’s weak meme launch, and a security report on a North Korean hiring tactic involving a $500 paid interview and in-person job substitution.

590

Claude

2026-09-15 07:35:10

Claude cracks a 370-year-old cipher in 44 minutes and then solves a second Urquhart puzzle

Claude Fable 5.1 has been reported to solve a 370-year-old cipher left by 17th-century English scholar Sir Thomas Urquhart, finishing the task in 44 minutes while using 176,000 tokens and without human intervention during the process. The first puzzle, known as the Cyphral Distich and published at the end of Urquhart’s 1653 work Logopandecteision, consists of two lines of 32 numbers each and had resisted attempts by scholars and hobbyists for centuries. According to the report, Claude identified that the key was not an external codebook but the book itself: each number pointed to a numbered petition and then to a word position, with the plaintext recovered by taking first letters. The output read, “O GOD UPHOLD KING CHARLS THE SECOND AND” and “MAKE HIM THE SUPREME RULER OF THIS LAND.” After that, the model applied the same book-as-index logic to Urquhart’s Cyphral Octastich in The Jewel, a longer puzzle built from 285 numbers, and used EEBO-TCP plus self-written Python scripts for layout calibration to reconstruct a full iambic eight-line poem. The report also linked this result to broader AI research work disclosed by Anthropic and comments from Vals AI on rapid iteration across archival tasks.

310

Engels Pause

2026-09-14 14:44:09

A long read reframes the Engels Pause as “social sarcopenic obesity” in the AI era

MarsBit published a long-form article by Lü Peng of Tencent Research Institute arguing that current debate around a possible new “Engels Pause” in the age of artificial intelligence is too narrowly framed as a distribution problem. In the author’s view, the deeper issue is not simply that a larger economic pie is being divided unevenly, but that the economy is growing “fatter” while society is shrinking. He describes that condition as “social sarcopenic obesity,” a state in which GDP, profits and capital stock expand while social organization, collective capacity and self-renewing dynamism weaken. The article revisits the original Engels Pause during Britain’s early Industrial Revolution, citing figures that show output per worker rose about 46% between 1780 and 1840 while real wages increased only 12%, even as profit rates and profit shares climbed sharply. It then argues that the old pause ended through a combination of capital accumulation and social institution-building, but that the modern version may not heal in the same way because today’s gains increasingly take the form of rent rather than productive profit, and because AI-era capital deepening can substitute for labor instead of complementing it. Lü extends the analysis to China, saying the country’s policy framework, including the “two synchronizations” and the newer emphasis on “investing in people,” gives it a different starting point from the West. Even so, he warns against relying on redistribution alone and calls for a broader agenda centered on rebuilding society’s “muscle” through organization, participation, labor bargaining capacity and local social institutions.

370

MarsBit

2026-09-10 04:44:09

MarsBit piece argues modern IT is still stacked on a fragile von Neumann foundation

MarsBit has published a long-form commentary that uses a single diagram to dissect the structure of modern IT infrastructure, arguing that the industry keeps adding new layers while leaving its oldest bottlenecks largely untouched. The article centers on the von Neumann bottleneck, describing it as the thin support column under today’s computing stack: CPUs keep getting faster, but memory and bus bandwidth remain limiting factors, and growing software complexity only increases data movement between memory and processors. The commentary walks from electricity and semiconductor fabs up through operating systems, high-bandwidth memory, CPUs and GPUs, C projects, JVM-based software, web tooling, and end-user applications. Its core claim is that each layer may look well engineered on its own, yet the full stack is less a unified design than an accumulation of temporary fixes, compatibility compromises, and maintenance patches. It also frames the current AI boom as a source of additional strain rather than a fix for the base layer. According to the article, AI training and inference sharply increase power demand, GPU usage, and data retrieval pressure, while AI-generated code and content can introduce quality and security problems into open-source and web ecosystems. The author extends that criticism to large tech companies, which are portrayed as repeatedly disrupting existing frameworks and standards, and to independent developers, whose fragmented projects can add flexibility but also disorder and maintenance risk.

860

AI

2026-09-06 02:46:14

System thinking, not knowledge, is becoming the scarce skill in the AI era

MarsBit has published a commentary by GritMeng arguing that the core bottleneck in the AI era is no longer access to knowledge, but the ability to think in systems. The article says large models are rapidly eroding the scarcity of professional skills such as coding, design, copywriting, and data analysis, while many companies are discovering that broader decision-making can still deteriorate even as every team becomes more productive with AI tools. According to the piece, the main failure comes from using AI as a tool for isolated optimization. Departments may improve procurement costs, inventory turnover, click-through rates, or other KPIs, yet still push the whole organization into weaker margins, damaged user experience, supply-chain fragility, or reputational stress. GritMeng describes this as a form of "Goodhart collapse," where a metric stops being useful once it is optimized too aggressively. The article breaks system thinking into three core capabilities: defining the right boundaries before optimization begins, seeing delayed feedback loops before they become destructive, and revising one’s own mental model when reality diverges from theory. It also argues that AI cannot generate this form of thinking on its own because system thinking depends on a broader human mental architecture, not just computational power. The piece closes by urging readers to move from being high-level executors to system architects who can set boundaries, build feedback loops, and pay attention to residual errors rather than just efficiency gains.

290

MarsBit

2026-09-05 06:24:14

MarsBit Weekly Editor’s Picks: Fed signals, Robinhood Chain trades and a tighter ETH float

MarsBit’s Weekly Editor’s Picks for Aug. 29 to Sept. 4 rounds up a broad set of stories spanning macro policy, crypto market structure, Robinhood Chain, Ethereum, Hyperliquid and the week’s most discussed headlines. At the macro level, the package focuses on Federal Reserve Chair Kevin Warsh’s first Jackson Hole appearance, where he argued that forward guidance had served its purpose in extraordinary times and that policy should return to data dependence and institutional discipline. The roundup also highlights renewed stagflation pricing as oil moved back toward $90, pressure from Treasury supply and debt issuance on long-end yields, and a sharp rise in concern over an unwind of yen carry trades after Japan’s 10-year government bond yield moved above 3%. On the crypto side, MarsBit places heavy emphasis on Robinhood Chain. The selection tracks which tokens may be capturing growth from the chain’s activity, the mechanics and limits of stock-paired meme coins, emerging AMMs for stock-token LPs, and the surge of PONS after strong issuance and trading growth. It also points to a temporary easing of a key market worry around Strategy and STRC, a tighter Ethereum supply picture driven by spot ETF inflows and staking, and the opening phase of HIP-4 competition in Hyperliquid’s prediction-market ecosystem. The roundup ends with a rapid-fire list of policy, company, market and security developments from the week.

1300

AI betting

2026-09-04 10:57:10

How AI Betting Disputes Turned Into Public Trials for Creators

A MarsBit article examines the rise of so-called "AI betting" disputes in online creator communities, using illustrator Xiaolin’s six-hour livestream as its starting point. In that case, Xiaolin was accused on social media of using AI to generate an artwork and was challenged to repaint it live. The wager was 4,100 yuan: if he succeeded, the accuser would pay; if he failed, he would have to compensate, apologize, and absorb the damage to his reputation. The report says this format has spread across platforms such as Xiaohongshu, where similar posts have become common. A recurring structure has emerged: the accused creator, the person raising the accusation, and a middleman holding the stake. According to the article, what began as a form of self-verification has increasingly become a public spectacle in which thousands of viewers scrutinize every pause, brushstroke, or wording choice as possible proof of AI use. MarsBit’s piece argues that the problem goes beyond online hostility. It says current AI detection logic is fundamentally unstable because neither an image nor a text can reliably prove whether it was made by a human or generated by AI. The author frames the trend as a mix of group psychology, social identity, and a broader trust crisis over what originality means in the AI era.

430