Bitcoin Price Falls Below $92,000

Beating AI News Flash: Huawei Cloud announced on November 1 that its two self-developed model services, openPangu-2.0-Flash and Pro, will be discontinued for sale on November 6 and fully retired on November 30. After full retirement, existing connected users will no longer be able to call the models. The two models were launched just a few months ago. The 92-billion-parameter Flash model was open-sourced on June 30, while the 505-billion-parameter Pro model was open-sourced on July 31. At the end of September this year, Huawei even open-sourced the pre-training, fine-tuning, and reinforcement learning codes for Pangu 2.0. Flash model users are advised to migrate to DeepSeek-V4.1-Flash-1017, while Pro model users are recommended to switch to Zhipu GLM-5.3. Huawei noted that a new generation of Pangu models is already in the planning stage, but has not yet disclosed specific models or launch timelines.

ZachXBT has updated his investigation into the suspected hack of a European institution, with over $27 million in abnormal withdrawals from centralized exchanges (CEXs) now tied to the case, bringing the total funds involved to more than $133 million. Some of these exchange withdrawals occurred after the incident was first disclosed. The investigator put forward an unconfirmed speculation that the case may involve a large Western European financial institution being attacked, with the situation possibly being covered up. He also stated that more than $30 million has been transferred and laundered due to inadequate incident disclosure, noting EU law enforcement agencies could have had the opportunity to freeze these funds.

Breaking news from Beating AI Insight: OpenAI has released a new evaluation framework called LifeVal, designed to test whether AI can assist ordinary people with daily life tasks. The assessment covers practical, step-by-step scenarios including finding kindergartens, grocery shopping, accessing government benefits, and caring for sick family members. To define the test content, OpenAI referenced survey data on daily life and time allocation from residents across multiple countries, and surveyed 3,995 people to identify the tasks they most want AI to handle. As a result, LifeVal spans areas including parenting, healthcare, shopping, finance, and education. The scores released by OpenAI show that GPT-5.2 scored an average of 49.3%, while GPT-6 Astra has improved to 80.6%. However, most progress came in the earlier stages: GPT-5.4 reached 73%, GPT-5.6 hit 80%, and the latest Astra only edged 0.6 percentage points higher. The company also showcased high-scoring cases: one earned a perfect 100 for helping a family find a home, with the AI locating properties meeting budget and commute requirements, calculating transfer taxes, and adding relevant documents to a cart for review. Another case, selecting a kindergarten after moving, scored 93.1 points, with the AI shortlisting three candidate schools and organizing enrollment information. LifeVal has previously been used to guide model improvements and remains an internal test designed and evaluated solely by OpenAI.

The World Gold Council’s Q3 2026 Global Gold Demand Trends report, released on November 3, shows that global physical gold ETFs returned to net inflows in the third quarter, totaling 211 tons, pushing gold ETF holdings to a record high of 4,256 tons. Over-the-counter and other gold demand stood at 143 tons, up from 101 tons in the same period last year. Additionally, global central banks net-purchased 321 tons of gold in Q3, marking the fastest pace of gold buying since the end of 2024. Year-to-date, central banks’ cumulative net gold purchases have reached 675 tons, slightly higher than the level in the same period last year.

动察 Beating AI Flash News: The Financial Times cites anonymous sources saying Nvidia is requiring global data center and cloud computing partners to sell at least 51% of their computing power services to non-Chinese clients. The requirement was communicated verbally to sales teams starting this summer, with no written documentation currently in place, and applies only to future business—existing signed leases remain unaffected. Previously, the U.S. restricted the export of advanced GPUs to China, but Chinese firms could still access high-end computing power via overseas cloud providers. For instance, chips are deployed in data centers in Singapore, Malaysia and other regions, with Chinese companies accessing them remotely via the internet, no need to ship GPUs back to China. This has been a key channel for some Chinese AI enterprises to obtain advanced computing power. Nvidia already tightened rules once in July this year, cutting the list of Asian customers approved to purchase AI chips by half, with a focus on reviewing cloud service providers that specialize in renting out GPUs. Staff even conduct on-site inspections of data centers, verify contracts and interview end users. Now the restrictions have expanded to the end buyers of computing power. Some Southeast Asian cloud providers are heavily dependent on Chinese clients, making it difficult to secure enough non-Chinese buyers in the short term. Some surveyed cloud providers worry they may have to scale back their data center operations if they fail to meet the requirement.

London-listed technology company The Smarter Web Company has announced a preferred stock dividend for its MORE shares, set at ￡0.2301 per share for the week ending November 6, 2026. Key dividend dates are: record date November 6, ex-dividend date November 5, and payment date November 23. The preferred stock currently carries an annual dividend yield of 12%, based on a ￡100 par value. As of the announcement date, 111,137 of the 211,112 issued shares are eligible to receive the dividend; the remaining shares are held by Tennyson Capital via an ATM facility and have waived dividend rights until they are sold. The company will provide a further update on the matter on November 5.