Analysis: Gold and Silver Undergo Sharp Corrections, Supply-Demand Imbalance Still Supports Precious Metals' Prospects

Hyperliquid’s TIA contract is currently trading at $0.75, up roughly 18% from the prior day. Against this market backdrop, two large holders with opposing positions have seen a stark divergence in profit and loss. One whale address, 0xe097…d0fe, holds approximately 1.1195 million TIA in long positions, using 5x full leverage, valued at around $820,000. It currently has an unrealized profit of about $289,000, a position return rate of roughly 272%, and remains unclosed. The other whale address, 0x07fd…689d, holds roughly 1.4295 million TIA in short positions, also with 5x full leverage, valued at approximately $1.07 million. It currently has an unrealized loss of about $342,000, with a position return rate of around -236%. This address added some 23,400 TIA to its short position at around $0.65 yesterday, and remains unclosed as of now.

Independent media outlets Court Watch and 404 Media have jointly reported that recently unsealed U.S. federal court documents reveal privacy-focused social messaging app Telegram provided specific user private data to U.S. authorities upon receiving valid legal requests. The documents include a September 2026 response from Telegram, which contains the user’s phone number, user ID, account display name, as well as the last login IP address and timestamp. Telegram described the disclosure as "voluntary", but stated it does not recognize U.S. judicial jurisdiction, and did not provide any chat content or contact information. Prior to this, Telegram adjusted its policy starting September 2024, allowing it to disclose IP addresses and phone numbers when judicial authorities confirm a user is suspected of criminal activity violating its terms of service. In all of 2024, the company responded to approximately 900 U.S. law enforcement requests involving 2,253 users.

Beating AI Express News: Gartner estimates that by 2030, AI data center owners will need to generate cumulative revenue of $8.5 trillion to cover capital expenditures, financing costs, and achieve a 12% return on invested capital (ROIC). However, revenue from AI businesses attributable to enterprises and consumers is projected to reach only $3.6 trillion, less than half of the target. The firm warns that AI infrastructure investment expansion may outpace revenue growth, and computing power assets could become obsolete before generating sufficient returns due to rapid depreciation, with over $3.5 trillion in AI infrastructure spending facing impairment risks. Separately, Global X research forecasts that the AI-related revenue of 15 AI platforms and service providers will exceed infrastructure depreciation around 2027, and reach approximately $1.5 trillion by 2030. The firm estimates that by 2030, every $1 in annual depreciation of AI-related assets will correspond to $2.36 in AI revenue. However, revenue covering depreciation does not mean it is sufficient to cover financing costs and meet the 12% ROIC requirement. Gartner also predicts that cost pressures and insufficient business differentiation will lead over 80% of new cloud service providers to cease independent operations by 2030. Meanwhile, CoreWeave’s CEO stated that the company’s order backlog is enough to meet its debt obligations. The firm’s Q2 backlog stood at $104.2 billion, with new customer commitments exceeding $25 billion in the third quarter.

Beating AI Insight News: NVIDIA CEO Jensen Huang recently told Bloomberg in a London interview that if customers develop insecure AI software, NVIDIA could hypothetically halt chip supplies to them, but will first communicate with clients before taking any action. Huang also emphasized that NVIDIA provides general-purpose technology, and has certain limitations in judging how customers use it. Meanwhile, NVIDIA pledged to provide $250 million worth of computing resources to the UK over the next five years to support AI safety research, focusing on agent security testing, with approximately $50 million allocated annually for this purpose. Supported projects include the Oxford-led BOLD initiative, as well as related research from Imperial College London and University College London. Huang advocated accelerating the integration of safety constraints and monitoring technologies into AI systems, noting that AI capabilities and safety must advance in tandem. The report notes that AI safety regulation could impact NVIDIA’s chip demand: if regulations restrict new model training or drive up development costs, AI processor purchases may be dampened; however, enhanced safety research will also help drive more reliable deployment of AI technology.

According to Bitget’s official data, the sixth batch of rewards for the platform’s "Peer Program" has been fully distributed, with total rewards amounting to 6,182,523 USDT and a total of 4,515,913 recipients. Additionally, after users visit the event page to claim their rewards, they will receive an extra chance to enter the 8th anniversary surprise blind box draw. The event will run until October 26.

Ethereum co-founder Vitalik Buterin published a post noting that roughly half of Bitcoin mining costs come from capital expenditure on mining hardware, with the other half attributed to electricity consumption. By contrast, AI inference costs are almost entirely dominated by computing hardware capital expenditure, a difference that reflects the distinct cost structures of the two computing businesses. Bitcoin miners must cover ongoing electricity bills, with profit margins determined by mining machine efficiency, Bitcoin prices, and changes in network difficulty; AI inference, meanwhile, relies far more on investments in expensive hardware such as GPUs, as well as device utilization rates and computing efficiency. The industrial boundaries between the two sectors are also shifting. As demand for AI computing power grows, some Bitcoin mining firms are repurposing existing power resources and data center facilities for AI and high-performance computing (HPC) services, aiming to transition from pure mining to computing infrastructure services. Reuters previously reported that some mining companies are leveraging power capacity originally built for mining to enter the AI data center market. However, this transition does not eliminate cost pressures: mining businesses are more directly exposed to fluctuations in electricity prices and Bitcoin prices, while AI computing operations face risks including GPU procurement, depreciation, and underutilization. Both sectors compete for power and computing resources, but their ultimate profitability depends on how effectively these resources are converted into revenue.