News

According to Foresight News, the Pentagon is set to begin utilizing xAI's Grok system later this month, as reported by DogeDesigner.

According to Foresight News, artificial intelligence company Anthropic has launched Cowork under its Claude platform. Cowork is designed to simplify task management for Claude users by eliminating the need for repetitive manual context provision and output formatting. Users can submit tasks in a queue, allowing Claude to handle them concurrently without waiting for the completion of current tasks before introducing new ideas or feedback. Cowork is now available as a research preview on the macOS app, accessible to Claude Max subscribers.

According to BlockBeats, on January 13, BlackRock released its 2026 global outlook report, emphasizing the significant scale of AI infrastructure investments. The report suggests that 'micro is macro,' indicating that these investments are substantial enough to influence the broader macroeconomic landscape, presenting challenges such as increased leverage and the illusion of diversification. BlackRock maintains a pro-risk stance, favoring U.S. equities, particularly those related to AI, and sees opportunities in active investments. The report outlines three core investment themes: Micro is macro: AI infrastructure is dominated by a few companies, with capital expenditures projected to reach $5-8 trillion between 2025 and 2030. This investment is expected to support U.S. economic growth in 2026, contributing three times the historical average, despite a cooling labor market. However, there is uncertainty about whether revenues will match expenditures and how much will return to tech giants. While AI may accelerate innovation, historical trends over the past 150 years suggest that major technological changes have not disrupted the U.S.'s long-term 2% growth trend. Nonetheless, a 'growth breakout' scenario is now conceivable. Leveraging up: The initial massive investments by AI builders, coupled with delayed revenues, lead to increased system leverage. High government debt adds to this vulnerability. BlackRock favors private credit and infrastructure financing while tactically underweighting long-term government bonds, such as U.S. Treasuries, due to the adverse effects of high leverage and rising capital costs on long-term debt. Diversification mirage: Under major trends, traditional diversification strategies may actually concentrate bets. Investors need to actively manage risk, maintain portfolio flexibility with a 'Plan B,' and seek unique returns from private markets and hedge funds. The report also highlights that BlackRock views digital assets, particularly stablecoins, as foundational to payment and settlement systems, rather than merely speculative assets. Stablecoins are seen as 'digital dollar rails,' evolving from crypto-native tools to bridges connecting traditional finance with digital liquidity, expanding into areas like cross-border payments and settlements, especially in regions where traditional systems are slow, costly, or fragmented. The report suggests that crypto is integrating into mainstream finance, with stablecoins maturing as infrastructure supporting global liquidity flows and overlapping with traditional finance.

According to PANews, OpenAI and SoftBank Group have jointly invested $1 billion in SB Energy, an infrastructure company collaborating with these tech giants to construct large-scale data centers in the United States to support artificial intelligence. In a statement released on Friday, OpenAI and SoftBank announced that each company will invest $500 million in SB Energy to aid its development as a data center developer and operator. OpenAI has also selected SB Energy to build and operate its 1.2-gigawatt data center in Milam County, Texas. A gigawatt of power can supply electricity to approximately 750,000 American households simultaneously. This collaboration with SB Energy is part of OpenAI's "Stargate Initiative."

According to PANews, Australian eSafety Commissioner Julie Inman Grant has expressed concerns over the increasing number of complaints related to involuntary gendered images generated by Elon Musk's AI model, Grok. Reports indicate that these complaints have doubled since late 2025, with some involving child pornography. Grok's 'Spicy Mode' feature has been criticized for facilitating AI deepfake misuse and has been deemed illegal by the European Union. Grant emphasized that under Australian regulations, platforms must enhance oversight of AI-generated content. The eSafety office plans to investigate and take legal action as necessary.

According to PANews, Elon Musk has announced that his AI model, Grok, will receive a significant upgrade next month. The update will enhance Grok Code's ability to generate multiple complex programming tasks simultaneously. Previously, Grok Code Fast 1 topped the OpenRouter AI model rankings, with usage reaching approximately 420 billion tokens this week.

According to PANews, OpenAI has announced the release of its 'OpenAI for Healthcare' product suite, designed to help medical institutions deploy AI in compliance with HIPAA requirements to improve clinical and administrative efficiency. This suite is already operational in healthcare systems such as AdventHealth, UCSF, Cedars-Sinai, and HCA. Based on the GPT-5.2 model, the product offers features like clinical documentation generation, evidence citation, institutional pathway integration, and role-based access management, without being used for model training. The initiative aims to enable AI to support large-scale, high-quality healthcare services. Yesterday, OpenAI also introduced 'ChatGPT Health,' which facilitates the connection between medical records and health applications.

According to PANews, OpenAI has unveiled a new health-focused feature called 'ChatGPT Health.' This service allows users to voluntarily connect their medical records with health applications like Apple Health and MyFitnessPal to receive personalized advice based on their health data. The feature aims to assist users in understanding test results, preparing for medical appointments, and optimizing diet and exercise plans. It ensures privacy by storing health data in a separate space. A waitlist is now open, with initial access limited to a small group of users.

According to PANews, artificial intelligence startup Anthropic is planning to raise $10 billion in its latest funding round, with the company's valuation reaching $350 billion prior to securing new investments. Sources indicate that Singapore's sovereign wealth fund GIC and Coatue Management are expected to lead this round of financing. The funding is anticipated to be finalized in the coming weeks, although the total transaction amount may vary. In a previous round last September, Anthropic raised $13 billion from investors, tripling its valuation to $183 billion.

According to Foresight News, Jack Clark, co-founder of Anthropic and former policy director at OpenAI, highlighted the significance of decentralized training in his weekly AI newsletter, Import AI. He emphasized that decentralized training enhances data privacy and system robustness by utilizing distributed learning across multiple nodes. Citing a study by Epoch AI, Clark noted that the computational scale of decentralized training is expanding at an annual rate of 20 times, significantly outpacing the fivefold growth rate of cutting-edge centralized training. Although decentralized training is currently about 1,000 times smaller than advanced centralized methods, it remains technically feasible and could potentially facilitate broader collaborative efforts to develop more powerful models.

According to Foresight News, sources indicate that Baidu's artificial intelligence chip company, ARM, is planning an initial public offering (IPO) in Hong Kong. The company aims to raise up to $2 billion through this IPO.

According to Odaily, xAI has successfully completed its Series E funding round, raising $20 billion, exceeding the initial target of $15 billion. The investment round included contributions from Valor Equity Partners, StepStone Group, Fidelity Management & Research, Qatar Investment Authority, MGX, and Baron Capital Group, with NVIDIA and Cisco participating as strategic investors.xAI announced that by 2025, the company aims to achieve significant advancements in computing power and product development. This includes the construction of the Colossus I and II supercomputing centers, with GPU capacity exceeding the equivalent of one million H100 units. The company is also advancing the training of the Grok 4 series models and plans to launch products such as Grok Voice and Grok Imagine, which will cater to applications in voice, image, and video generation. Currently, xAI's product suite covers approximately 600 million monthly active users across its X and Grok applications.The funds from this round will primarily be used to accelerate the development of computing infrastructure, support the training of the next-generation Grok 5 models, and continue the development of consumer and enterprise-level products. xAI also plans to expand its team to support ongoing research and product implementation.

According to BlockBeats, the AI project Giza, supported by the Ethereum Foundation, has seen its AI Agent's asset management scale surpass $40 million, marking an increase of over 60% in the past month.Giza focuses on developing infrastructure for verifiable AI Agents within the DeFi sector. These AI Agents autonomously conduct market analysis, optimize assets, and execute trading strategies without human intervention, ensuring transparency and security through zero-knowledge proofs.

According to Foresight News, Nvidia introduced its new automotive platform, Alpamayo, at the CES 2026 event. CEO Jensen Huang stated that the platform enables vehicles to perform 'inference' in real-world scenarios. Potential users can adopt the Alpamayo model and retrain it independently. This free service aims to develop vehicles capable of handling unexpected situations, such as traffic light malfunctions. The onboard computer analyzes inputs from cameras and other sensors, breaking them down into steps to derive solutions.Additionally, Nvidia's new Rubin data center product is set to launch this year, allowing customers to test the technology. Six Rubin chips have been delivered from manufacturing partners and have passed several key tests, indicating progress toward customer deployment. The latest Rubin accelerator offers a 3.5-fold improvement in training performance and a fivefold increase in AI software speed compared to its predecessor, Blackwell. Systems based on Rubin are expected to have lower operational costs than those based on Blackwell.

According to Cointelegraph, Nvidia CEO Jensen Huang emphasized the escalating demand for computing resources driven by the rapid evolution of artificial intelligence (AI) models. Speaking at a live event in Las Vegas, Huang described the situation as an "intense race" towards the next technological frontier. He highlighted the fierce competition within the AI sector, noting that the computational requirements for AI are increasing exponentially. Huang stated that the demand for Nvidia GPUs is soaring as AI models grow by an order of magnitude each year. He remarked, "Everybody's trying to get to the next level and somebody is getting to the next level. And so therefore, all of it is a computing problem. The faster you compute, the sooner you can get to the next level of the next frontier." The rapid growth and adoption of AI have prompted several Bitcoin mining companies to shift their focus towards AI, either partially or entirely, over the past few years. This shift is partly due to the increasing difficulty of Bitcoin mining, while AI offers miners an opportunity to optimize their resources and potentially increase their revenues beyond Bitcoin. The rising demand for AI computing power could make this transition even more appealing for Bitcoin miners. During his address, Huang also discussed Nvidia's next-generation Rubin Vera chips, which are currently in full production and on schedule. He explained that the combination of Rubin and Vera chips, designed to work in tandem, will deliver five times greater AI computing performance compared to previous models. This advancement underscores Nvidia's commitment to staying at the forefront of AI technology and meeting the growing computational demands of the industry.

Binance Blog published a new article, introducing Binance's groundbreaking vision-language model designed to generate fact-checked narratives for meme tokens. This innovative tool creates backstories from names, symbols, logos, and social media posts in under two seconds across multiple blockchain platforms. Each narrative undergoes rigorous evaluation for factuality, fluency, style, and relevance, ensuring high-quality and trustworthy outputs. Integrated into the Meme Rush feature, these narratives provide users with instant insights into token origins and motivations, significantly reducing research friction.The rapid creation of meme tokens has led to an information gap, with new tokens appearing frequently without context about their origins or creators. This leaves traders guessing, trying to decipher a constant influx of new coins with minimal information. Binance Wallet's Meme Rush addresses this issue by offering AI-generated narratives for each token, allowing users to understand the "why" behind each meme token instantly. The system continuously scans both on-chain and off-chain sources to gather essential details for new meme tokens from over ten launch platforms. This automated process enables Binance to index new tokens swiftly.The AI narrative system, built using a compact vision-language model, generates concise and easy-to-read narratives. To maintain speed during peak traffic, these narratives are cached in a high-performance database for quick access. Users can view the narrative instantly by hovering over the narrative icon on a token profile, providing quick context without the need for additional searches. This feature not only enhances user engagement but also positions Binance as a pioneer in integrating AI narratives with Web3 discovery, potentially boosting trust and community energy in the memecoin market.To ensure the narratives capture not just what a token is but why it exists, Binance has fine-tuned its vision-language model to read both text and logos, trained it across multi-chain meme ecosystems, and built an efficient data pipeline to track emerging token trends. The AI system is chain-agnostic, offering instant context for newly launched meme tokens without users leaving Binance Wallet. It understands different naming conventions, memes, and cultural trends, ensuring consistent and clear stories regardless of the blockchain a token originates from.Looking ahead, Binance plans to enhance the Meme Rush narrative system by continuously updating the model to capture the latest memes, token launches, and evolving language trends. The goal is to make the AI not just reactive but proactively insightful, connecting trends, culture, and markets in real-time across the Web3 ecosystem. As AI-driven insights become increasingly crucial in the crypto space, Binance is at the forefront, providing users with tools to navigate the complex world of meme tokens effectively. However, users are reminded to conduct their own research and consider their risk appetite before making investment decisions.

According to ChainCatcher, Israeli financial daily Calcalist reported on Tuesday that Nvidia (NVDA.O) is in advanced negotiations to acquire Israeli AI startup AI21 Labs for a sum between $2 billion and $3 billion. A 2023 funding round valued AI21 Labs at $1.4 billion, with Nvidia and Google's parent company Alphabet (GOOG.O) participating in the investment.Founded in 2017 by Amnon Shashua and two others, AI21 Labs is among the startups benefiting from the surge in interest in artificial intelligence. Shashua is also the founder and CEO of Mobileye, a developer of autonomous vehicle technology. AI21 Labs has been seeking a sale for some time, and discussions with Nvidia have made significant progress in recent weeks.Nvidia's primary interest in AI21 Labs appears to be its team of approximately 200 employees, most of whom hold advanced degrees and possess rare expertise in AI development. The acquisition price translates to roughly $10 million to $15 million per employee.

According to PANews, SoftBank has fulfilled its $40 billion investment commitment to OpenAI, as reported by CNBC citing unnamed sources. This significant financial move underscores SoftBank's confidence in OpenAI's potential and its strategic importance in the technology sector.