# Сопутствующие статьи по теме AI

Новостной центр HTX предлагает последние статьи и углубленный анализ по "AI", охватывающие рыночные тренды, новости проектов, развитие технологий и политику регулирования в криптоиндустрии.

After Institutional Support and Price Surge, Revisiting the True Value of Bittensor's 128 Subnets

After removing institutional support and price increases, this article re-evaluates the real value of Bittensor's 128 subnets. Bittensor operates as a decentralized AI ecosystem where each subnet functions like an independent startup with its own token (Alpha), revenue model, and team. There are two primary ways to earn: TAO emissions (protocol subsidies based on staking inflows) and Alpha token PnL (capital gains from subnet performance). Since the Taoflow update in November 2025, subnets with negative net staking flow receive zero emissions, creating a competitive environment. Approximately 3,600 TAO (around $960k daily) is distributed, with the top 10 subnets controlling 56% of emissions. Key case studies include Chutes (SN64), which demonstrates product-market fit with 400k users and 9.1 trillion tokens processed at 85% lower cost than AWS, and Templar (SN3), which offers asymmetric upside by training frontier LLMs in a fully decentralized manner. The investment framework positions TAO as an index fund for the entire network, while Alpha staking represents concentrated bets on specific subnets. The ecosystem is attracting institutional interest, with significant holdings from DCG and Polychain Capital. The conclusion emphasizes evaluating subnets based on product utility, staking flow, team execution, organic demand, and liquidity conditions.

marsbit03/17 13:32

After Institutional Support and Price Surge, Revisiting the True Value of Bittensor's 128 Subnets

marsbit03/17 13:32

Stargate Project Pivots: OpenAI Abandons Construction for Leasing, Wakes from 1.4 Trillion Compute Empire Dream

OpenAI has significantly restructured its ambitious Stargate project, shifting from a plan to build its own data centers to leasing computing power from cloud providers like Microsoft Azure, Oracle, and Amazon AWS. Initially announced in January 2025 with a projected investment of up to $1.4 trillion, the project's budget has been reduced by 57% to around $600 billion through 2030. The change comes after more than a year of delays, with no employees hired and no construction started, largely due to lenders' reluctance to finance the project given OpenAI’s ongoing substantial operating losses. The restructured Stargate is now organized into three teams overseeing commercial partnerships, technical engineering, and physical operations, all managed by former Intel executive Sachin Katti. Despite the shift in strategy, OpenAI still aims to deploy nearly 7 gigawatts of server capacity over three years, with investments totaling over $400 billion. The company is also developing its own AI chips in collaboration with Broadcom to reduce inference costs. Annual spending is projected to reach about $120 billion, matching Microsoft's yearly capital expenditure, even as OpenAI's revenue remains significantly lower and profitability is not expected before 2030.

marsbit03/17 13:08

Stargate Project Pivots: OpenAI Abandons Construction for Leasing, Wakes from 1.4 Trillion Compute Empire Dream

marsbit03/17 13:08

Huang Reinforces Nvidia’s Role, Pushing AI Cryptocurrencies upwards

Nvidia CEO Jensen Huang's keynote at CES 2026 emphasized the company's pivotal role in the expanding artificial intelligence industry, projecting a $1 trillion chip demand through 2027. This announcement boosted Nvidia's stock by 1.63% and positively influenced AI-related cryptocurrencies. Tokens such as FOMO, LISA, and AIN saw significant gains, rising over 30%, while major AI cryptocurrencies like TAO and NEAR also advanced. The broader crypto market rally, including Bitcoin and Ethereum, further supported the upward trend. Additionally, Nvidia's collaboration with Samsung on the Groq LP30 chip, expected to ship in late 2026, was highlighted.

TheNewsCrypto03/17 09:37

Huang Reinforces Nvidia’s Role, Pushing AI Cryptocurrencies upwards

TheNewsCrypto03/17 09:37

Cango Releases 2025 Financial Report: Advancing Towards AI Infrastructure

Cango Inc. (NYSE: CANG) released its unaudited financial results for Q4 and full year 2025, highlighting its transition into a Bitcoin mining company and its strategic pivot toward AI infrastructure. In 2025, the company reported total revenue of $688.1 million, with Bitcoin mining contributing $675.5 million. A total of 6,594.6 Bitcoin were mined throughout the year. However, the company reported a net loss from continuing operations of $452.8 million, influenced by one-time transition costs and fair value adjustments on Bitcoin-collateralized receivables. Adjusted EBITDA for the year was $24.5 million. In Q4, revenue was $179.5 million, with a net loss of $285 million and negative EBITDA of $156.3 million. The company ended the year with $41.2 million in cash and equivalents, $663 million in non-current Bitcoin receivables, and $557.6 million in related-party long-term debt. To reduce leverage, the company sold 4,451 Bitcoin in February 2026. CEO Paul Yu emphasized 2025 as a foundational year, noting the completion of structural adjustments and the establishment of a global mining network. The company is now advancing its transformation into an AI infrastructure provider through its EcoHash platform, aiming to offer flexible and cost-efficient AI inference services. CFO Michael Zhang highlighted efforts to optimize the balance sheet and secure new capital to support growth in high-potential areas like AI.

marsbit03/17 06:38

Cango Releases 2025 Financial Report: Advancing Towards AI Infrastructure

marsbit03/17 06:38

FMAC IPO Approved by SEC: Raises $100 Million, Targets AI and Web3 Leaders

Future Money Acquisition Corp (FMAC) has received approval from the U.S. SEC for its IPO, marking a significant milestone. The company plans to raise $100 million by offering 10 million units at $10.00 each and will list on Nasdaq under the ticker "FMAC." FMAC is a SPAC focused on acquiring leading companies in high-growth sectors including artificial intelligence (AI), Web3, and advanced manufacturing. The firm aims to identify globally oriented industry leaders with strong potential for growth and strategic acquisitions. Although the management has strong Asian investment experience, FMAC's acquisition strategy is global and not limited to mainland China. The leadership team includes Siyu Li as Chairman and CEO, founder of FutureMoney Group and advisor to Nasdaq-listed BTC Digital, and Steven Markscheid as CFO and Director, an experienced financial expert with a background in SPACs and cross-border mergers. With SEC approval secured, FMAC moves into its public offering phase, positioning itself as a bridge between traditional capital markets and the digital economy.

marsbit03/17 04:37

FMAC IPO Approved by SEC: Raises $100 Million, Targets AI and Web3 Leaders

marsbit03/17 04:37

Kalshi Issues a $1 Billion Free Lottery Ticket, Remember to Scratch It

Kalshi, a prediction market platform, has announced a "Perfect Bracket Challenge" inspired by Warren Buffett, offering a $1 billion grand prize to any user who correctly predicts every game outcome in the upcoming NCAA "March Madness" basketball tournament. The tournament begins on March 18, featuring 68 college teams competing in a single-elimination format. "March Madness" is one of the most popular sporting events in the U.S., drawing widespread public engagement due to strong school allegiances. This year's tournament is especially notable as a strong NBA draft class, including top prospects like Cameron Boozer, is raising its profile. Prediction markets like Polymarket and Kalshi have launched related events. Polymarket currently lists Duke University as the favorite to win at 21%. Meanwhile, Kalshi’s free-to-enter contest also includes a $1 million consolation prize for the best bracket and an additional $1 million donation to charity if no one wins the grand prize. The odds of a perfect bracket are extremely low—often estimated at around 1 in 120 billion—since games are not purely random and upsets are common. Warren Buffett has run a similar challenge for his employees since 2014, with no one ever claiming the top prize. The announcement has sparked discussion online, with some suggesting the use of AI and automated systems to attempt mass entries. Despite the near-impossible odds, users are encouraged to submit their brackets for free.

Odaily星球日报03/17 04:19

Kalshi Issues a $1 Billion Free Lottery Ticket, Remember to Scratch It

Odaily星球日报03/17 04:19

After 315, It's Time to Remove the Filter of AI Recommendations

After the 315 Gala exposed AI recommendation manipulation through Generative Engine Optimization (GEO), this article examines how GEO services allow brands to artificially boost their visibility in AI-generated responses. Similar to traditional SEO, GEO involves flooding the internet with promotional content that AI models scrape and present as authentic recommendations. The piece highlights a case where a fictional product with absurd features was promoted via a low-cost GEO tool and quickly appeared atop AI recommendations. It draws parallels with historical search engine manipulation, such as Baidu’s paid ranking scandals involving unregulated medical ads, and notes that GEO has already attracted significant capital and speculative stock market interest in China. The article also contrasts illicit “AI poisoning” with sanctioned commercialization efforts, citing OpenAI’s introduction of ads in ChatGPT. It concludes by warning users to maintain critical thinking and not blindly trust AI-generated recommendations, as the line between organic and paid content continues to blur.

比推03/17 04:11

After 315, It's Time to Remove the Filter of AI Recommendations

比推03/17 04:11

Stratechery Overturns the AI Bubble Theory: What Should We Use AI For?

Stratechery's Ben Thompson revises his stance on the AI bubble debate, arguing that current AI investment reflects structural growth driven by fundamental technological shifts, not speculation. He identifies three key transitions in LLM development: ChatGPT’s debut (making AI broadly usable but unreliable), OpenAI’s o1 (introducing reasoning and reliability), and the recent emergence of true Agent systems (e.g., Anthropic’s Opus 4.5 and OpenAI’s GPT-5.2-Codex). The critical innovation is the "agent harness"—a control layer that autonomously schedules models, uses tools, and verifies outcomes, reducing human intervention. This transforms AI from an assistive tool into an actionable infrastructure, enabling complex task execution. Agentic systems drive surging demand for compute, as they require repeated model calls and higher usage frequency. Thompson emphasizes that AI demand now depends not on user numbers, but on per-user agent utilization. Enterprises are adopting AI not only for efficiency gains but for structural workforce reduction, as agents amplify the impact of key employees while lowering coordination costs. He concludes that massive capital expenditures are justified by real demand, and profits will flow to integrated model-harness providers rather than commoditized standalone models.

marsbit03/17 03:57

Stratechery Overturns the AI Bubble Theory: What Should We Use AI For?

marsbit03/17 03:57

Intelligent Computing Convergence: The Deep Integration Architecture, Paradigm Evolution, and Application Landscape of AI and Cryptocurrency Industries

The deep integration of AI and cryptocurrency represents a fundamental paradigm shift, moving beyond mere technological convergence to reshape economic and computational infrastructures. By 2025, the crypto market cap surpassed $4 trillion, signaling its maturation, while AI evolved from centralized models toward decentralized, transparent “open intelligence.” Key architectural innovations include decentralized physical infrastructure networks (DePINs) like Render and Akash, which aggregate global idle GPU resources, and platforms like Ritual that embed AI models into blockchain execution environments. Verification mechanisms such as ZKML and TEE ensure computational integrity and privacy. Bittensor introduces a token-incentivized marketplace for machine intelligence, using its Yuma consensus to reward high-performing models dynamically. AI agents have transitioned from tools to autonomous on-chain entities, capable of managing finances and executing DeFi strategies via protocols like x402 and Olas. Privacy advancements through FHE (e.g., Zama), ZKML, and TEE enable confidential on-chain computations, critical for high-stakes applications. AI also enhances security via automated smart contract auditing and real-time threat prevention systems. This fusion drives enterprise efficiency through cost reduction and secure data processing, while empowering individuals via intent-based agents and data monetization. The future points to “intelligent ledgers” where AI and blockchains are deeply architecturally coupled, enabling a fairer, decentralized digital economy.

marsbit03/17 03:13

Intelligent Computing Convergence: The Deep Integration Architecture, Paradigm Evolution, and Application Landscape of AI and Cryptocurrency Industries