# Сопутствующие статьи по теме LLM

Новостной центр HTX предлагает последние статьи и углубленный анализ по "LLM", 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

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

KuCoin has Launched Skills Hub, Advancement for AI Agent

KuCoin has launched Skills Hub, an open skills marketplace designed to broaden access to AI agents and LLM tools. The platform enables these agents to interact directly with the exchange’s data via standardized APIs, aiming to assist retail users, institutions, developers, and quantitative teams through intuitive natural language interactions. Built on principles of trust, transparency, and execution, the hub seeks to simplify complex crypto workflows by converting them into structured, deployable skills. KuCoin CEO BC Wong described the launch as a significant step toward a more practical and AI-enabled crypto ecosystem. Future plans include strengthening open infrastructure and expanding the library of skills, data sources, and protocol integrations.

TheNewsCrypto03/16 09:28

KuCoin has Launched Skills Hub, Advancement for AI Agent

TheNewsCrypto03/16 09:28

Firecrawl Launches Web Scraping Tool for Agents, NVIDIA Releases Nemotron 3 Super: What's the English Community Discussing Today?

Over the past 24 hours, key discussions in the English-speaking crypto and AI communities centered on several major developments. Firecrawl launched a CLI toolchain specifically for AI agents, enabling efficient web scraping and data extraction, though its pricing drew some criticism. Nvidia released Nemotron 3 Super, a 120B-parameter open-weight model with a 1M-token context window, raising both excitement and concerns over latency and safety. Google introduced Nano Banana 2, a high-speed image generation model, though its naming was met with mixed reactions. In AI agent infrastructure, Base44’s Superagent entered the cloud-based agent automation space, intensifying competition with local solutions like OpenClaw and raising debates over security and centralization. Ramp’s AI Index suggested Anthropic is gaining traction as the preferred enterprise AI vendor over OpenAI. In crypto, Solana continued to strengthen its infrastructure with DoubleZero Edge’s real-time market data via multicast technology and led in stablecoin transfer volume after filtering wash trading. Jupiter launched its Season 2 rewards program with a $2M JupUSD pool. Ethereum saw progress in L2 interoperability with on.eth addressing cross-chain identity fragmentation. Base ecosystem projects like Noise.xyz and rip.fun attracted significant attention, while Circle experimented with AI agents autonomously managing a hackathon using USDC. Perp DEX Lighter introduced a revised market structure to improve fairness, and prediction market platform Kalshi noted growing institutional engagement, with Marco Rubio emerging as an early favorite for the 2028 U.S. presidential election. Overall, themes included the shift toward cloud-based AI agents, model capability races, enterprise AI adoption, and the maturation of on-chain trading and prediction markets.

marsbit03/12 15:41

Firecrawl Launches Web Scraping Tool for Agents, NVIDIA Releases Nemotron 3 Super: What's the English Community Discussing Today?

marsbit03/12 15:41

AI Jargon Dictionary (March 2026 Edition), Recommended to Bookmark

AI Jargon Dictionary (March 2026 Edition): A Practical Guide This article provides a clear glossary of essential AI terminology, perfect for anyone looking to quickly understand the field. It breaks down key concepts into two categories: foundational and advanced terms. Foundational terms (12) include: - **LLM (Large Language Model)**: Deep learning models trained on vast data to understand and generate language. - **AI Agent**: Systems that understand goals, use tools, and execute multi-step tasks. - **Multimodal**: Models that process and generate multiple content types (text, image, audio, video). - **Generative AI (AIGC)**: AI focused on creating new content. - **Token**: The basic unit for processing input/output, affecting cost and performance. - **Context Window**: The amount of data a model can consider at once. - **RAG (Retrieval-Augmented Generation)**: Enhances responses using external, up-to-date data sources. Advanced terms (18) cover: - **Transformer**: The architecture behind modern LLMs. - **Hallucination**: When models generate incorrect but confident responses. - **Agentic Workflow**: Systems that autonomously break down and execute tasks. - **Vibe Coding**: Generating code through natural language commands without manual coding. The guide is designed to help readers confidently navigate AI discussions and implementations.

marsbit03/11 11:52

AI Jargon Dictionary (March 2026 Edition), Recommended to Bookmark

marsbit03/11 11:52

AI Jargon Dictionary (March 2026 Edition), Recommended to Bookmark

The "AI Jargon Dictionary (March 2026 Edition)" is a practical guide for those new to the AI field, especially crypto enthusiasts looking to stay relevant. It covers essential and advanced AI terms to help readers understand key concepts and avoid confusion in industry discussions. The dictionary is divided into two parts: **Basic Vocabulary (12 terms):** - Core concepts like LLM (Large Language Model), AI Agent (intelligent systems that execute tasks), Multimodal (handling multiple data types), and Prompt (user instructions). - Key technical terms: Token (processing unit), Context Window (token capacity), Memory (retaining user data), Training vs. Inference (learning vs. execution), and Tool Use (calling external tools). - Generative AI (AIGC) and API (integration interface) are also explained. **Advanced Vocabulary (18 terms):** - Technical foundations: Transformer architecture, Attention mechanism, and Parameters (model scale). - Emerging trends: Agentic Workflow (autonomous systems), Subagents, Skills (reusable modules), and Vibe Coding (AI-assisted programming). - Challenges: Hallucination (incorrect outputs), Latency (response time), Guardrails (safety controls). - Optimization techniques: Fine-tuning, Distillation (model compression), RAG (Retrieval-Augmented Generation), Grounding (fact-based responses), Embedding (vector encoding), and Benchmark (performance evaluation). The article emphasizes practicality, urging readers to learn these terms to navigate AI conversations confidently. It highlights terms like RAG and Grounding as critical for enterprise AI, while newer buzzwords like MCP (Model Context Protocol) and Vibe Coding reflect evolving trends. The goal is to provide a concise yet comprehensive reference for understanding AI jargon in 2026.

Odaily星球日报03/11 11:36

AI Jargon Dictionary (March 2026 Edition), Recommended to Bookmark

Odaily星球日报03/11 11:36

Lobster Key 11 Questions: The Most Easy-to-Understand Breakdown of OpenClaw Principles

"OpenClaw Demystified: A Beginner's Guide to AI Agent Principles" explains the popular OpenClaw AI assistant by breaking down its core functions into 11 key questions. The article first clarifies that the underlying large language model is merely a "text prediction engine" with no real understanding, memory, or senses. OpenClaw acts as a "shell" around this model, creating the illusion of memory by appending massive prompts containing its personality files (AGENTS.md, SOUL.md, USER.md) and the entire conversation history before each interaction. This mechanism is why it's "expensive"—each query processes thousands of tokens of context, not just the latest message. A core differentiator is tool use. The model itself only outputs text; OpenClaw parses this output for specific structured commands (e.g., `[Tool Call] Read("file.txt")`) and executes the corresponding action (reading the file) locally on the user's machine. This allows it to act, not just advise. For complex tasks, it can even write and run its own Python scripts, a powerful but dangerous capability. To manage limited context windows and complex tasks, OpenClaw uses sub-agents. A main agent can spawn sub-agent to handle a sub-task and return a summarized result, preventing the main context from being overloaded. Crucially, sub-agents cannot spawn their own to avoid infinite loops. Unlike standard chatbots, OpenClaw is proactive due to its heartbeat mechanism, which periodically prompts the model to check for tasks. It can also "sleep" via cron jobs to wait for long-running tasks, saving resources. The guide ends with critical security warnings. OpenClaw has extensive local access, making it a significant risk. It can malfunction (e.g., deleting emails uncontrollably) or fall victim to prompt injection attacks, where malicious input from the web is mistaken for a user's command. The strong recommendation is to run it on a dedicated, isolated "sacrificial" computer with minimal permissions and mandatory human confirmations for destructive actions.

Odaily星球日报03/11 09:53

Lobster Key 11 Questions: The Most Easy-to-Understand Breakdown of OpenClaw Principles

Odaily星球日报03/11 09:53

After the Lobster Comes Ashore, the Next Game in AI Hardware Lego

The article "Lobster Comes Ashore: The Next Game in AI Hardware Lego" discusses the growing influence of OpenClaw, an open-source AI framework, as it extends from software into the physical hardware world, reshaping the development and functionality of smart devices. OpenClaw enables hardware products to be combined like Lego blocks, creating diverse intelligent devices. Examples include Rokid AI glasses, which can now connect to any backend system like OpenClaw via an SSE interface, and Apple Watch, which acts as an AI control terminal for tasks like managing notifications and sending commands. WHOOP wearable devices use OpenClaw to provide personalized health advice, while companies like Songling Robotics integrate it into robotic arms for natural language control. Individual developers are also experimenting, such as combining OpenClaw with Meta’s Ray-Ban smart glasses for visual AI agents, or enhancing robot dogs like Vbot for autonomous tasks. These innovations are expanding possibilities but also raise concerns around security and token costs. The trend is particularly strong in China, where OpenClaw has sparked enthusiasm among companies, developers, and policymakers. In Shenzhen, public installations and events around OpenClaw have drawn large crowds, and electronics market Huaqiangbei has started selling modified "Lobster boxes." This movement is also driving the growth of Chinese large language models (LLMs) internationally. Data from OpenRouter shows Chinese models now account for half of global token consumption, with MiniMax M2.5 leading in usage. MiniMax’s market value has surged, exceeding Baidu’s, and its revenue is now over 70% from international markets. Similarly, Kimi2.5 has seen a spike in paid users and overseas revenue since being adopted as OpenClaw’s primary free model. The integration of OpenClaw is blurring traditional boundaries between hardware makers, developers, and AI companies, creating a new ecosystem for AI-powered hardware innovation.

比推03/11 06:49

After the Lobster Comes Ashore, the Next Game in AI Hardware Lego

比推03/11 06:49

A Monthly Salary of 20,000 RMB, But Can't Afford to Keep a 'Lobster'?

The article "Can a 20,000 Yuan Monthly Salary Not Afford a 'Lobster'?" discusses the viral trend and misconceptions surrounding OpenClaw, an open-source personal AI agent framework. It highlights two contrasting views: one praising its potential as a virtual team member, and another criticizing the hype and privacy concerns. Key points include: 1. OpenClaw experiences vary widely based on deployment methods (local hardware, cloud servers, personal PCs, or vendor-hosted services), affecting functionality and cost. 2. High permissions granted to OpenClaw pose significant security risks, including data breaches and malicious attacks. 3. Performance depends heavily on the underlying AI model (e.g., GPT, Claude) and its associated costs, with token consumption potentially leading to high expenses. 4. OpenClaw is still an immature, technically complex tool requiring substantial time, financial investment, and expertise to use effectively. 5. Users are advised to assess their actual needs, technical capability, and risk tolerance before adoption, as the AI serves as an "amplifier" rather than a standalone solution. The conclusion emphasizes cautious experimentation and independent thinking amidst the hype.

比推03/09 17:12

A Monthly Salary of 20,000 RMB, But Can't Afford to Keep a 'Lobster'?

比推03/09 17:12

Shenzhen Longgang District Plans to Launch Special "Lobster Ten" Policy

Shenzhen's Longgang District is drafting a special policy package, "Ten Measures for OpenClaw" (龙虾十条), to support the development of the open-source AI agent project OpenClaw and One Person Companies (OPCs). The policy aims to position Longgang as a global hub for AI agent entrepreneurship. OpenClaw, a popular local-first AI automation gateway that connects various large models and messaging platforms (like WeChat and DingTalk), allows users to perform tasks via natural language commands. The proposed measures include free deployment services, subsidies for data, hardware (e.g., 30% off AI NAS devices), and model usage (30% reimbursement for calling major domestic LLMs, up to ¥1 million annually). It also offers up to three months of free computing power for new OPCs, rewards for innovative applications (up to ¥2 million), and support for talent (including relocation subsidies, housing, and office space discounts). Additional features include capital investment (up to ¥10 million for early-stage projects), export services, and competitions to attract global developers and individual entrepreneurs. The policy is set to take effect in 2026 for a three-year term.

marsbit03/09 01:41

Shenzhen Longgang District Plans to Launch Special "Lobster Ten" Policy

marsbit03/09 01:41

活动图片

Популярные категорииright-arrow

Популярные тегиright-arrow

Technology Trends

Regulatory Policy