# Сопутствующие статьи по теме AI Agents

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

Research on Commercialization Infrastructure for Crypto Agents: In-depth Analysis of Stablecoin as the Core "Native Currency Layer" and Settlement Network

This article explores the commercialization of AI Agents and the critical "payment gap" they face within traditional financial systems. It argues that stablecoins (like USDC, USDT) provide a superior, native "monetary layer" for AI, enabling programmable, permissionless, 24/7, and transparent value transfer essential for autonomous agents. The piece details infrastructure initiatives from key players: Coinbase's AgentKit and Agentic Wallets for on-chain payments; Circle's CCTP for cross-chain USDC transfers and AgentStack for micro-payments; and Stripe's stablecoin APIs bridging traditional commerce. Collaborations like AWS-Stripe-Coinbase and Google-Coinbase are also highlighted. Key application scenarios are analyzed: 1) DeFi yield optimization, where agents autonomously manage capital across protocols; 2) Ultra-micro-payments (e.g., per API call) enabled by low-fee stablecoin protocols like x402 and Gateway; 3) Automated yield generation through yield-bearing stablecoins, transforming agents into self-sustaining economic units. Major challenges to scaling are identified: private key security and risks like prompt injection; regulatory grey areas regarding agent identity (KYA) and liability; and technical risks including smart contract vulnerabilities and ensuring AI intent alignment during financial operations. In conclusion, the fusion of AI Agents and stablecoins is fundamentally reshaping digital commerce settlement. While security and regulation are immediate hurdles, the infrastructure being built paves the way for a self-operating, agent-driven on-chain economy, shifting humans from transaction approvers to system designers.

marsbit05/26 01:04

Research on Commercialization Infrastructure for Crypto Agents: In-depth Analysis of Stablecoin as the Core "Native Currency Layer" and Settlement Network

marsbit05/26 01:04

The Real Progress and Investment Opportunities of Decentralized AI Computing Power Networks in 2026

In 2026, the AI compute market is marked by centralized GPU consolidation and a significant GPU shortage for smaller players. In this context, Decentralized Physical Infrastructure Networks (DePIN), valued at $9.4B+, have emerged as a viable, revenue-generating alternative. Leading protocols like Aethir ($150M ARR), io.net (130k+ GPUs), Akash, Bittensor, and Render are carving out distinct niches, moving beyond hype to deliver verifiable income primarily from non-crypto-native clients. The key advantage of decentralized GPU networks lies in serving latency-tolerant, cost-sensitive workloads like AI inference, fine-tuning, data preprocessing, and agent operations, offering substantial cost savings (45-80%) compared to major cloud providers. However, reliability variance, lack of robust SLAs, and fragmented tech stacks remain significant adoption hurdles. The sector is maturing with critical 2026 shifts: 1) Evolution of tokenomics towards demand-driven, revenue-linked models (e.g., Render's BME, io.net's IDE), and 2) Clearer enterprise adoption pathways, with traditional firms integrating decentralized compute. For new entrants, opportunities are now concentrated in specialized tooling layers (orchestration, verification, SLA management), vertical applications (e.g., bio-med, content generation), and innovative token designs tied to real usage, rather than generic GPU aggregation. The convergence with the emerging AI Agent economy presents a significant future growth vector.

marsbit05/25 08:01

The Real Progress and Investment Opportunities of Decentralized AI Computing Power Networks in 2026

marsbit05/25 08:01

AlphaGo's Creator Puts AI into a 23-Year-Old Artificial Society: All Three Toughest Challenges for AI Agents Are Here

Demis Hassabis, CEO of DeepMind, has embarked on a new AI research venture by partnering with the long-running space MMO, EVE Online. This collaboration, announced in early May, aims to use the game's 23-year-old, player-driven persistent universe as a testbed for tackling three core challenges in AI agent research: long-horizon planning, memory, and continual learning. Unlike previous DeepMind environments like AlphaGo (Go) or AlphaStar (StarCraft II), EVE Online features no fixed end state. Its single-shard universe has fostered complex, emergent player societies with real economies, political alliances, and wars that can span months or years. These conditions naturally demand the very skills—long-term strategic planning, maintaining memories over extended periods, and adapting to constant change—that are hardest for current AI agents to master. The research will initially use an offline version of EVE, providing a controlled, complex sandbox without interfering with the live player server. This move continues DeepMind's trajectory of using increasingly complex and open-ended virtual worlds for AI training, from Atari games and Go to StarCraft II and the SIMA project. The EVE environment represents a significant step towards testing AI in a persistent, socially complex, and continuously evolving world shaped by human behavior over decades.

marsbit05/25 00:08

AlphaGo's Creator Puts AI into a 23-Year-Old Artificial Society: All Three Toughest Challenges for AI Agents Are Here

marsbit05/25 00:08

Anthropic Major Release: "The Founder's Playbook" - All 4 Stages of Entrepreneurship, Completely Reimagined with AI

**Anthropic Releases "The Founder's Playbook," Reimagining the Four Stages of Startups with AI** The logic of entrepreneurship is being fundamentally reshaped by AI. Anthropic's new handbook, "The Founder's Playbook: Building an AI-Native Startup," defines the AI-native startup as a new species: not a traditional company with AI tools, but a venture driven by AI from day one. The founder's role is transforming from a hands-on builder to a conductor or architect, orchestrating AI agents for execution while focusing on high-level judgment and strategy. Anthropic outlines a product matrix of Claude tools for different tasks: Claude Chat for interactive research, Claude Code for generating production-ready code, and Claude Cowork for automating knowledge-intensive workflows. The handbook structures the startup lifecycle into four stages, detailing core goals, pitfalls, and AI applications for each: 1. **Idea Stage**: Focuses on validating a real problem. The core challenge is avoiding confirmation bias. AI practices include using Claude as a "structured devil's advocate" to challenge assumptions and for automated market/competitor research. 2. **MVP Stage**: Aims to gather early signals of Product-Market Fit (PMF). Key risks are technical debt and scope creep due to rapid AI-assisted development. Recommended AI uses include maintaining project memory documents (e.g., CLAUDE.md), using Claude Code for structured coding, and automating user feedback analysis. 3. **Launch Stage**: Centers on establishing scalable growth, operations, and compliance. Challenges include accelerating technical debt and founders becoming bottlenecks. AI should be used to build an "operating system" for launch—automating routine tasks (scheduling, reporting, content) and code audits—freeing founders for critical decisions. 4. **Scale Stage**: Focuses on achieving sustainable business operations. The main challenge is delegating operational control. AI should be leveraged for differentiated marketing, operational optimization, and building competitive moats through data network effects. The handbook concludes that in the AI era, "Can we build it?" is no longer the primary constraint. The advantage shifts back to foundational strengths: **insight, judgment, and a deep understanding of a specific problem and audience.**

marsbit05/22 13:58

Anthropic Major Release: "The Founder's Playbook" - All 4 Stages of Entrepreneurship, Completely Reimagined with AI

marsbit05/22 13:58

Sui Launches Gasless Stablecoin Transfers With Support From Fireblocks

Sui has launched a protocol-level feature enabling gasless stablecoin transfers on its network. This allows users and businesses to send supported stablecoins like USDsui, suiUSDe, and USDC without paying gas fees or needing to hold SUI tokens, effectively reducing transfer fees to $0.00. Supported by Fireblocks, this initiative aims to remove a major friction point for mass stablecoin adoption by simplifying payment workflows. Sui leadership highlights this as a step toward making the network a global payments rail for businesses, consumers, and AI agents. The feature is a permanent structural change, not a temporary promotion, and builds on Sui's rapidly growing stablecoin ecosystem, which has surpassed $1 trillion in transfer volume.

TheNewsCrypto05/20 21:21

Sui Launches Gasless Stablecoin Transfers With Support From Fireblocks

TheNewsCrypto05/20 21:21

Five Core Forms of AI Agent in YC's Eyes

The article outlines five core architectural patterns for effective AI Agents, emerging from tools like Codex and Claude, that move beyond simple prompts towards reusable, process-based capabilities. 1. **Skills**: Reusable, parameterized workflows that function like method calls, allowing a single process (e.g., "/investigate") to handle various tasks based on input parameters. 2. **Thin Harness**: A lightweight execution framework (~200 lines) that manages the AI model's "hands and feet"—handling loops, file I/O, and context—without becoming bloated. 3. **Resolvers**: Routing tables that map tasks to specific Skills, preventing "context corruption" when managing dozens of Skills and ensuring outputs go to the correct locations. 4. **Latent vs. Deterministic Layer**: A critical separation where LLMs handle judgment, synthesis, and pattern recognition, while deterministic code handles tasks requiring precision, consistency, and low cost (like calculations). 5. **Memory**: A persistent, accumulating knowledge base (e.g., a markdown folder) with a "current trusted conclusion" section and an append-only timeline, enabling the system to learn and retain context over time. Together, these patterns create a "process power"—a durable competitive advantage. Unlike one-off prompt-based applications whose value quickly commoditizes, a well-designed AI Agent system encodes experience into reusable, parameterized workflows, offloads stable rules to code, and continuously learns through memory. This creates a structured, hard-to-replicate capability that can provide sustained value for individuals or businesses, such as an accountant automating client reviews while preserving privacy and accumulating expertise.

marsbit05/20 07:46

Five Core Forms of AI Agent in YC's Eyes

marsbit05/20 07:46

After Developer Numbers Halved: Crypto Isn't Dead, It's Just Giving Up Talent to AI

The title "After a 50% Drop in Developer Count: Crypto Isn't Dead, It's Just Ceding Talent to AI" suggests a shift, not an end. The article analyzes GitHub data showing a significant drop in overall Crypto developer activity from a peak of 45K monthly active developers in 2022 to about 23K in 2026. However, this masks a deeper trend of "talent deleveraging." The exodus consists mainly of newcomers who entered during the bull market for hype-driven roles (e.g., NFT contracts, forked DeFi protocols), with over 50% of developers with less than one year of experience leaving. In contrast, established developers (2+ years of experience) have hit record highs, contributing roughly 70% of the code. They are consolidating in ecosystems with real users and revenue, like Bitcoin and Solana. These experienced builders possess unique skills forged in Crypto's "code is law" environment: the ability to build trust and functional systems from scratch in the absence of external authority or rules, with zero tolerance for error. The article argues that AI's scaling faces structurally similar trust, coordination, and verification problems—particularly regarding compute aggregation, multi-agent incentive alignment, and autonomous payments. Crypto builders are already applying these skills in AI. Examples include CoreWeave (mining to AI compute), OpenRouter (NFT marketplace routing to AI model routing), and projects like Hyperbolic (using crypto-native mechanisms for decentralized compute verification) and EigenLayer (applying restaking logic to AI agent governance). Stablecoin infrastructure is becoming critical for AI agent micro-payments (e.g., x402 protocol). The role of these builders is evolving from writing smart contracts to "designing trusted mechanisms for autonomous AI systems." This shift is reflected in new hiring trends at major exchanges and significant venture capital flowing into the crypto-AI convergence (e.g., funds from Paradigm, Haun Ventures). The article concludes that while developer numbers have halved, the core density of talent has increased, and their uniquely cultivated skills are finding a new, larger stage in the AI era.

marsbit05/18 13:41

After Developer Numbers Halved: Crypto Isn't Dead, It's Just Giving Up Talent to AI

marsbit05/18 13:41

After the Developer Count Halved: Crypto Is Not Dead, It's Just Ceding Talent to AI

Following a significant decline in the total number of open-source crypto developers, from a peak of 45K in 2022 to approximately 23K by 2026, this article argues the industry is undergoing a "talent deleveraging" rather than a collapse. The exodus primarily consists of newcomers who entered during the bull market, while the core of experienced developers (2+ years) has grown to a record high, contributing around 70% of code. These established builders are concentrating in ecosystems with real users and revenue, like Bitcoin and Solana. The article posits that crypto has cultivated a unique skill set in building trustless, autonomous systems with near-zero tolerance for error—a capability now finding high demand in the AI era. As AI scales, it faces structural gaps in decentralized compute aggregation, multi-agent coordination/incentive alignment, and autonomous payment infrastructure. Crypto builders are transitioning their expertise to address these exact problems. Examples include CoreWeave (mining to AI compute), Hyperbolic (decentralized compute verification), EigenLayer (extending restaking mechanisms to AI agent governance), and the x402 protocol (enabling AI agent micro-payments via stablecoins). The role of the crypto builder is evolving from writing smart contracts to designing the rule-based, trust-minimized frameworks necessary for AI-native systems. Venture capital is increasingly funding this convergence, viewing it as a structural opportunity rather than a narrative shift. The core talent and systemic design principles from crypto are not disappearing but being re-priced and applied to the foundational challenges of scalable AI.

链捕手05/18 13:37

After the Developer Count Halved: Crypto Is Not Dead, It's Just Ceding Talent to AI

链捕手05/18 13:37

Morning News | VanEck and Grayscale Submit BNB ETF Amendments on the Same Day; BlackRock in Talks to Invest Billions in SpaceX's IPO; Michael Saylor Posts Bitcoin Tracker Info Again

Crypto & Tech Roundup: VanEck and Grayscale both submitted revised applications for a BNB ETF. BlackRock is reportedly in talks to invest tens of billions of dollars in SpaceX's upcoming IPO. Michael Saylor hinted at another Bitcoin purchase announcement by MicroStrategy, expected next week. The Abu Dhabi sovereign wealth fund Mubadala increased its holdings of BlackRock's IBIT spot Bitcoin ETF to nearly $660 million. In other news, Grafana disclosed a GitHub security breach where code was stolen and a ransom demanded, while Circle's CEO expressed interest in funding teams building AI agents capable of signing legal contracts using its technology stack. Separately, Chinese customs in Zhejiang province cracked down on crypto mining rig smuggling, seizing over 400 machines. Popular memecoins on ETH, Solana, and Base networks were also listed. The roundup concludes with highlights from a Vitalik Buterin podcast on blockchain as a "sanctuary" from AI and an interview on US banks re-engaging with crypto.

链捕手05/18 01:33

Morning News | VanEck and Grayscale Submit BNB ETF Amendments on the Same Day; BlackRock in Talks to Invest Billions in SpaceX's IPO; Michael Saylor Posts Bitcoin Tracker Info Again

链捕手05/18 01:33

MuleRun CTO: The Moat of Agents Lies in Data Density and User Memory

In a speech titled "Handing AI's Keys to the On-Chain Controllers," MuleRun CTO Shu Junliang discussed the evolution and security of AI Agents in finance and Web3. He outlined six dimensions for a complete AI assistant: dialogue, data input, agent capability, execution environment, user memory, and continuous learning. MuleRun's product integrates these through features like multi-platform IM bots, real-time multi-asset data, smart model routing, cloud sandboxes, persistent user profiles, and a shared knowledge network. Shu emphasized that while AI Agents are advancing from assisting to autonomously executing decisions—potentially enabling individuals to operate like small funds—safety remains paramount. He detailed MuleRun's security measures, including local key handling, isolated sandboxes, full audit trails, and strict permission controls. However, he acknowledged persistent risks like data exposure, model hallucinations, prompt injection, and the "black box" nature of AI decisions, advising manual confirmation for financial operations. He identified key trends: the shift from human-led to Agent-led on-chain interactions requiring infrastructure redesign; the erosion of information advantages replaced by competition in execution speed and strategy; and the balancing effect of Agents, which democratize access but ultimately advantage those with superior judgment. Shu concluded that an Agent's true moat lies in data density and accumulated user memory, not easily replicable technology, and that while Agents will reshape finance and Web3, human oversight over critical decisions must remain.

marsbit05/14 08:50

MuleRun CTO: The Moat of Agents Lies in Data Density and User Memory