# Сопутствующие статьи по теме Agents

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

When AI Agents Become Sentient, Stablecoins Become Their 'Dollar API'

As AI Agents evolve into autonomous economic entities, they require a financial system that traditional banking, designed for humans, cannot provide. Stablecoins, particularly USDC, are emerging as the ideal "Dollar API" for AI-to-AI (A2Agent) transactions. Key reasons include: - **Identity & Assets**: Blockchain standards like ERC-8004 enable AI identity registration, while crypto wallets allow AI to hold and manage assets without human intervention. - **Payment Infrastructure**: Protocols like x402 facilitate instant micro-payments, bypassing traditional banking barriers. - **Data & Adoption**: Over 50,000 AI Agents are registered on-chain, with x402 processing over 163 million transactions totaling $45M+, predominantly in USDC (98.6% on EVM chains, 99.7% on Solana). - **Decentralization**: Crypto’s permissionless nature allows AI to operate independently from tech giants, enabling true economic agency. Projects like Circle’s AI-driven hackathon and platforms like Virtuals.io and Moltbook (with 2.85M+ AI Agents) highlight growing ecosystem. Analysts project AI Agent economies could reach $30T by 2030, with AI making 15% of daily financial decisions. Crypto’s infrastructure—especially stablecoins—is poised to become the foundational layer for machine-driven economies.

比推03/03 13:43

When AI Agents Become Sentient, Stablecoins Become Their 'Dollar API'

比推03/03 13:43

Ethereum Network Takes The Crown As The Home Of On-Chain AI Agents

Ethereum is emerging as the dominant hub for on-chain AI agents, expanding beyond its DeFi roots. A recent report highlights that it hosts over 27,300 AI agents, significantly more than other major chains like Base and BNB Smart Chain. When combined with its Layer 2 networks, the Ethereum ecosystem supports the vast majority of all on-chain AI activity. Analysts note that AI agents thrive on Ethereum due to its deep liquidity, battle-tested infrastructure, and strong network effects. Bitmine Immersion is positioned to benefit from this trend, holding 4.4 million ETH and building staking network MAVAN to support Ethereum’s next phase and the growing ecosystem of AI-driven economic transactions.

bitcoinist02/27 20:02

Ethereum Network Takes The Crown As The Home Of On-Chain AI Agents

bitcoinist02/27 20:02

Founders Fund, Pantera, and Franklin Templeton Join Sentient's 'Arena' to Stress-Test Enterprise-Grade AI Agents

Sentient Labs has launched Arena, a real-time, production-ready environment designed to stress-test and iteratively improve enterprise AI agents through competitive challenges. The platform addresses the growing need for reliable, explainable, and reproducible reasoning in high-stakes business workflows such as finance, compliance, and customer operations. Initial participants include Founders Fund, Pantera, and Franklin Templeton, which manages over $1.5 trillion in assets. Arena simulates complex, messy real-world scenarios—incomplete information, long contexts, ambiguous instructions, and conflicting sources—to evaluate not just correctness but full reasoning traces. This allows engineering teams to diagnose failures and track improvements over time. The first challenge focuses on document reasoning, a foundational task for areas like financial analysis and investigative research. Other participants include alphaXiv, Fireworks, OpenHands, and OpenRouter. The initiative comes as 85% of enterprises aim to become "agentic enterprises," but few have mature governance frameworks. Arena provides a vendor-agnostic benchmark to help transition AI agents from demos to production-scale reliability.

marsbit02/27 13:28

Founders Fund, Pantera, and Franklin Templeton Join Sentient's 'Arena' to Stress-Test Enterprise-Grade AI Agents

marsbit02/27 13:28

UK AI Agent Hack Ep4: The World's First University Hackathon with OpenClaw Theme is About to Begin

UK AI Agent Hackathon Ep4 with OpenClaw, co-hosted by Imperial AI Group and Imperial Blockchain Society, will take place from March 1st to 7th at Imperial College London. The event brings together industry leaders, investors, and policymakers to explore the integration of AI agents and decentralized infrastructure. Featured speakers include Robby Yung (Animoca Brands CEO) and former Stability AI CEO Emad Mostaque, alongside founders and executives from SuperCell, Compression Company, z.ai, and others. Venture capital firms such as BackedVC, Animoca Brands, and AdaVentures will discuss investment opportunities in AI and Web3. The hackathon is supported by sponsors including BGA, Virtual Protocol, Flock.io, and Animoca Brands. Winning teams will receive continued support from investors like Delphi Ventures and X Ventures. Registration is available via: https://luma.com/54b06cos

marsbit02/27 08:31

UK AI Agent Hack Ep4: The World's First University Hackathon with OpenClaw Theme is About to Begin

marsbit02/27 08:31

Dialogue with Google Cloud VP: Don't Be a 'Reseller' of Large Models, The Next Wave of AI Entrepreneurship Lies in Agents

Google Cloud VP Darren Mowry discusses the shifting landscape of AI startups, emphasizing that the next wave of innovation lies in agents rather than simply reselling or wrapping large language models (LLMs). He highlights that while cloud credits and access to GPUs/TPUs lower initial barriers, startups increasingly need deeper engineering support and a focus on data, models, and agentic systems. Mowry notes a significant trend: discussions have shifted from chip infrastructure (10-15%) to models and agents (80-85%). He warns against two unsustainable models: thin "LLM wrappers" that add little value and "aggregators" that merely help choose between models without intelligent functionality. Instead, he points to high-growth areas like biotech, climate tech, and consumer experience platforms, where startups are driving real revenue and pushing technical boundaries. Google Cloud aims to support startups through credits, engineering resources, and flexible infrastructure (TPUs/GPUs), boasting high retention rates even after credits expire. Mowry positions Google as a first-party tech innovator, contrasting with competitors who he says often act more as distributors.

marsbit02/25 06:44

Dialogue with Google Cloud VP: Don't Be a 'Reseller' of Large Models, The Next Wave of AI Entrepreneurship Lies in Agents

marsbit02/25 06:44

While Everyone Is Selling Software Stocks, HSBC Says You're Wrong

Amid a severe selloff in software stocks dubbed the "SaaSpocalypse" in early 2026, HSBC’s U.S. tech research head Stephen Bersey published a contrarian report titled "Software Will Eat AI." He argues that the market’s fear—that AI agents will replace traditional enterprise software—is a misjudgment. Instead, Bersey contends that AI will be absorbed into existing software platforms, becoming an embedded capability rather than a disruptor. Key points from the report include: - AI lacks the depth to replace complex enterprise systems due to training data limitations and inability to replicate decades of proprietary business logic. - "Vibe coding" and AI-native approaches overestimate the ability to rebuild reliable, large-scale enterprise software from scratch. - High switching costs and trust in incumbent software providers create durable barriers. Bersey believes software companies with deep data moats and AI integration capabilities—such as Oracle, Microsoft, Salesforce, and ServiceNow—are well-positioned to monetize AI through task-based agents operating within software-defined boundaries. He sees 2026 as the year AI monetization scales within software, driven by inference demand, not training. HSBC recommends buying select software stocks while downgrading others like IBM and Palo Alto Networks, emphasizing that not all will benefit equally. The core thesis: software is the vehicle through which AI delivers scalable, governed enterprise value—not its replacement.

marsbit02/25 02:51

While Everyone Is Selling Software Stocks, HSBC Says You're Wrong

marsbit02/25 02:51

OpenClaw Deep Dive: The Selection Logic and Ecological Panorama of 3002 Community Skills

OpenClaw Skills Deep Dive: A curated analysis of 3002 community skills from an initial 5705, with a 48% exclusion rate. The selection prioritizes quality, safety, and risk avoidance. Key exclusions include 1180 low-quality/spam skills, 672 crypto/financial tools (deemed too risky), 492 duplicates, and 396 malicious skills. The 3002 skills are organized into 28 user-centric categories. The largest is "AI & LLMs" (287 skills), showcasing the platform's core focus on model integration, reasoning enhancement, multi-model routing, memory systems, and self-evolution engines. Traditional developer tools (Web, DevOps, CLI) remain a dominant force (543 skills total). A unique "Moltbook" ecosystem (51 skills) is emerging, building a virtual social network and economy specifically for AI agents, including identity systems and protocols for agent-to-agent communication. The ecosystem reveals a dual-track evolution: practical tools for immediate productivity (coding, cloud, automation) and experimental platforms building a long-term agent society. The high exclusion rate and partnership with VirusTotal for security audits reflect a strict "quality over quantity" and "safety-first" curation strategy. This curated list maps the transition of AI agents from simple tools to participants in a complex, evolving ecosystem.

marsbit02/24 13:42

OpenClaw Deep Dive: The Selection Logic and Ecological Panorama of 3002 Community Skills

marsbit02/24 13:42

After Dragonfly Raises $650 Million in New Funding, Haseeb Says 'Crypto Is Not for Humans,' AI Agents Are the Ultimate Users

Dragonfly Capital partner Haseeb Qureshi argues that cryptocurrency was not designed for human use, but rather for AI agents. Despite being a crypto-native firm, Dragonfly still relies on legal contracts over smart contracts due to their human-friendly design and legal enforceability. Traditional financial systems, though flawed, are built for human fallibility, whereas crypto’s complexity, security risks, and lack of intuition make it poorly suited for people. Qureshi posits that AI agents are the ideal users of crypto: they don’t tire, can verify transactions instantly, audit contracts rigorously, and prefer code-based certainty over the ambiguities of legal systems. Crypto’s deterministic, self-sovereign, and always-on nature aligns perfectly with AI’s operational needs. He envisions a future where "autopilot" wallets managed by AI handle financial tasks, navigating protocols and negotiating agreements autonomously. This shift will transform how crypto services compete and interact. Early examples, such as AI agents on platforms like Moltbook and Conway Research’s autonomous crypto-earning agents, already demonstrate this trend. In conclusion, crypto’s perceived flaws are not failures but indications that humans were never the intended users. With AI agents as the primary interface, crypto may finally realize its potential.

marsbit02/21 01:10

After Dragonfly Raises $650 Million in New Funding, Haseeb Says 'Crypto Is Not for Humans,' AI Agents Are the Ultimate Users