# 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.

IOSG|Decentralized AI: Ethereum's Next Decade Bet

A Glimpse into Decentralized AI: Ethereum's Next Decade Bet In a future scenario, AI assistants could autonomously handle complex tasks like booking flights by coordinating with specialized AI agents. However, a critical challenge emerges: how can AI determine which other agents to trust? Current AI agents operate in isolated ecosystems (e.g., OpenAI, Google), unable to communicate or verify each other’s reliability—a "trust crisis" akin to early fragmented email systems. While protocols like Google’s A2A (Agent-to-Agent) and Anthropic’s MCP (Model Context Protocol) enable AI-to-AI communication, they lack a trust mechanism. ERC-8004 proposes a solution by leveraging Ethereum to provide AI agents with: 1. **Identity**: A unique, verifiable on-chain ID (as an NFT) to prevent forgery. 2. **Reputation**: A transparent, immutable rating system (like Uber or Taobao reviews) recorded on-chain. 3. **Validation**: For high-risk tasks, third-party verification via cryptographic proofs or trusted execution environments. Ethereum’s neutrality is key: it offers a decentralized, tamper-proof foundation for AI identity and reputation, avoiding reliance on any single corporation. The Ethereum Foundation’s dedicated dAI (Decentralized AI) team aims to position Ethereum as the settlement and coordination layer for the AI economy, marking a strategic shift from DeFi/NFTs to "on-chain intelligence." The ecosystem is already advancing, with 1,100+ developers, 70+ demos, and integrations like Taiko L2 and x402 (a payment protocol by Coinbase/Cloudflare). Combined, ERC-8004 and x402 could enable a closed-loop economy where AI agents autonomously complete tasks and receive payments. In the long term, this could allow users to hire AI agents effortlessly, monetize their own agents, and enjoy interoperable AI assistants beyond walled gardens. While success isn’t guaranteed, ERC-8004 represents a pivotal narrative shift for Ethereum—potentially becoming the "TCP/IP" for decentralized AI.

marsbit01/19 13:59

IOSG|Decentralized AI: Ethereum's Next Decade Bet

marsbit01/19 13:59

a16z: Three Major AI Trends for 2026

a16z: Three AI Trends for 2026 1. AI Takes on Substantial Research Tasks: AI models are evolving to handle complex, abstract instructions and assist in research, particularly in reasoning. They are beginning to solve difficult problems and foster a new "generalist" research style that focuses on connecting ideas and making inferences from hypothetical answers. This requires new "nested agent" workflows where models collaborate and refine each other's outputs, though better model interoperability and compensation methods (potentially via blockchain) are needed. 2. The Shift from KYC to KYA (Know Your Agent): The bottleneck in the agent economy is shifting from intelligence to identity verification. With non-human identities vastly outnumbering humans in finance, there's a critical need for a "Know Your Agent" infrastructure. Agents require cryptographically signed credentials to transact, linking them to their principals, constraints, and liabilities. 3. Solving the Open Web's "Invisible Tax": AI agents are disrupting the economic foundation of the open web by extracting data from ad-supported sites (the context layer) while bypassing their revenue models. This creates an "invisible tax" that threatens content creation. Solutions are needed to automatically reward content creators, moving from static licensing to real-time compensation systems using technologies like nanopayments and attribution standards, potentially blockchain-enabled.

marsbit01/12 08:09

a16z: Three Major AI Trends for 2026

marsbit01/12 08:09

Visa Crypto Head: Eight Major Evolution Directions for Crypto and AI by 2026

Cuy Sheffield, Head of Crypto at Visa, outlines eight key themes for the evolution of cryptocurrency and AI by 2026, emphasizing a shift from theoretical potential to practical, reliable implementation. Cryptocurrency is transitioning from a speculative asset class into a high-quality technology. Its underlying infrastructure has become faster, cheaper, and more reliable, shifting its primary value from speculation to utility, particularly for payments and settlement. Stable币 are the clearest example of this, succeeding on objective merits like cost, speed, and global reach, and enabling adoption without ideological buy-in. As crypto becomes infrastructure, distribution capabilities and existing customer relationships—often held by large, regulated institutions—will matter more than pure technical novelty. For AI, the focus is shifting from raw intelligence to trust and reliability. AI agents are proving most valuable not as autonomous entities but as tools that reduce coordination costs in knowledge work—spanning research, analysis, and operations, not just coding. Their current limitation isn't capability but trust, requiring systems that are verifiable, consistent, and transparent. Successful AI integration is now a systems engineering challenge, relying on architecture, state management, and monitoring, not just model prompts. This development is creating a tension between the capital-intensive, centralized development of frontier models and the rapid iteration of open-source alternatives, leading to unresolved governance questions. Finally, the convergence of these fields is enabling new economic interactions. Programmable money, like stablecoins, is emerging as the native currency for AI agents, allowing for automated, fine-grained, and continuous payment flows between machines, opening the door to novel economic behaviors. The overarching trend is a move from flashy technological novelty to a focus on reliability, governance, and distribution, as both technologies become deeply embedded into real-world systems and workflows.

marsbit01/07 12:10

Visa Crypto Head: Eight Major Evolution Directions for Crypto and AI by 2026

marsbit01/07 12:10

Messari 2026 Crypto Thesis: Why Speculation Is No Longer Enough (Part 1)

Messari's 2026 Crypto Thesis argues that the market is shifting from pure speculation to systemic integration. The report highlights several key trends: First, it identifies a valuation trap in new Layer 1 blockchains. Many VC-backed L1s with high fully diluted valuations lack fundamental revenue, as token issuance far exceeds gas fee income. The market is expected to strip away the "monetary premium" of these tokens, with only a few ecosystems like Solana and Base maintaining real traction. Second, chain abstraction is emerging as a critical strategy, where blockchains become backend infrastructure. Users interact without needing to know which chain they're on, shifting competitive advantage from execution speed to control over user flow and liquidity. Third, the rise of an agent economy is predicted, with AI agents potentially driving up to 80% of on-chain transactions by 2026. This emphasizes the importance of APIs and machine-readable financial primitives over user-facing interfaces. Fourth, equity perpetuals (e.g., on Hyperliquid) are gaining traction as synthetic derivatives for global stock exposure, offering a more scalable alternative to tokenized stocks. Finally, DePIN is highlighted as the sector most likely to generate hundreds of millions in verifiable revenue by 2026, driven by real demand for AI compute resources rather than speculative hardware deployment. The report underscores that while fundamentals are becoming more critical, market narratives and liquidity flows remain key drivers of returns.

marsbit01/02 13:41

Messari 2026 Crypto Thesis: Why Speculation Is No Longer Enough (Part 1)

marsbit01/02 13:41

Written at the End of 2025: Code, Power, and Stablecoins

"Stablecoins have firmly established themselves as the foundational infrastructure for the next decade of financial services, with the market surpassing $300 billion in 2025. This growth is driven by a fundamental shift in trust: relying on transparent, verifiable code and math rather than opaque promises from centralized intermediaries, as starkly illustrated by the Synapse bankruptcy. Self-custody models change risk dynamics, eliminating intermediary risk (though not issuer risk) and reducing the necessity for traditional insurance like FDIC. Stablecoins offer inherent global reach, with the main bottleneck being local fiat on/off-ramps rather than rebuilding entire banking stacks per country. The emergence of payment-specific blockchains like Tempo and Arc faces the challenge of building trust from scratch, competing with the established security of networks like Solana and Ethereum. The real potential of 'agentic finance' lies in automating mundane financial tasks through smart contracts with enforced permission boundaries, providing security that traditional systems cannot. However, the rapid growth attracts teams with inadequate security practices, a critical misstep for financial infrastructure. Furthermore, as real business activity moves on-chain, solving for privacy through selective disclosure—not full anonymity—becomes crucial to prevent competitive intelligence leaks. The true opportunity lies not just in rebuilding existing fintech more efficiently but in leveraging programmable money and internet-native capital markets to reimagine financial services entirely."

marsbit12/29 01:35

Written at the End of 2025: Code, Power, and Stablecoins

marsbit12/29 01:35

What Should the New Financial Infrastructure of the AI Era Look Like?

The article explores the limitations of current prediction markets, which, despite their success in aggregating information through risk-sharing (e.g., accurately predicting election outcomes), suffer from a flawed economic model: their most valuable output—information—becomes a free public good once generated. This restricts their viability to entertainment-driven domains like elections and sports, while critical areas (geopolitical risk, regulatory outcomes, etc.) remain unaddressed. The author proposes "Cognitive Finance," a new infrastructure designed from first principles for the AI and crypto era. Key components include: - **Private Markets**: Using trusted execution environments (TEEs) to keep prices confidential, enabling entities (e.g., hedge funds, corporations) to pay for exclusive signals without leakage to competitors. - **Combinatorial Markets**: Moving beyond isolated events to maintain a joint probability distribution, where trades update correlated outcomes simultaneously, akin to a neural network. - **Agent Ecosystems**: AI-native markets where specialized agents (trading, evaluation, information acquisition) operate with strict isolation between price access and information sourcing to prevent self-cannibalization. - **Human Intelligence**: Interfaces allowing humans to contribute knowledge via natural language without seeing prices, compensated based on predictive accuracy. The vision is a decentralized, composable infrastructure where AI systems and humans collaboratively build a continuously updated, probabilistic world model. This transcends today’s prediction markets, aiming to transform decision-making in finance, supply chains, geopolitics, and beyond by making uncertainty tradable and knowledge liquid.

marsbit12/26 11:06

What Should the New Financial Infrastructure of the AI Era Look Like?

marsbit12/26 11:06

活动图片