Skepticism Mounts Over AI Agent 'LARP'
In a recent podcast, the founder of human verification protocol Ethos argued that most current crypto AI agent narratives are a "LARP." He claims today's agents lack true autonomy and are just following human instructions, suggesting a need for verification layers to distinguish humans from bots as the agent economy grows.
The core of the "LARP" argument is that most AI agents are just sophisticated automation, not truly autonomous economic actors. True autonomy requires agents to own assets, pay for their own resources, and execute complex tasks without human intervention, a reality limited by the current centralized internet's structure which requires human proxies for financial transactions. Projects like Fetch.ai and Autonolas (Olas) are building the foundational layers for more advanced agents. Fetch.ai provides frameworks for creating autonomous agents that can perform tasks and make decisions independently, aiming for a decentralized digital economy. Olas offers a unified network for off-chain services like automation and co-owned AI, enabling complex logic to be executed decentrally. Crypto infrastructure is the unlock for genuine agent autonomy. By using crypto wallets for exclusive fund control and decentralized identifiers (DIDs) for building on-chain reputation, agents can move from being mere tools to sovereign market participants that can transact and build trust permissionlessly. This circumvents the need for traditional bank accounts and human legal identities. The Solana ecosystem is a growing hub for AI agent development, with toolkits like the Solana Agent Kit, Eliza, and GOAT framework enabling developers to connect AI to on-chain actions. Projects are emerging that focus on everything from AI-powered trading and lending to privacy-focused agent coordination using ZK proofs. For instance, Cainam Ventures uses Solana-based AI agents for DEX arbitrage between platforms like Jupiter and Orca. This rise of non-human on-chain participants introduces a critical need for verification and reputation, which is the problem Ethos aims to solve. As agents begin to manage treasuries and execute trades, a standardized method to verify an agent's identity and assign accountability—a "Know-Your-Agent" (KYA)—becomes essential to prevent manipulation and build trust in an AI-driven economy. Despite the skepticism, the agent economy is expanding rapidly, with over 20,000 active AI agents deployed across blockchains as of early 2026, a 300% increase from late 2025. Analysts project that machine-driven transactions could influence up to thirty trillion dollars in global purchases by 2030, positioning public blockchains as a primary settlement layer for autonomous systems.
