Decentralized AI Infrastructure Projects Gain Funding and Traction

- The $11.5 million Series A for Based was led by crypto-native VC firm Variant, with notable participation from Cyber Fund, 1kx, and angel investors like former Coinbase CTO Balaji Srinivasan and Ethereum researcher Justin Drake. - This funding round for Based established a post-money valuation of approximately $85 million for the company, which aims to address the AI industry's "GPU bottleneck" by creating a secondary market for underutilized data center capacity. - Acurast's network leverages the Trusted Execution Environments (TEEs) within smartphones to enable confidential AI inference, meaning tasks can be processed without exposing sensitive data to the phone's owner or other external parties. - The Acurast network already handles over 1 million daily on-chain transactions and spans more than 140 countries, providing a globally distributed alternative to centralized cloud providers. - A key feature of the Acurast integration on Base is the support for native USDC payments, allowing AI agents to autonomously pay for compute resources in real-time without intermediaries. - The decentralized AI sector is currently valued between $12 billion and $30 billion, a fraction of the estimated $12 trillion valuation of centralized AI giants, highlighting a significant valuation gap that investors are targeting. - The core technology behind Acurast is known as a Decentralized Physical Infrastructure Network (DePIN), a model that incentivizes individuals to share their physical hardware resources to create a decentralized network. - A forthcoming feature on the Based mainnet, called "Auto-Lease," will permit autonomous AI agents to directly manage their own compute budgets by interacting with the marketplace, essentially allowing an AI to pay for its own training and inference.