According to CoinMeta, Vitalik.ETH disclosed in a post on Farcaster that he is testing a privacy-protected AI health system that combines local models, zkapi, and Tor. This system processes personal health and travel data through a three-layer architecture. It utilizes a local model Qwen with 3.8B data to build queries, zkapi to hide payment information, and Tor to conceal IP addresses. Currently, the system faces issues such as high Tor delays, insufficient running speed of the local model, and trade-offs between data protection and model assistance capabilities. The relevant code has been submitted to the Ethereum zkapi repository.