Ethereum founder Vitalik Buterin posted on X that he is conducting an AI self-experiment. The goal is to use personal health and travel data to obtain personalized diet and exercise recommendations through cutting-edge AI models, while avoiding the disclosure of any private information. The experiment employs a three-layer privacy approach: using a local model (Qwen 3.8 Flash Next) to write queries to avoid disclosing personally identifiable information or writing style; using zkAPI to prevent identity exposure through payment channels; and leveraging Tor to provide network and IP layer privacy. Buterin stated that the experiment was successful and yielded recommendations improved by cutting-edge models, but also pointed out major drawbacks such as insufficient optimization of Tor in request de-correlation, slow local model speed, and the trade-off between data privacy and model assistance capabilities.