Ethereum Creator Vitalik Buterin Reveals Privacy-Focused AI Experiment

Ethereum creator Vitalik Buterin tests three-layer privacy model for frontier AI.

Disclaimer: The opinions expressed by our writers are their own and do not represent the views of U.Today. The financial and market information provided on U.Today is intended for informational purposes only. U.Today is not liable for any financial losses incurred while trading cryptocurrencies. Conduct your own research by contacting financial experts before making any investment decisions. We believe that all content is accurate as of the date of publication, but certain offers mentioned may no longer be available.

In a recent X post, Ethereum co-creator Vitalik Buterin revealed he was doing a bit of a self-experiment.

The goal of the experiment was to use his personal health and travel data to provide personalized diet and exercise recommendations for himself, using frontier models but in a way that avoids leaking any private information to them.

For the experiment strategy, Buterin used a local model (Qwen 3.8 Flash Next) to orchestrate and also use powerful remote models via tool calls to benefit from higher-level thinking and knowledge that the local model lacks.

There is a three-layer approach to privacy: first, to avoid leaking personally identifiable information, or leaking his identity through his writing style, as the local model writes the queries to the frontier model.
The second layer uses zkAPI to prevent his identity from being exposed through the payment channel, while the third uses Tor to provide network and IP-level privacy. “You need all three,” Buterin wrote.

Doing a bit of a self-experiment.

Goal: use my personal health and travel data to provide personalized diet and exercise recommendations for me, using frontier models but in a way that avoids leaking to them any private information.

Strategy: use a local model (Qwen 3.8 Flash Next) to orchestrate, and use powerful remote models as a tool call to benefit from higher-level thinking and knowledge that the local model does not have.

Three-layer approach to privacy:

Avoid leaking personally identifiable information, or leaking my identity through writing style -> local model writes the queries to the frontier model, not me
Avoid leaking who I am through the payment channel -> zkAPI
Avoid leaking who I am through networking / IP -> Tor

You need all three (and finally we have all three, at least to some extent)

A skill file teaches the local model when and how to construct minimally-data-revealing requests to remote models. Use zkAPI-wrapped-with-Tor as a CLI tool.

And everything works! I got the recommendations back, info from frontier models helped to improve them.

Main deficiencies:

* Tor is really not optimized for request-by-request de-linking, which is the only form of network-layer privacy that really makes sense in today’s world (long-running identifiers are too fragile). Probably not private enough, and latency 10-100x higher than it could be, at the same time.
* The skill file’s request construction strategies are definitely far from optimal.
* Qwen 3.8 Flash Next is still too slow for comfort. It’s comfortably running at 20-30 TPS, but it would only really start to feel fast at 100+
* There is a tradeoff: the more careful you are about what data you give to a remote model, the less it can help you

https://t.co/1g81Mp1sWW

A skill file teaches the local model when and how to construct minimally data-revealing requests to remote models. Buterin said he used zkAPI wrapped with Tor as a command-line tool.

According to Buterin, the experiment worked and returned recommendations that were improved using information from the frontier models: “And everything works! I got the recommendations back, info from frontier models helped to improve them.”

Buterin outlined three main deficiencies, highlighting the trade-off; the more careful you are about what data you give to a remote model, the less it can help you.

While Buterin’s latest discourse on X appears to focus solely on AI, his vision for Ethereum remains unchanged, with him envisioning it as a ‘cryptographic world computer‘ in the years ahead.

In a post towards September’s close, Buterin noted that Ethereum’s technology has evolved greatly over the last fifteen years, and is poised to evolve even more over the next three — to the point where it becomes fair to call what Ethereum is moving towards a qualitatively different kind of system.

According to Buterin, next year’s Pectra hard fork might likely be Ethereum’s last “normal” fork, with features and technology that would be recognizable to someone in 2015. Everything after that involves recursive STARKs, automated formal verification, highly optimized consensus algorithms, and making it all quantum-safe. With PeerDAS, Ethereum began its transition from being just a blockchain to being something much more powerful. Starting after Pectra, this transformation becomes Ethereum’s primary story.

The outcome is much cheaper, more scalable, and more private high-security computation than anything that could be done with the previous era’s technology alone: the cryptographic world computer.