I built Btfy, an experimental blockchain that uses weather observations as a source of unpredictability for consensus

The motivation is simple. Proof of Work generates uncertainty by performing computations. I wanted to explore whether we could obtain some of that uncertainty from the real world instead.

Btfy combines weather observation data with a Verifiable Delay Function (VDF):

Weather Observation Data -> VDF -> Block -> Hash Chain

In the current implementation, observation data is collected from multiple locations, a VDF is computed, and the result is used to generate a block. Furthermore, because each block depends on the hash of the previous block, making it difficult to replay previously observed weather data to construct an alternative history.

This is strictly an experimental project and not a production-ready consensus protocol. I am particularly interested in the security assumptions here: specifically, how much uncertainty weather observation data can actually provide, how an attacker might manipulate or predict the beacon, and whether the VDF plays a meaningful role in this design.

Btfy is still a very early-stage project and requires various forms of validation. However, I hope this sparks exploration into the possibilities of new blockchains based on real-world information. I would love to hear your feedback.

Thanks!