I love to see more decentralized storage solutions popping up, but here is the thing: 

NO, YOU CAN'T CREATE A PUBLIC GOOD DATASET WITHOUT A REAL, DECENTRALIZED VALIDATION LAYER. NONE OF THESE SOLUTIONS CAN DELIVER THAT.

 Here is why ⤵️

Head of Ecosystem at KYVE Network don't trust, verify & validate.

Binance Angel

1/ When you create a public good dataset, you want everyone to be able to access the data freely and trust that they can rely on it. 

To achieve this, they need to be sure that the data they are using is accurate and has not been corrupted, whether intentionally or not.

2/ Even when multiple data sources upload to a decentralized storage solution, there’s no guarantee you’re retrieving the correct data. (all examples are with Arweave, but work for any solution)

3/ With a provenance layer, you would only ensure the origin of the data, not whether the main data source itself is corrupted  so the trust issue would still remain.

4/ This is where a validation layer comes in. It cross-checks multiple independent data sources through validators who verify whether the uploaded data matches their own. Once the quorum is reached, the data is marked as valid and is verified and verifiable.