Incentive Design for Open Data

[davidgasquez]

I'm not entirely sure how useful the "packaging side" is for the world. Being a public good, it suffers from the tragedy of the commons and the likes.

That said, there might be effective mechanisms to incentivize data generation and sharing. I'm thinking about all the new "web3 mechanisms" like Hypercerts, Retroactive Public Good fundings, Gitcoin Rounds, ...

[davidgasquez]

Dune opened their spellbook and people happyly jumped into the code and started adding models and metrics without any rewards besides having better data.

[davidgasquez]

How can a project like this resits the incentives of going the enterprise/private data direction? Does it need to resist it or should it embrace and build with that in mind (e.g: ACLs)?

[davidgasquez]

There should be incentives at the different levels:

• Dataset Production
• Dataset Curation
• Dataset Consumption

For the production of datasets, an idea inspired by al the recent crypto-economics developments might be to use an Ethereum compatible L2 Chain (e.g: Optimism, Base) as the dataset registers, and, at the same time, have a way to request those datasets and reward creators maintainers of those datasets.

Very handwave-y but using a blockchain for provenance helps at the time of distributing financial incentives. If you know user X uploaded the dataset and that Y curated it, once Z queries / access the data, you can reward them appropiately¹.

Now, for the consumer to have an incentive on all of this... I can only think of extreme data quality and easy UX.

¹ 👋 👋 📓

[davidgasquez]

Why and how places like Wikipedia and the Arch User Repository work? Can a dataset based tool/product/framework use a similar setup?