Hello all. I am new here and I apologize if this is the incorrect location for this post. I also realize that I have no reputation yet here, so I apologize if this looks like spam.
I have blogged about the idea of trying to get public funding for Fediverse instances:
https://dynamorando.pages.dev/blog/the-public-web/
The main takeaway -
Public contracts to be bid on by your favorite tech corporations to host public Fediverse instances on their cloud, renewable every 4 years. Basically pre-paid reserved instances, to be provisioned out by your favorite civil three letter agency. Silicon Valley hosts the infastructure, and provides infastructure services to support it (“YouGov: DDoS protection by Cloudflare”.)
Semi-public institutions provision, administer, and moderate their own Fediverse flavors: NPR, PBS, University of Whatever, non-profits like Wikipedia, Mozilla, and so on. Practical cost keeping measures could be implemented: allowing only text based communications, or rate limiting posts, or enabling targeted data retention policies (posts deleted after N days), etc.
Here’s the kicker: For any institution wanting to take advantage of this program, they would be subject to the following restrictions:
- Their code hosting the social media derivative must be open source.
- The data being submitted by their users must have a public and easy to understand data governance plan.
- If the user submitted content would be licensed for example as Creative Commons, describe what is being collected, how it may be used elsewhere, etc. For example, if NPR hosted their own Fediverse instance, could the users opt-in in allowing some of their content to be used to train AI models? Etc. This would provide a public data set for companies to leverage.
- The data collected by their social media derivative must be subject to user data rights: Users have a right to read, update, delete their data. They have the right to move their data elsewhere. And so on.
I choose the Fediverse (really, the ActivityPub protocol) because I don’t want to tend towards N number of walled sites that can’t interoperate their content: I want to try and keep the benefits of centralization and decentralization.
Would this be something the Mozilla Foundation would support?