Dan Romero
@dwr.eth
New doc from @v on what’s next for the protocol. Worth a 5 min read. https://hackmd.io/@farcasterxyz/SyeXtT4fh https://i.imgur.com/iVW04Dk.jpg
9 replies
0 recast
0 reaction
Jackson 🎩🍖
@jacks0n
Amps proposal harms (1) and (2) for spurious benefits to (3) imo. Why not just make publishing follows easier for devs? Protocol should prioritise simplicity of core primitives for devs to play with. No devs no users no protocol https://hackmd.io/IP-8snyMQfOGxV3LUjlJbA
1 reply
0 recast
0 reaction
Dan Romero
@dwr.eth
What happens if a TikTok like client becomes dominant and the default form factor is no follows? Show me the incentive I’ll show you the outcome. :)
1 reply
0 recast
0 reaction
Jackson 🎩🍖
@jacks0n
As a protocol builder is your goal to decide what clients people should build? Amps changes the thesis from “The social graph is powerful but it needs to be decentralised” to “the canonical social graph design all the major platforms have landed on (follows) is wrong, it creates bad incentives”
1 reply
0 recast
0 reaction
Dan Romero
@dwr.eth
The job of a protocol builder is to figure out what parts of the protocol have strong enough incentives to be followed in a decentralized manner. And “norms” is not a good enough answer imo. Too much tech history we’re well-intentioned protocols get captured or corrupted. :)
1 reply
0 recast
0 reaction
Jackson 🎩🍖
@jacks0n
True, but there’s a diff between norms and a good data model. Does replacing the follow primitive with an algorithm the inflates away social capital actually encourage clients to publish? As a dev it looks like a basic feature missing that now I have to build myself, I lose the benefit of shared infra
1 reply
0 recast
0 reaction
Jackson 🎩🍖
@jacks0n
As a user it feels like it harms the portability of my audience, maybe engagement is a better proxy for that though and clients can project that knowledge more usefully into some form of audience metric. Totally open to being wrong 😊
0 reply
0 recast
0 reaction
