Recently, @stani.lens started a discussion about how to best provide context to posts, especially when they are controversial (link to post in comments). It's a very important topic, so here are my 2 cents. There's a single poll question at the end, please cast your vote after reading this 👇.
There's a tested and open source model for that, as @0xmoe.lens , @prayag.lens , me and few others pointed out - Twitter Community Notes. It works well for Twitter, but can it work for Lens? Let's dig in.
🐦Twitter Community Notes work in the following way, roughly speaking. There's a curated but broad group of members who write notes and vote for them. Votes are given weights, based on member's voting history, then counted, and notes that pass a defined threshold are shown. There's more to it, but essentially, votes that do not correlate with your prior voting pattern get more weight. It is incredibly effective in selecting high quality notes. The model makes sense for many reasons and becomes even more robust with time and scale, which is quite incredible.
💭However, there's more context when it comes to Lens ecosystem. The main premise of decentralized social is that users are in control. Any kind of censorship, with good or bad intentions, will undermine community trust. Because of that, for #LensNotes (here it is, I coined another one 😏) to work, the following must be true:
➡️ The criteria to become #LensNotes contributor must be transparent and unopinionated. Ideally, apart from filtering out bots and activity threshold, Anybody should be able to vote. This means that model weights would have to be tuned accordingly, given wider participation.
➡️ Statistical model itself, including all the weights, must also be transparent and unopinionated. Ideally, there should be a clearly defined process of tuning it (probably via LIPs).
➡️ Voting data must be open and preferably on-chain. This will give a mass of data to analyze and allow public tuning of the model, if necessary.
➡️ Notes cannot influence the visibility of posts in any way. As apps control their own feed algorithms, this means that feed algos themselves would have to be transparent and verifiable. The latter is a big one, but I believe we will get there.
➡️ Users can turn it on/off in settings.
If ALL of the above is true, I believe #LensNotes could blend-in to the Lens ecosystem and become a valuable tool. Fail a single measure, and there are potential problems down the road.
Bonus thought 🤔. If #LensNotes are supported on protocol level, then every app will have an easy path to approach it, given the guidelines are established. However, if one of the apps implements singlehandedly before protocol level support, the situation could get really interesting. #LensWars 🚀 would get a completely new dimension (another link in comments).
Thanks for reading 🙏. Please consider to:
🤞Tag @stani.lens @lensprotocol.lens and others on Lens team to make sure they definitely see this.
🤞Like, collect and mirror the post.
🤞Answer the poll bellow. It will work as a community temperature check.
Does Lens protocol need Twitter style Community Notes? If yes, what version?