0 reply
20 recasts
20 reactions
For real bots and "uninteresting" accounts, the goal is to reduce users' exposure to these accounts while highlighting genuinely interesting ones.
One approach is using social reputation systems like social capital scores/ranks or openrank, though this can make it difficult for new users to be discovered. A common solution is to rely on recommendation algorithms that give all content initial exposure and judge its quality based on user preferences, customizing content delivery accordingly. Adding a downvote button, similar to Reddit, could help identify low-quality accounts. These user interactions, combined with manual tagging, could serve as training data for LLMs to determine low-quality content.
Different methods have their merits and drawbacks, making it hard to choose the best one. Warpcast alone might struggle with this, so third-party clients could experiment with different approaches. Effective methods could then be widely adopted, benefiting the entire Farcaster ecosystem. 3 replies
0 recast
44 reactions
1 reply
0 recast
1 reaction
