Our top priority this week is fixing spam in popular channels. The core problem is: 

- Channels with large distribution attract casters
- Some casters post spam or even just slightly off topic stuff
- Moderating it is exhausting and thankless

We want to find a solution that can be decentralized across  clients.

Technowatermelon. Elder Millenial. Building Farcaster. 

nf.td/varun

Moderation is hard. Automation tools are key, and LLMs are surprisingly useful to define and apply policies (might be a good lever here)

Building https://srcbook.com // creator of the https://blockchainsmokers.xyz

Indeed. Especially when we can train some models on prior moderators actions (which we'd capture onchain).

So combining human and ai moderation would continually improve the moderation per channel(!).