Hi Michael, Any thoughts on incorporating AI into Spamsieve to enhance categorizing the level of Spamminess? Recently I’m flooded with hundreds of Spam emails that appear to be written by AI to use words that don’t normally trigger being flagged as spam.
I’d like to delete and empty trash everything that is high probability of being Spam so I don’t have to scroll down through it all. It’s hundreds to thousands per day for my systems. Was hoping maybe AI could help identify known AI created Spam.
SpamSieve is already using AI (but not LLMs, if that’s what you’re referring to). Are the messages already being correctly categorized as spam but you disagree with SpamSieve about the level of spamminess?
Yes, SpamSieve is catching 99% but level of spamminess seems kinda random. It could be since I’ve not retrained in a while. My goal currently is to never see the “100% sure it’s spam” by deleting not just move to trash by using a mail rule. I don’t want to throw shade on SpamSieve at all since it does a great job. I’m just trying to avoid that daily look through the hundreds for false positives.
I completely agree with your request.
I’ve set the SS score to 80; in the filter tab, all options are checked; in the Apple Mail tab, I move messages to the trash if the score is greater than 80.
Then, when the trash contains recurring spam on a specific topic, I add an automatic deletion rule. It’s imperfect and tedious, but I haven’t found a better solution.
It would be helpful to have an additional feature in SS with a list of keywords (based on AI analysis) that would trigger automatic deletion.