Don Marzetta
@donmarzetta.bsky.social
Gmail’s spam filter doesn’t fail because of weak AI. It fails because Google hides uncertainty. Legit emails from contacts get buried in Spam, forcing users to scan thousands of junk messages. That’s bad human-in-the-loop design, not “safety.”
January 15, 2026 at 4:18 PM
Gmail’s spam filter doesn’t fail because of weak AI. It fails because Google hides uncertainty. Legit emails from contacts get buried in Spam, forcing users to scan thousands of junk messages. That’s bad human-in-the-loop design, not “safety.”
Gmail’s spam filter doesn’t fail because of weak AI. It fails because Google hides uncertainty. Legit emails from contacts get buried in Spam, forcing users to scan thousands of junk messages. That’s bad human-in-the-loop design, not “safety.”
January 15, 2026 at 4:17 PM
Gmail’s spam filter doesn’t fail because of weak AI. It fails because Google hides uncertainty. Legit emails from contacts get buried in Spam, forcing users to scan thousands of junk messages. That’s bad human-in-the-loop design, not “safety.”
Gmail’s spam filter doesn’t fail because of weak AI. It fails because Google hides uncertainty. Legit emails from contacts get buried in Spam, forcing users to scan thousands of junk messages. That’s bad human-in-the-loop design, not “safety.”
January 15, 2026 at 4:17 PM
Gmail’s spam filter doesn’t fail because of weak AI. It fails because Google hides uncertainty. Legit emails from contacts get buried in Spam, forcing users to scan thousands of junk messages. That’s bad human-in-the-loop design, not “safety.”
Gmail’s spam filter doesn’t fail because of weak AI. It fails because Google hides uncertainty. Legit emails from contacts get buried in Spam, forcing users to scan thousands of junk messages. That’s bad human-in-the-loop design, not “safety.”
January 15, 2026 at 4:16 PM
Gmail’s spam filter doesn’t fail because of weak AI. It fails because Google hides uncertainty. Legit emails from contacts get buried in Spam, forcing users to scan thousands of junk messages. That’s bad human-in-the-loop design, not “safety.”
Gmail’s spam filter doesn’t fail because of weak AI. It fails because Google hides uncertainty. Legit emails from contacts get buried in Spam, forcing users to scan thousands of junk messages. That’s bad human-in-the-loop design, not “safety.”
January 15, 2026 at 4:14 PM
Gmail’s spam filter doesn’t fail because of weak AI. It fails because Google hides uncertainty. Legit emails from contacts get buried in Spam, forcing users to scan thousands of junk messages. That’s bad human-in-the-loop design, not “safety.”
Gmail’s spam filter doesn’t fail because of weak AI. It fails because Google hides uncertainty. Legit emails from contacts get buried in Spam, forcing users to scan thousands of junk messages. That’s bad human-in-the-loop design, not “safety.”
January 15, 2026 at 4:13 PM
Gmail’s spam filter doesn’t fail because of weak AI. It fails because Google hides uncertainty. Legit emails from contacts get buried in Spam, forcing users to scan thousands of junk messages. That’s bad human-in-the-loop design, not “safety.”
Gmail’s spam filter doesn’t fail because of weak AI. It fails because Google hides uncertainty. Legit emails from contacts get buried in Spam, forcing users to scan thousands of junk messages. That’s bad human-in-the-loop design, not “safety.”
January 15, 2026 at 4:13 PM
Gmail’s spam filter doesn’t fail because of weak AI. It fails because Google hides uncertainty. Legit emails from contacts get buried in Spam, forcing users to scan thousands of junk messages. That’s bad human-in-the-loop design, not “safety.”
Gmail’s spam filter doesn’t fail because of weak AI. It fails because Google hides uncertainty. Legit emails from contacts get buried in Spam, forcing users to scan thousands of junk messages. That’s bad human-in-the-loop design, not “safety.”
January 15, 2026 at 4:11 PM
Gmail’s spam filter doesn’t fail because of weak AI. It fails because Google hides uncertainty. Legit emails from contacts get buried in Spam, forcing users to scan thousands of junk messages. That’s bad human-in-the-loop design, not “safety.”
Gmail’s spam filter doesn’t fail because of weak AI. It fails because Google hides uncertainty. Legit emails from contacts get buried in Spam, forcing users to scan thousands of junk messages. That’s bad human-in-the-loop design, not “safety.”
January 15, 2026 at 4:10 PM
Gmail’s spam filter doesn’t fail because of weak AI. It fails because Google hides uncertainty. Legit emails from contacts get buried in Spam, forcing users to scan thousands of junk messages. That’s bad human-in-the-loop design, not “safety.”
Gmail’s spam filter doesn’t fail because of weak AI. It fails because Google hides uncertainty. Legit emails from contacts get buried in Spam, forcing users to scan thousands of junk messages. That’s bad human-in-the-loop design, not “safety.”
January 15, 2026 at 4:09 PM
Gmail’s spam filter doesn’t fail because of weak AI. It fails because Google hides uncertainty. Legit emails from contacts get buried in Spam, forcing users to scan thousands of junk messages. That’s bad human-in-the-loop design, not “safety.”