Numerous　emerging　applications　in　modern　society　require　humidity　sensors　that　are　not　only　sensitive　and　specific　but　also　durable　and　intelligent.　However,　conventional　humidity　sensors　do　not　have　all　of　these　simultaneously　because　they　require　very　different　or　even　contradictory　design　principles.　Here,　inspired　by　camel　noses,　we　develop　a　porous　zwitterionic　capacitive　humidity　sensor.　Relying　on　the　synergistic　effect　of　a　porous　structure　and　good　chemical　and　thermal　stabilities　of　hygroscopic　zwitterions,　this　sensor　simultaneously　exhibits　high　sensitivity,　discriminability,　excellent　durability,　and,　in　particular,　the　highest　respond　speed　among　reported　capacitive　humidity　sensors,　with　demonstrated　applications　in　the　fast　discrimination　between　fresh,　stale,　and　dry　leaves,　high-resolution　touchless　human-machine　interactive　input　devices,　and　the　real-time　monitoring　humidity　level　of　a　hot　industrial　exhaust.　More　importantly,　this　sensor　exhibits　typical　synapse　behaviors　such　as　paired-pulse　facilitation　due　to　the　strong　binding　interactions　between　water　and　zwitterions.　This　leads　to　learning　and　forgetting　features　with　a　tunable　memory,　thus　giving　the　sensor　artificial　intelligence　and　enabling　the　location　of　water　sources.　This　work　offers　a　general　design　principle　expected　to　be　applied　to　develop　other　high-performance　biochemical　sensors　and　the　next-generation　intelligent　sensors　with　much　broader　applications.　©