To　meet　the　requirements　of　ubiquitous　connectivity,　the　sixth-generation　is　expected　to　support　higher　frequencies　such　as　millimeter　waves　and　terahertz,　as　well　as　high-mobility　scenarios　like　unmanned　aerial　vehicles　and　vehicular　networks.　However,　communication　scenarios　under　these　conditions　may　suffer　from　beam　misalignment　due　to　high　mobility,　especially　in　high-frequency　channels　with　narrow　beam　characteristics.　Integrated　sensing　and　communications　(ISAC)　enable　hardware　and　spectrum　resource　sharing　between　radar　sensing　and　wireless　communication,　and　exhibit　great　potential　in　sensing-assisted　communication.　One　pressing　problem　is　that　performance　metrics　for　sensing　and　communication　are　often　contradictory　when　they　share　the　same　hardware　or　wireless　resources.　Therefore,　how　to　optimize　the　performance　metrics　between　both　is　the　key　to　fulfilling　specific　requirements.　This　article　focuses　on　ISAC　enabled　predictive　beamforming,　and　investigates　the　technical　characteristics,　system　architecture,　and　optimization　schemes　involved　in　the　performance　trade-offs.　We　first　provide　an　overview　of　the　unique　characteristics　of　ISAC　and　predictive　beamforming　and　highlight　the　benefits　of　combining　them.　We　then　investigate　the　design　architectures　and　propose　an　optimization　scheme　that　can　achieve　a　performance　trade-off　between　sensing　and　communication　to　maximize　throughput.　Finally,　we　highlight　some　challenging　issues　that　need　to　be　addressed　in　practical　implementations.　IEEE