Clumping Solstice Singalongs in Groups of 2-4 — LessWrong