Since it works over any kinded value this means that in the Future when we have integrations like Spark Data sets etc which are also kinded and can provide instances for those type classes you will be able to use the same DSL to reduce a list locally than you use to reduce a distributed remote data set.