Spark SQL generic dataset reading

Question

I would like to create a generic trait for loading Datasets:

case class Foo(name: String)

trait Loader[T] {
  def load(implicit spark: SparkSession): Dataset[T]