Alternately shamelessly repurposing the implementation (uniq_list/3) of Enum.uniq/1:

defmodule Demo do
  def duplicates?(list),
    do: duplicates?(list, %{})

  defp duplicates?([], _) do
    false
  end

  defp duplicates?([head|tail], set) do
    case set do
      %{^head => true} ->
        true
      _ ->
        duplicates?(tail, Map.put(set, head, true))
    end
  end

  def run(list) do
    list
    |> duplicates?()
    |> inspect()
    |> IO.puts()
  end
end

list = [1, 2, 3, 3, 2, 1]
uniq_list = Enum.uniq(list)
Demo.run(list)      # true
Demo.run(uniq_list) # false

A lot can be learned by scouring through Elixir’s source code. The important thing is to understand how it works.

But reducing and finding the elements in a list again results in a runtime of O(n log n), using a MapSet is O(n).

Maybe something like this:

def has_duplicates?(list), do: Enum.uniq(list) != list

def has_duplicates?(list) do
  list
  |> Enum.reduce_while([], fn x, acc ->
    if x in acc, do: {:halt, 0}, else: {:cont, [x | acc]}
  end)
  |> is_integer()
end

@l00ker in your case list will be iterated twice - first to make uniq and second to compare with original list