YES

Problem:
 first(0(),X) -> nil()
 first(s(X),cons(Y,Z)) -> cons(Y,n__first(X,activate(Z)))
 from(X) -> cons(X,n__from(n__s(X)))
 first(X1,X2) -> n__first(X1,X2)
 from(X) -> n__from(X)
 s(X) -> n__s(X)
 activate(n__first(X1,X2)) -> first(activate(X1),activate(X2))
 activate(n__from(X)) -> from(activate(X))
 activate(n__s(X)) -> s(activate(X))
 activate(X) -> X

Proof:
 WPO Processor:
  algebra: Max
  weight function:
   w0 = 0
   w(0) = 1
   w(n__from) = w(n__s) = w(from) = w(n__first) = w(activate) = w(cons) = w(
   s) = w(nil) = w(first) = 0
  status function:
   st(n__from) = st(n__s) = st(from) = [0]
   st(n__first) = [0, 1]
   st(activate) = [0]
   st(cons) = [0, 1]
   st(s) = [0]
   st(nil) = []
   st(first) = [0, 1]
   st(0) = []
  subterm penalty function:
   sp(n__first, 1) = sp(first, 1) = 2
   sp(n__from, 0) = sp(n__s, 0) = sp(from, 0) = sp(n__first, 0) = sp(
   activate, 0) = sp(cons, 1) = sp(cons, 0) = sp(s, 0) = sp(first, 0) = 0
  precedence:
   activate > s > from ~ first > n__from ~ n__s ~ n__first ~ cons ~ nil ~ 0
  problem:
   
  Qed