Searle's argument depends on our intuition of the distinction between syntax and semantics, but this distinction seems very outdated to me. It's an easy intuition to have in 1980, but its hard to defend in 2018. It's true that syntax itself isn't enough for semantics, but syntax is a component of a system that recognizes semantics, i.e. low level rule based operations that in aggregate entail semantics. So to say that because computers operate based on syntactic structures means the computer can never understand semantics is just a mistake on many levels.