Now that computer scientists have discovered that the statistics and probability research of the past few decades completely applies to any interesting problems, there should be at least a few more good years left.

Regardless, computer science would never actually die. It would just deadlock or attempt to solve the halting problem.