“I can assure you that no string theorist would be interested in working on string theory if it were somehow permanently beyond testability. That would no longer be doing science.”