As I mentioned earlier, we have predicted that with full deployment we can save 80% of collisions and deaths, but not 100%. The technology is not perfectly safe. It never will be—I'm an engineer—but what I feel is that if we delay the deployment of AVs too much, we will delay the benefits of saving those lives.
There are two sides of the coin here. One of the issues is that the public and most other people in the know have expectations that are too high. The technology is going to be much better than humans, but never perfect. If we delay too much, we will in fact allow those accidents and collisions to happen. Those lives might otherwise be saved.