Where did he said that? A new experiment can be done, responsibly, right?

The difficulty in getting to an agreement on what "responsible" means, and on how a "failure" should be handled means that said experiment will never occur. Nothing is ever safe enough and no failure is acceptable -- so no test which can fail can be run.

A situation which would be in obvious conflict (negative expected value) with the goal of saving the man hours and lives currently being cost by manual driving.