3 ms·
The difference, of course, is that an airbag can't take control of a car and run someone over. More to the point, the report notes that new methods to determin
by YeGoblynQueenne 9y ago
The difference, of course, is that an airbag can't take control of a car and run someone over.
More to the point, the report notes that new methods to determine the safety of self-driving cars are required.
Which the industry is not exactly falling head over heels trying to develop.
- dpark 9y agoThe report also says that these hypothetical new methods may not be able to prove safety. It’s not a straightforward problem. How do you prove that you’ve reduced (or at least not increased) a problem that occurrs so infrequently? Realistically no one will trust “new methods” and establishing their relevance is really difficult. I would imagine that most of these companies are running lots of simulations, because why wouldn’t you? But how many people will see that and trust it more than data gathered on the road?
- YeGoblynQueenne 9y agoIndeed, there are very few reasons to trust simulations to tell us anything about safety in the real world.