In my experience, the Waymo cars are WAY better than the Cruise ones. The ride is less swervy (hardly at all), and they seem to have very good ‘situational awareness’ of nearby cars and pedestrians. I’ve never felt that the Waymo cars acted ‘sketchy’ when I’ve been in them, but the Cruise cars seem a bit less ‘solid’ with their driving.
I think that Cruise is giving self driving cars a bad reputation in San Francisco. I see them getting stuck WAY more often than Waymos, and I have never seen a Waymo fail to yield to pedestrians.
I’d just like, say, a flashing purple light or something on top of the driverless cars so I know what I’m dealing with. I mean, you can obv already recognize them by the cameras and lidars, but a big fat “Robot Student Driver” sign would warn other street users that a robot was currently at the controls.
When you see the news reports, tho, it’s almost always Cruise cars that are causing trouble.
Also, a street with extremely high traffic and lots of pedestrians after a big parade wandering around and UP TO the self driving cars is not going to be easy for these cars to safely navigate right now. They should probably just steer clear of high traffic areas for the time being.
The Waymos, at least in my experience, are extremely cautious around pedestrians, especially those moving erratically into the street.
Also, I would rather these cars err on the side of caution rather than trying to power through dense fog or a crowd of people?
If we take ZERO risks, we will have ZERO progress. They are safer drivers NOW than many humans on the road. These cars are definitely good enough for testing. And it is exactly this sort of ‘real world’ testing that shows where they still need to improve.
They are not good yet at very unusual traffic situations, such as unexpected/unlisted road closures, construction/emergency workers walking around in the middle of the street, etc. They definitely need to get better about routing them away from such areas until they can negotiate them successfully w/o getting stuck.
But I am hardly concerned about a few stuck cars now for the potential progress of having most of the cars on the road be self driving.
There is taking a risk that leads to progress and there is making things more difficult for everyone else to try out a solution to a problem that never existed in the first place.
If you want less traffic on the road, then that actual solution is more effective public transit, that is either very cheap or free.
Or… more public transit. More heavy and light rail. ban cars in dense urban cores with some exceptions. Build out massive high speed rail nationally. That’s the actual solution we need to this problem. A bunch of self-driving cars on the road is only going to INCREASE traffic. It’s not a solution to the problems we face with traffic in America. At all.
Again, as long as the legal repurcussions for dangerous failure extend to an actual human being responsible for the robotaxis, then great. Until then, who gave them permission to perform dangerous experiments on the public?
But they are brilliant techies breaking thin… I mean solving important problems! They don’t need permission from the plebs!!! /s
But again… let’s build out our public transit that will benefit everyone, not just a few rich assholes in silicon valley. We did it in the 40s and 50s with the interstate system, we can do it with a public transit system, too.
It was never necessary to test these on public roads in order to find out, for example, that they are entirely incapable of following instructions from a human traffic control officer. That’s not some bizarre edge case that nobody could have foreseen in advance.
“Self-driving” shouldn’t be the goal. “Safer, more efficient transportation” should be the goal. We shouldn’t start from the assumption that self-driving vehicles are critical to, or even necessarily compatible with, reaching that goal.
Self driving isn’t the goal: destroying public transport and externalising costs for the private system that will replace it is. Not paying workers is the goal of pretty much all “AI” these days because it’s an obsession of capital.
