I haven’t watched the video and I can’t watch it, I feel ill just thinking about it. They hunted him down like he was an animal, without so much as even an afterthought. It made me think of every time I had police unexpectedly appear out of nowhere and just how nervous and afraid I was. The last time was while I was…

I knew him too doing mics around NYC and he was always very kind and supportive. No where on earth will you find more crabby people than in comedy, and everyone has a negative story about everyone else.

Take note of where this vehicle was damaged:

It saddens me that on Jalopnik, a site supposedly dedicated to car enthusiasts and aficionados, that this comment has more stars than the comment to which you are replying.

The “20ft” you are using is hogwash. You can see the woman’s shoes in the video quite clearly just as the time turns 5 seconds, she is hit sometime after 6 seconds. 1 second at 45 mph is 66 feet. This woman was visible in the crummy color night camera 66 feet in front of the car and it still hit her. Just for funsies…

I would hope the sensors don’t rely on vision to detect obstacles. Does it have to be a sunny f**king day for this crap to work?

Problem is the Uber car never even attempt to brake. If it jumped on the brake and still couldn’t stop in time, that’s understandable. But the car never even tried to brake.

What you’re seeing in the video is NOT a true representation of what is visible with the human eye.

Dash cams have utterly terrible low light capability.

I’ve slowed for and avoided many small animals at night that move faster than this large individual and bicycle.

These cars are supposed to be able to see in the dark, or so I thought. If they don’t, that’s a MAJOR strike against the idea that they are better than human drivers*

The sensors have multispectral cameras (near-infra-red and visible light), lidar and radar. They would have seen her heat signature, the radar would have had a bounceback and the lidar would have mapped her. This was a failure of every single system.

Could it have completely avoided an accident? Maybe not, but by all accounts it appears to have hit her at full speed. Even a human watching that road would have done some level of breaking before hitting her. That didn’t happen because the guy did what most people do when given semi-autonomous cars, which is stop…

If it can’t detect beyond 20ft at 45mph, it shouldn’t be going 45mph 🤷🏻‍♂️

Am I the only one who thinks this EXACT scenario should be test number fucking one? If the system can’t handle this....

The video and your eyes are two different things. Anyone who uses a dashcam at night and reviews footage would know that your eyes see much better then the dashcam.

And yet, they’re starting to roll out buses without drivers, steering wheels or brake pedals. While the tech has yet to be proven. And soon they’ll want to move to big rigs as well.

the other issue in addition to this is what looks like an over-reliance on the system—it’s not just that he car itself wasn’t sufficient, it was that the person was putting more trust in it than it deserved, something we’re seeing time and time again with this tech.

Car is equipped with LIDAR and multiple other night and day sensors to detect exactly that sort of thing. Those sensors are absurdly expensive and if they couldn’t detect a person in the middle of the road, then what is the point of those sensors at all? At first I was defending the tech, but from this video it was…

She appears “invisible” thanks to the video’s poor resolution. In real life, the safety operator would have been able to see her a couple of seconds sooner (if he had been watching).

She’s invisible to us on a video recording. She shouldn’t be invisible to the sensors on an autonomous vehicle.

It’s supposed to have LIDAR or similar sensors which detect obstacles in front of it - whether or not they’re illuminated. Part of the point of these autonomous vehicles is that they’re supposed to have better ways to detect potential problems than a human driver, and this shows that wasn’t the case.