Hacker News Reader

Comment by curiousllama | Hacker News Reader

Seems like a great reason to test, gather data, and improve in NYC, no?

Sounds like a great reason to train the model on a closed road with a trained (and consenting) driver playing the role of "lane-blocking-driver" (in that example).

But that would take about a billion years, with all the variations and edge cases.

It would take the same amount of time if they're on a closed course or in a city. The difference is you'd need to pay a lot more people to work on it.

Or you could have a large fleet of a million plus instrumented vehicles gathering data as they drive, then you could use that real world data to analyze and learn from all sorts of situations.

Yeah, and the real world has advantages as the response of human drivers is likely more realistic.

I'd bet they use this data to feed their simulators.

They also train in simulation, where they modify real scenarios to run through variations and edge cases. [1]

As of last year, they'd driven 15 billion miles in simulation - which is more than a typical person will drive in a million years. Still a ways to go to get to a billion though.

[1] https://blog.waymo.com/2020/04/off-road-but-not-offline--sim...

Interesting. What is the validity of the simulations? Are they randomized? Based on actual driving events?

This isn't Tesla we're talking about, Waymo has a large testing facility where they do exactly that sort of thing. Here's an indepth report on it from 4 years ago https://www.theatlantic.com/technology/archive/2017/08/insid...

Interesting, thank you

Sounds more like a great reason to stay away from some of the most chaotic traffic in the US until it can be considered safer. There are far, far more pedestrians in NYC than in Phoenix, and I do not trust a Waymo car to not hit some of them.

There are people behind the wheel.

In the article (only a few paragraphs btw) it specifically says the goal is to map and have safety operators manually controlling all the vehicles.

When I got my NYC drivers license, the full end to end test was 8 minutes (I got a printout) on empty roads.

I’d take a high quality autonomous vehicle over a tired Uber driver any day as a pedestrian

Well I think "high quality" is doing a lot there. Obviously autonomous vehicles would be safer if they work perfectly -- the concern is that they will not.

Yeah, approximately 10 minutes for me to pass driving test (from second attempt). And then around a year of driving (effectively training myself) in city to get more or less comfortable with traffic.

I trust a Waymo car way-more than Tesla's autopilot...

And as evidence for that, you use ... an example of Waymo being overly cautious?

I routinely drive and walk near these vehicles, they do not seem in danger of hitting anyone.

Generally, self driving cars are very safe, too safe actually. There are some high profile accidents, most of them involving Tesla because of the scale and way they approach the problem, plus an infamous one by Uber, but generally, self driving cars will not go anywhere unless they are sure the way is clear.

And this is the problem, self driving cars don't force their way, and they break when anything is vaguely in a collision course. If there is a problem putting them in NYC traffic is that they will stop all the time and create traffic jams, maybe get rear ended by other (human) drivers and fail to get their passengers to the destination because of some minor obstruction.

I can't wait to see one and punch the hood and yell "I'm walkin' here!"

I don't know about New York but I live in a major city and I wonder what will happen when the homeless population figures out they can mess with them.

What is this population currently doing with cars operated by humans and how would it be any different?

In my city, mostly throwing glass bottles at vehicles and screaming.

It is not uncommon to see someone cut across the road and scream at people in cars or throw things.

Depending on the driver lots of different stuff happens. Some will drive around them, others get out and confront them etc.

I suspect an automated vehicle will just come to a stop an wait.

Worst case, add one more person armed with a crowbar and then we have the smash and grab thieves that seem to rampant in San Francisco with a very compliant automated driver stuck behind the accomplice just standing in front of the vehicle.

Envisioning a "keep summer safe" approach…

"Mess" with them? WDYM?

I live in SF by the way.

But to what end?

Gathering more data on a problem you don't know how to solve won't give you a solution to solving it. We don't have any evidence to support that current AI is capable of the advanced problem solving the city driving requires.

Reply on news.ycombinator.com