Welcoming our first riders in San Francisco(blog.waymo.com)
blog.waymo.com
Welcoming our first riders in San Francisco
https://blog.waymo.com/2021/08/welcoming-our-first-riders-in-san.html
669 comments
Waymo is currently the only SAE Level 4[0] self-driving car in use, right? The wheel must not be used. Even though it is limited to two cities.
Their technology seems an order of magnitude above the rest, although factoring in its cost could make it look worse. (Can it turn a profit, when including the development costs?)
Tesla FSD is at best SAE level 3 (can need human fallback), and at worst level 2 (needs constant human monitoring).
[0]: https://blog.waymo.com/2020/10/revealing-our-approach-to-saf...
Their technology seems an order of magnitude above the rest, although factoring in its cost could make it look worse. (Can it turn a profit, when including the development costs?)
Tesla FSD is at best SAE level 3 (can need human fallback), and at worst level 2 (needs constant human monitoring).
[0]: https://blog.waymo.com/2020/10/revealing-our-approach-to-saf...
It's fantastic to see Waymo's progress. SF is a real nightmare to drive. If they can nail it there, that's 2/2 for busy, urban street driving (SF) and "boring" suburban driving (Chandler, AZ). They've been quietly very confident of their tech, but this is a real test.
Tangentially, I've also noticed that Waymo has picked up pace ever since the recent leadership changes. They are publishing more blog posts, offering more insights into their tech and generally seem to have increased their PR game. I wonder if that was a mandate from Alphabet leadership to show some urgency.
Tangentially, I've also noticed that Waymo has picked up pace ever since the recent leadership changes. They are publishing more blog posts, offering more insights into their tech and generally seem to have increased their PR game. I wonder if that was a mandate from Alphabet leadership to show some urgency.
As someone who lives in Bath, an old european city with roads (and other drivers) that can give experienced human drivers a nervous breakdown I'm rather more interested in how they do in SF than their current deployment in Phoenix.
From what I've heard repeatedly, SF sounds much more irregular and messy than Phoenix, so it should be something of a stepping stone to making them usable more widely if they can crack it.
I've been expecting it to happen for literally decades, but always been disappointed.
From what I've heard repeatedly, SF sounds much more irregular and messy than Phoenix, so it should be something of a stepping stone to making them usable more widely if they can crack it.
I've been expecting it to happen for literally decades, but always been disappointed.
Kind of interesting to consider how this adjusts priors on these outcomes:
- Self driving is not possible
- Self driving is not possible anytime soon
- Self driving is possible, but requires LIDAR
- Self driving is possible, and can be done with normal cameras
I've engaged with a lot of people who presume self-driving reduces onto AGI, therefore it will not be achieved anytime soon if ever. I wonder which of their assumptions is wrong, if this ends up being successful.
- Self driving is not possible
- Self driving is not possible anytime soon
- Self driving is possible, but requires LIDAR
- Self driving is possible, and can be done with normal cameras
I've engaged with a lot of people who presume self-driving reduces onto AGI, therefore it will not be achieved anytime soon if ever. I wonder which of their assumptions is wrong, if this ends up being successful.
Disclosure: I work at Waymo.
This is big news, and I encourage you to apply if you're in San Francisco. For the HN crowd, I'd also recommend last week's blog post [1] which includes some more technical material on "how" we're driving.
[1] https://blog.waymo.com/2021/08/MostExperiencedUrbanDriver.ht...
This is big news, and I encourage you to apply if you're in San Francisco. For the HN crowd, I'd also recommend last week's blog post [1] which includes some more technical material on "how" we're driving.
[1] https://blog.waymo.com/2021/08/MostExperiencedUrbanDriver.ht...
I'll never forget I almost got murdered terminator style by a Cruise while riding my bicycle around SF... hopefully they figured that out.
I just signed up and actually skimmed the privacy policy (because hey, it's Alphabet). They record video of you during the ride and didn't mention anything that I caught about ever deleting it. I get that they need video in case I do bad stuff in the back, but it's really disappointing that they didn't bother reassuring you like "we delete it after a month" or something.
I'm hoping I just missed that line. But I signed up anyway, so I guess that says more.
I'm hoping I just missed that line. But I signed up anyway, so I guess that says more.
Kudos to the Waymo team.
I was a self driving skeptic 2-3 years back. Now given the advances in both hardware and NNs, I do see possible solutions in the next decade.
I was a self driving skeptic 2-3 years back. Now given the advances in both hardware and NNs, I do see possible solutions in the next decade.
> We can’t wait to hear from more San Franciscans as they experience the Waymo Driver themselves. Beginning later today, San Franciscans can sign up for the Trusted Tester program and help us shape the future of mobility in this city. Just download the Waymo One app to get involved.
Anyone else got this to work? I only see a message that says "It looks like you're not in our service area..."
Anyone else got this to work? I only see a message that says "It looks like you're not in our service area..."
I feel like people here are missing the other factor that will drive adoption of self driving vehicles: economics.
Safety is obviously important, but we already accept a decrease in safety over money and convenience (otherwise, we would have a 30kph speed limit everywhere).
But most people use their car for commuting, groceries, dropping the kids to sports, etc. They may make one or two road trips a year for vacations or business.
To facilitate this, a middle class family will often own two vehicles, pay tax, insurance and maintenance on said vehicles, pay for parking, etc.
If that suddenly wasn't necessary, that would be a massive saving for most people. You no longer need the capital investment in something that loses value.
Even better, you don't need to bring little Suzy to football, you can put her in a waymo instead (once she's old enough obviously).
And when you want to go drive to see grandma, you rent a car for the longer trip.
All of that depends on affordable self driving transport of course.
Safety is obviously important, but we already accept a decrease in safety over money and convenience (otherwise, we would have a 30kph speed limit everywhere).
But most people use their car for commuting, groceries, dropping the kids to sports, etc. They may make one or two road trips a year for vacations or business.
To facilitate this, a middle class family will often own two vehicles, pay tax, insurance and maintenance on said vehicles, pay for parking, etc.
If that suddenly wasn't necessary, that would be a massive saving for most people. You no longer need the capital investment in something that loses value.
Even better, you don't need to bring little Suzy to football, you can put her in a waymo instead (once she's old enough obviously).
And when you want to go drive to see grandma, you rent a car for the longer trip.
All of that depends on affordable self driving transport of course.
Many comments in this thread are variations on a theme of "self-driving cars don't need to be perfect, they just need to be better than human drivers, who aren't actually all that great." I think it would be nice if this were true, and I suppose it is from an actuarial perspective, but it's also an extremely flawed point.
From a public confidence perspective, it doesn't matter if a self-driving car crashes one tenth, even one one-hundredth as often as human drivers.
If you see a self-driving car cause an accident, particularly a lethal one, in a situation that almost any human driver would have avoided, you've totally destroyed any and all confidence in this car's driving ability, because "I would never, ever have crashed there."
As we've seen, there are lots of scenarios like this. The Tesla crash from last year, where the car simply didn't see a white truck against a light background. Or imagine an adversarial image attack, where some tiny insignificant detail is placed onto a stop sign or a "do-not-enter" that turns it into nothing from the perspective of the AI driver.
These kinds of scenarios obliterate public confidence in self-driving cars, because intuitively, you immediately realize that you're "a much better driver" than this car! Even if that's untrue 99/100 times, it only takes one visceral example to drive this kind of wedge.
Self-driving cars don't just have to be better than human drivers. They have to be as close to perfect as is possible, because that's what people will expect.
From a public confidence perspective, it doesn't matter if a self-driving car crashes one tenth, even one one-hundredth as often as human drivers.
If you see a self-driving car cause an accident, particularly a lethal one, in a situation that almost any human driver would have avoided, you've totally destroyed any and all confidence in this car's driving ability, because "I would never, ever have crashed there."
As we've seen, there are lots of scenarios like this. The Tesla crash from last year, where the car simply didn't see a white truck against a light background. Or imagine an adversarial image attack, where some tiny insignificant detail is placed onto a stop sign or a "do-not-enter" that turns it into nothing from the perspective of the AI driver.
These kinds of scenarios obliterate public confidence in self-driving cars, because intuitively, you immediately realize that you're "a much better driver" than this car! Even if that's untrue 99/100 times, it only takes one visceral example to drive this kind of wedge.
Self-driving cars don't just have to be better than human drivers. They have to be as close to perfect as is possible, because that's what people will expect.
What's the biggest technical challenge that they have right now? And how much time and effort will it take to overcome?
I feel like there is a lot of disconnect between the different media pieces and opinions I come across specifically when it comes to Waymo's self-driving - from "we're already there" to "it won't happen in the next 10 years".
I feel like there is a lot of disconnect between the different media pieces and opinions I come across specifically when it comes to Waymo's self-driving - from "we're already there" to "it won't happen in the next 10 years".
Last week Waymo said they are driving 100,000 miles per week in San Francisco. That figure is just bonkers. SF MTA only operates about 450,000 miles per week, and that was before COVID-19 shut them down. When one is on the streets of San Francisco does a Waymo Jaguar just drive by every couple of minutes? I honestly haven't been over there in a few months.
I wonder if the cars will always have a human chaperone. Not as a safety driver, but to prevent riders from trashing the car.
This is both incredibly exciting and a little terrifying. As a pedestrian, I'm still a little weary around these cars.
I'm curious how Waymo deals with the human aspects of cab hailing. With a human driver, you can say "hey looks like traffic is bad between where the car currently is and the pick up spot, can I meet you at X instead and save us both 10 minutes?"
To "good morning dear, do you think you could help me load my bag in the trunk"
All the way up to completely degenerate scenarios like "my friend was drunk and tried to take over command of the car to see if it would work" or "I had to call customer support because when my ride arrived to pick me up, there was a homeless guy hogging the backseat and cursing at me".
Surely there's more to commercializing the tech than just not killing pedestrians. The absence of a driver should create some new weird dynamics.
IIRC Uber et al spend a considerable amount on customer support. I wonder if Waymo can break free from Google's bad reputation on that front.
To "good morning dear, do you think you could help me load my bag in the trunk"
All the way up to completely degenerate scenarios like "my friend was drunk and tried to take over command of the car to see if it would work" or "I had to call customer support because when my ride arrived to pick me up, there was a homeless guy hogging the backseat and cursing at me".
Surely there's more to commercializing the tech than just not killing pedestrians. The absence of a driver should create some new weird dynamics.
IIRC Uber et al spend a considerable amount on customer support. I wonder if Waymo can break free from Google's bad reputation on that front.
Why does Waymo use a niche British luxury car for their platform? Must be very expensive and what are the benefits?
What if self-driving cars develop personalities? Like, let's say to be able to drive, you need x level of intelligence or whatever, and just happens to be the level where consciousness emerges
I'm sure I got all the science wrong but do we really want cars with souls? Do they become non-human humans? Do they get the right to vote? Are they considered slaves?
Even if its just horse-level smarts, that's still enough to develop consciousness, maybe. Let's just say sure. But then, what kinda relationship do you have with your car then?
Another thing is that, now that everyone's working from home, there's a lot less incentive to buy cars. Who are the going to sell these cars too? Maybe they'll all be self-driving taxis. And that's a whole can of proverbial worms on its own. Politically, self-driving taxis are dead ducks.
I'm sure I got all the science wrong but do we really want cars with souls? Do they become non-human humans? Do they get the right to vote? Are they considered slaves?
Even if its just horse-level smarts, that's still enough to develop consciousness, maybe. Let's just say sure. But then, what kinda relationship do you have with your car then?
Another thing is that, now that everyone's working from home, there's a lot less incentive to buy cars. Who are the going to sell these cars too? Maybe they'll all be self-driving taxis. And that's a whole can of proverbial worms on its own. Politically, self-driving taxis are dead ducks.
Given the number of people here saying self-driving cars require AGI (a.k.a. sand that thinks a.k.a. fairy magic), I would say we have definitely hit the bottom of the trough of disillusionment. But since Waymo is launching in a second city, this is perhaps the very start of the slope of enlightenment.
You cannot predict the future with patterns from the past, regardless of how many patterns you crunch - the world is always changing. ML can just never solve FSD - conceptual understanding and causality is needed to achieve human level driving capability. We instantly classify things conceptually from our perception, instantly understand how those things may behave, and then learn the specifics of what that thing is doing to instantly predict what might happen and how it is interacting with other causally connected things in the environment.
If ML could predict the future, there would be some very rich stock brokers right now.
Autonomous driving efforts need a major pivot to start working with concepts, causality, and the integrated space time model of the world we humans use.
If ML could predict the future, there would be some very rich stock brokers right now.
Autonomous driving efforts need a major pivot to start working with concepts, causality, and the integrated space time model of the world we humans use.
I can’t wait for truly autonomous cars. I live in the countryside and have to drive to collect and deliver kids, or abstain as a designated driver and so on. Self driving will change this.
The aspect I dislike is the move towards hiring cars on demand instead of owning them, though.
The aspect I dislike is the move towards hiring cars on demand instead of owning them, though.
Are there any thought pieces on how cities can adapt roads to accommodate SAE level 4 cars to facilitate rolling out a wider deployment more quickly? I'm thinking of something like cities could designate specific routes and lanes for L4 cars with designated drop-off locations and also only allow the service to be operational during clement weather conditions. It seems it would be beneficial to have fully automated routes to and from specific spots even today, for example an airport to a downtown location, for many cities to help reduce transportation costs and safety issues as compared with Uber/Lyft.
I guess all that driving around the avenues in the middle of the night paid off.
Interesting timing given this article which was just posted a day or so ago on HN
https://www.autoblog.com/2021/08/22/waymo-is-99-of-the-way-t...
https://www.autoblog.com/2021/08/22/waymo-is-99-of-the-way-t...
Would getting hit by one of these things basically mean you won the lottery?
I'm surprised the numbers for Phoenix are so low (10's of thousands of rides). I'm not sure if that fleet is just very small or my sense of how many rides/day a major metro can generate.
Does the Waymo One app require a Google account? If Google arbitrarily freezes my Google account, will I be unable to use Waymo and unable to talk to any human about it?
I'll know that we've achieved full self driving when it can brave thru the streets of a South American city :)
This tells me that we’re still a long way from full level 4 (and certainly level 5) autonomy in a busy city like San Francisco. The edge cases requiring immediate human attention are still too frequent for the human safety driver to be remote, as is the case in Phoenix.
Also, just a reminder that Waymo in Phoenix is nowhere close to being level 5, since it is still heavily geofences and requires those remote safety monitors. I still think that true level 5 (i.e. ability to drive autonomously everywhere with zero human oversight with a safety record equivalent to the median human driver) requires AGI. Would love to be proven wrong!