Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

NY Times Rides in Waymo Robotaxis in SF

nytimes.com

Post reply on HN
Print view View original

81–90 of 348 posts

Re: NY Times Rides in Waymo Robotaxis in SF

#81

Post by pengaru »

abhayhegde wrote: post #57

> Though there’s no driver, each ride is supported by staff at a Waymo site that can be summoned if a car runs into trouble. Unless they have multiple rides looked after by a staff, what are they planning to save by not offering a driver? Also, is there any path to profitability given how many billions were poured into the technology? Not that I care so much though, Alphabet has enough profitable ventures to recoup o…

I think it goes without saying that the remote observer likely sits at something resembling a trading desk with a wall of LCD panels, monitoring heaps of vehicles like a glorified security guard.

It'd be interesting to know what the permissible overcommit ratio is though, if there even is a limit codified by the state/city somewhere.

Re: NY Times Rides in Waymo Robotaxis in SF

#82

Post by paul7986 »

bastawhiz wrote: post #73
paul7986 wrote: post #35

Earlier quoted context omitted.

Sure yet a human driver is not a device that can be hacked and then used as a weapon or weapons.

Yeah it's a good thing that cities aren't filled with retail locations that sell liquids that, if ingested, turn humans into bad drivers that think they're great drivers. And it's a good thing that those hypothetical liquids don't form debilitating addictions that make the humans want them more.

hmmm your analogy doesn't totally fit and you might have drank too much of the AI tech bro kool aid ;-)

These cars can be hacked as used as weapons ... for terrorist attacks. As well malfunction and lock you in driving into the bay or off a cliff or etc.

Open to your rebuttal ;-)

Re: NY Times Rides in Waymo Robotaxis in SF

#83

Post by AlotOfReading »

unyttigfjelltol wrote: post #48
ChatGTP wrote: post #36

Earlier quoted context omitted.

It's a ride in a car, it's boring, even if it's driven by GPS and Lidar.

It's a kind of Rube Goldberg contraption for getting around town..... The most complex gizmo possible for the most mundane of tasks.

Moravec's paradox. The things humans do easily end up being the hardest problems for machines.

Besides, you could call a smartphone a complicated rube Goldberg machine for the purpose of reading text when you could just buy a newspaper instead, but we don't because they're meaningfully different things.

Re: NY Times Rides in Waymo Robotaxis in SF

#84

Post by macintux »

dekhn wrote: post #50
macintux wrote: post #45

Earlier quoted context omitted.

Cars are dramatically safer for passengers than they used to be. Miles per death isn’t reflective of the skill/attentiveness levels of drivers. As the sibling comment points out, miles driven per death for pedestrians isn’t necessarily being improved by the mass and visibility and safety factors.

I believe my numbers include pedestrians, based on the definitions I can see.

Here’s a detailed breakdown (although no deaths/mile figures).

Proportional to population, pedestrian fatalities have risen significantly since 2010.

https://www.iihs.org/topics/fatality-statistics/detail/pedes...

Re: NY Times Rides in Waymo Robotaxis in SF

#85

Post by Dylan16807 »

proc0 wrote: post #76
_acco wrote: post #14

> My ride was so smooth, the novelty began to wear off, turning a trip to the future into just another journey across town. This is the only nod in the article to how futuristic and revelatory this innovation. The journalists sound so bored. I wonder if this is a function of how “around the corner” this tech has felt for so long? Have we somehow become jaded to this innovation before it has even arrived? Were journal…

There's a term for this, or at least it's a known phenomena in the AI industry for a while. It's how AI becomes transparent after wide adoption. I'm still amazed at video filters that transform your face in real time, and that is now a common feature in so many apps that most people might not consider AI.

> It's how AI becomes transparent after wide adoption.

I don't like this framing for video filters.

Go back ten years and nobody thought of that kind of visual effect as AI. Then someone made an AI version, and people talked about that for a bit, then they went back to considering the actual effect.

This is very different from situations where a task is inherently thought of as an AI task when people are anticipating it.

Re: NY Times Rides in Waymo Robotaxis in SF

#86

Post by xnx »

barbariangrunge wrote: post #21

But… why couldn’t we just build more and better trains?

Self driving cars are effectively "more and better trains". They also go wherever you want whenever you want, don't subject you to other passengers, and are less expensive per passenger mile.

Re: NY Times Rides in Waymo Robotaxis in SF

#87

Post by dekhn »

macintux wrote: post #84
dekhn wrote: post #50

Earlier quoted context omitted.

I believe my numbers include pedestrians, based on the definitions I can see.

Here’s a detailed breakdown (although no deaths/mile figures). Proportional to population, pedestrian fatalities have risen significantly since 2010. https://www.iihs.org/topics/fatality-statistics/detail/pedes...

It has deaths/mile, in that they provide deaths, and we know the miles.

I do not disagree the numbers are trending up (although, I believe rates/miles driven might show a different story, but I didn't grab the miles driven for years 1975-2021 to normalize).

Americans drive 400 MILLION miles before a pedestrian death.

Let that sink in. 400 million miles per ped death. That's just absolutely gob-smacking amazingly good for something involving direct human control of an enormous ton+ vehicle.

Re: NY Times Rides in Waymo Robotaxis in SF

#88

Post by olyjohn »

quickthrower2 wrote: post #23

A concept I look forward to is not just a driverless car, but driverless "various size vehicles" from a 1-person car to a 100-person bus. In this future we could get the advantages of public transport and car-free living into rural and suburban areas, while in the city these systems could adjust and use AI to find optimal routes that are a lot less slower than current buses (or even trains that require 2 or 3 connect…

Why does having no driver suddenly enable this?

You can do all this now, you just need a driver. Why is that the part that needs to be eliminated before you can do all these things? All it is, is route planning based on demand. Uber, etc should be able to do all this now. Isn't that what Uber Pool was supposed to deliver? Nobody uses that shit for anything but a commute. Nobody wants to sit in a car with other random-ass people stopping 30 times to pick them up. That's half the reason people who drive don't take transit now.

What is it about having no driver that suddenly makes cars green, free up space on the road, etc?

Why would someone who owns a car now, want to wait around for a driverless taxi to pick them up? They can go into their driveway and be driven to their destination right away in their own personal driverless car. And if the car is gonna park itself, that means it's driving around, taking up more space on the road just looking for a parking spot.

I mean think about it. If you can work for an hour while your car drives you to the office, and you don't have to stress or worry about the traffic, why would you even bother with anything else?

Re: NY Times Rides in Waymo Robotaxis in SF

#89

Post by jsnell »

abhayhegde wrote: post #57

> Though there’s no driver, each ride is supported by staff at a Waymo site that can be summoned if a car runs into trouble. Unless they have multiple rides looked after by a staff, what are they planning to save by not offering a driver? Also, is there any path to profitability given how many billions were poured into the technology? Not that I care so much though, Alphabet has enough profitable ventures to recoup o…

Of course there's not one on-duty staffer per car. You'd only need that many people if all rides required human support at exactly the same time. But in reality, most rides wouldn't require it at all. Even if there's likely to be some correlation between support requests, it'd be all cars on one street or block needing help at the same time, not all cars in the same city.

> Also, is there any path to profitability given how many billions were poured into the technology?

Sure, let's do it back of the envelope.

First, the unit economies. Let's assume that a taxi is driven for 16 hours per day (two 8 hour shifts), and the average wage for a driver is $15 / hour. That works out to 16 hours / day * $15 / hour * 365 days/year = $87600/year. This is how much less it costs to operate a self-driving car is than one driven by a human. Since human-operated taxis won't be running at a loss, this means a self-driving taxi would make a minimum of ~$90k/year profit if it got as many drives as a human.

Would they get as many drives? Would people prefer a self-driving taxi over a human-driven one? At least if the self-driving taxi is cheaper, I don't see why not. It'd be easy for the self-driving taxis to undercut human drivers on price, while still remaining very profitable. The big challenge is the service area, where self-driving cars currently have no flexibility.

The unit economies seem to check out. Is the market big enough to run this as a proper business, or is it depending on unrealistic growth? The numbers I'm finding aren't super consistent, but it seems that the taxi + ride-sharing business is something like $50B-$100B / year just in the US. The addressable market is definitely sufficiently large to pay for billions in R&D.

So really it's just a question of how fast they can scale it up, and how quickly the competition catches up.

Re: NY Times Rides in Waymo Robotaxis in SF

#90

Post by Vecr »

quickthrower2 wrote: post #11
jwr wrote: post #5

As a programmer, I won't put myself into that thing. I'm also not happy to be an unwilling participant in this experiment. Given the state of technology and our capabilities to write bug-free software (Seriously, we can't even get things like buying airline tickets correct), I think it's way too early to approve those vehicles on city streets.

Do you avoid flying too?

In GA, yeah. On commercial passenger jets with a couple well-trained pilots and a multiple-times lower fatality rate per hour than cars, there's no safer place to be if you need to cross the continental US. I doubt a self driving car is anywhere near as safe as that, especially considering how unlikely violent crime is to get anywhere on a US domestic flight, for obvious reasons.
Post reply on HN