The robotaxis are coming... the driverless car, AV thread. Waymo, Zoox, and others

we should prob have a separate thread for driverless cars given the Waymo growth?

― sleeve, Wednesday, June 25, 2025 1:48 PM (five hours ago)

As someone who bikes a LOT in San Francisco, I'll say this much: the Waymos are by far and away the safest cars to bike near and around. They are very cautious, drive slow and politely, always let me move in front of them, and I generally feel super safe around them. This has NOT been the case for many years around most human drivers and especially old school cabbies when they were more of a thing prior to 2015. Nearly got run over by them a LOT. If Ubers are replaced with robots like they replaced cabbies, as a cyclist I'm all for it.

I've yet to ride in a Waymo though.

the previous discussion

As an SF pedestrian my opinion on Waymos is...neutral. But not thrilled, honestly. For a couple of years in the early pandemic, as I did my morning walk, I saw a prototype regularly out at the same time I was, with a human driver clearly putting it through its general paces, getting it used to the blocks and areas, things like that. Vaguely interesting to note. But the couple of times I've been near them when out and about is weirdly uncanny, and I can't see myself ever actually using one. As was said in the discussion sleeve linked, knowing that any accident will involve Alphabet hiding behind as many lawyers as possible to avoid either paying up or admitting fault doesn't thrill me much. (And I don't knock octobeard's point at all but my sis, who lives in the city and is a biker herself, was in an accident the other year -- while driving in this case. The other driver was very much at fault and there were recordings to readily prove it, and while it took a while for insurance claims to go through and the legal niceties to be observed, pretty much that other driver's insurance knew they'd have to pay and did, earlier this year. I half suspect if it were a Waymo then by now Alphabet would be on its twentieth motion of 'but what IS an accident really' and trying to fob my sis off with a much smaller settlement.)

I do vaguely worry about them being hacked and turned into 1.5 tonne killing machines.

https://www.youtube.com/watch?v=SOVhz1PllJU

you're leaving out the part where a driverless car does not have a driver

― a (waterface), Wednesday, 25 June 2025 20:40 (yesterday) bookmarkflaglink

I didn't really get this as a slam-dunk, is the point a) you need a soul to drive, who will be listening to Springsteen records if it's all robots or b) robots will be able to drive better than people but it's important that more people die on the roads.

it's not meant to be a slam dunk. cars should have drivers. don't know why you're being all cute with the Springsteen reference--which doesn't really make sense. i also don't recall saying anything about people dying--but if that's your point, the idea that robots can be better drivers than people, and will kill less people on the roads--is foolish

sorry, how is that foolish ?

how about we take the tactic where you explain to me and give evidence that robots will eventually be better drivers than people and that would should continue to invest time money and infastructure in cars versus other forms of public transportation

Is there anything that robots cannot do better than humans ? We're early in the technology and safety is already presented as an argument in favor of AV, so imagine in 5-10 years.
https://www.nature.com/articles/s41467-024-48526-4
https://citiesofthefuture.eu/driverless-cars-far-safer-than-human-drivers/

Maybe you meant more as an ethical concern though ? I certainly feel differently for Elaine Herzberg (first person to die in an AV road accident) than Bridget Driscoll (first person to die in a road accident).

Ethical is not quite the right word, I mean establishing the chain of responsibility / liability.

Of course it comes down to ethics, because people are not going to treat these cars the same. Example here, go to about 4:40. Waymos are programmed to be safe--which has a limitation when you're trying to merge on a highway and for a few seconds, a human driver would do a slightly "unsafe" thing like butting into the merge lane versus the Waymo which just sits there. I don't know how you fix that one.

https://www.youtube.com/watch?v=Q4ldcJmf1a0

Is there anything that robots cannot do better than humans ?

lol

make a meal
write a book
create a piece of art
love another human

also that second study you linked to WAS WRITTEN BY WAYMO

Recently saw this article about bringing these things to London

https://www.businessinsider.com/i-took-chaotic-robotaxi-ride-through-london-impressive-one-question-2025-6

The city's hodgepodge of Roman and Victorian roads are a mess of cycle lanes and pedestrian crossings, with complex road layouts that often serve more as a rough guide than a rulebook for the millions of drivers passing through the city each day.

For Wayve, that complexity is the point. The company says its AI driver — which runs on an end-to-end AI model, an approach also adopted by Tesla — is capable of generalizing and reacting to the physical world in the same way a human would, unlike rivals like Waymo, which rely on high-definition maps and sensors.

Kendall said that this allows Wayve's software to drive anywhere, even places it hasn't seen before, and deal with the kind of unexpected encounters that are an everyday occurrence on the streets of a major city like London.

"I can't wait to see another autonomy company come into London because I think it's extremely challenging," said Kendall.

"The advantage of starting in London is that we've been forced to develop a system that can operate on complex roads and deal with all of these unexpected scenarios," he added.

In the first few minutes of our drive, we encountered multiple jaywalkers, including several who darted out across the street without warning in front of the robotaxi. We also had to inch through narrow gaps between rows of parked cars.

And yeah this thing is going to cause crashes on the North Circular within hours.

Also there have by definition never been any "jaywalkers" in England and the fact that this prick doesn't know that say a lot about quite what a prick he is.

Also 'hodgepodge' instead of 'hotchpotch'

The simple reality is that while AVs might drive more safely than humans under optimal conditions, conditions are rarely optimal, either due to human, infrastructure, or environmental factors. This fact alone means that they will never be able to fully integrate into current systems.

That they also stifle investment in public transportation infrastructure and further silo people away from each other is another compelling argument against them.

And finally, I admit that I am also opposed to them because unlike many people here, I actually *enjoy* driving, and I always have. I walk and ride my bike quite a bit, and I take public transit quite often, too, but I love my little ten year old Subaru. Even in the context of Philadelphia, which has some of the worst roads and scariest drivers of any city in the US, I still love driving.

I guess this is maybe for the controversial opinions thread but my thinking is that there are way too many dangerous/distracted/bad drivers out there and it seems very likely to me that autonomous vehicles will on average be much better than humans at driving. For every suboptimal condition where an autonomous vehicle might perform worse than human drivers there are probably dozens of totally normal conditions where human drivers make mistakes or drive dangerously where a robot driver will do much better so overall I think AVs come out ahead (if not now, then eventually, inevitably).

All that being said, I of course think investing in mass transit will always be a better use of resources than investing in individual cars, no matter how they are being driven.

Aren't computers way worse than people at interpreting and making decisions about the VAST NUMBERS OF THINGS we see and interact with in everyday life?

My understanding, which might be out of date now, was that the only practical use for fully autonomous driving would be for ex long-haul trucking on optimized highways where traffic largely follows norms and drivers interact with each other much less. So trucks would drive between nexuses where freight would have to be picked up by human drivers. (Honestly I wish this would happen in NYC because full size semis regularly go down streets they're not cleared to use and get stuck.)

I guess I should make clear that I don't think a robot driver will be necessarily be better than a human driver at his best, just that a robot driver will be better than a typical tired and distracted driver who is speeding because they are late for work. Humans are very good at plenty of things, including driving, but we're just not always operating at 100%.

I don't know, maybe it's just that I live in a city with a bad driver reputation. I feel like the robots would do better even if I think they will be far from perfect.

Remember that driverless taxis are operating in multiple cities in the US right now (and have been driving around in those cities for 2+ years now). This isn't an argument about IF they can handle real-world conditions, they already are. More data is needed on their safety record vs human drivers but early signs are that while they get in more minor accidents, they cause fewer serious/fatal injuries per driver mile at their current level of ability and are getting better. Human drivers, it needs to be said, are getting worse, especially post-COVID.

it's not meant to be a slam dunk. cars should have drivers. don't know why you're being all cute with the Springsteen reference--which doesn't really make sense. i also don't recall saying anything about people dying--but if that's your point, the idea that robots can be better drivers than people, and will kill less people on the roads--is foolish

― a (waterface), Thursday, June 26, 2025 6:43 AM (four hours ago) bookmarkflaglink

sorry, how is that foolish ?

― Naledi, Thursday, June 26, 2025 6:45 AM (four hours ago) bookmarkflaglink

how about we take the tactic where you explain to me and give evidence that robots will eventually be better drivers than people and that would should continue to invest time money and infastructure in cars versus other forms of public transportation

― a (waterface)

airplanes

autopilot on airplanes

humans are horrible drivers, and we're getting worse

i also think ai-guided autonomous everything is bad. i also wish there were no cars, that everyone walked and biked, and that public transportation in the country i live in wasn't destroyed in order to facilitate as many cars as possible. just getting that part out there, because that counterpoint always comes.

but in the meantime, it might be fun to go back to when autopilot became a feature on airplanes and see what people said about it

(yes, i also i understand that autopilot on a plane is different than the problems of doing it on highways with cars and a million different objects and weird situations)

(but i also think it's a hilarious self-own when people jump on some self-driving car accident/death as proof that it will never work, while ignoring the tens of thousands who die every year in the u.s. from their own terrible driving, let alone the much larger number of people who get severly injured)

again not saying anything about accidents, deaths, injuries, etc. i am just saying show me where a robot is going to be a better driver than a human and show me your work.

humans are horrible drivers, and we're getting worse

https://www.nhtsa.gov/press-releases/nhtsa-2023-traffic-fatalities-2024-estimates

The U.S. Department of Transportation’s National Highway Traffic Safety Administration today released its early estimates of traffic fatalities for 2024, projecting that 39,345 people died in traffic crashes. This represents a decrease of about 3.8% compared to the 40,901 fatalities reported in 2023 and marks the first time since 2020 that the number of fatalities fell below 40,000.

The quarterly fatality declines that began in the second quarter of 2022 also continued, with the fourth quarter of 2024 marking the 11th consecutive quarterly decrease in traffic fatalities.

“It’s encouraging to see that traffic fatalities are continuing to fall from their COVID pandemic highs. Total road fatalities, however, remain significantly higher than a decade ago, and America’s traffic fatality rate remains high relative to many peer nations,” NHTSA Chief Counsel Peter Simshauser said. “To reduce fatalities further, USDOT is working closely to partner with the law enforcement community to enhance traffic enforcement on our roads, including speeding, impairment, distraction, and lack of seatbelt use.”

but there ya go, there's your accident stats. we are still high compared to other countries, but i would imagine that's because we have more drivers

AND BIGGER MORE DANGEROUS CARS

yeah, the baseline here (speaking just about the USA) is that with human-piloted cars there is a vehicle fatality about every 12 minutes. consider just some categories of fatalities that would be entirely eliminated with driverless vehicles - distracted driving, speeding, and drunk driving are the top three causes of vehicle fatalities and autonomous cars do not do those things.

what is driving off the road and into a fire hydrant if not "distracted driving"?

yeah, the baseline here (speaking just about the USA) is that with human-piloted cars there is a vehicle fatality about every 12 minutes.

There are plenty of ways to solve this problem that don't involve robot cars and making money for huge corporations to sell us cars and taxis that drive themselves

consider just some categories of fatalities that would be entirely eliminated with driverless vehicles - distracted driving, speeding, and drunk driving are the top three causes of vehicle fatalities and autonomous cars do not do those things.

sure but now replace drunk driving with a new category "a computer made an error and drove a Cybertruck onto a sidewalk and killed a bunch of pedestrians not to mention the people in the car"

xp
those aren't my accident stats. there's a lot of ways to look at stats, and a lot of ways to cite them.

check out "national statistics", here, and look at the excel sheet: https://cdan.dot.gov/tsftables/tsfar.htm#

it only goes through 2023 and back to 2010, but per capita (number of people, number of vehicle miles traveled, etc), fatalities have only gone up. injury rates have gone slightly down.
--

but again, i think all of that is pointless. is 30,000-50,000 deaths per year from cars the gold standard? plus 2 million or so injuries? is there a way to transport people that doesn't kill that many, every year? again, i fucking hate the ai stuff everywhere, i'm just saying, i wouldn't mind if less people died

sure but now replace drunk driving with a new category "a computer made an error and drove a Cybertruck onto a sidewalk and killed a bunch of pedestrians not to mention the people in the car"

well, if you include "fantasies I had about autonomous vehicles killing tons of people" in your stats then they're gonna skew towards human drivers being better but I was thinking we'd stick to those causes of vehicle fatalities that actually happen in large numbers

sure but now replace drunk driving with a new category "a computer made an error and drove a Cybertruck onto a sidewalk and killed a bunch of pedestrians not to mention the people in the car"

and sorry (this is my least popular opinion here, or one of them i think), but yeah go ahead and do that! replace the drunk driving category, which is a really high number, and replace it with "computer error", which i think will end up being a much lower number

xp

there is a way, it’s called rapid rail transport and local public transit infrastructure

xpost once again i will remind you that i agree with you that fewer people should die and that is not the argument i am having, but it seems to be the argument everyone wants to have because having robots drive cars is kind of indefensable but sure go ahead and continue to bring up the grim spectre of death and avoid the idea that you have to mount up a defense of why technology should be allowed to drive cars

there is a way, it’s called rapid rail transport and local public transit infrastructure

^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

also did it occur to any of you robotstans the idea that there are fewer crashes with the robot cars because THERE ARE FEWER ROBOT CARS ON THE ROAD

i also think ai-guided autonomous everything is bad. i also wish there were no cars, that everyone walked and biked, and that public transportation in the country i live in wasn't destroyed in order to facilitate as many cars as possible. just getting that part out there, because that counterpoint always comes.

― z_tbd, Thursday, June 26, 2025 11:23 AM (thirty-nine minutes ago) bookmarkflaglink

also did it occur to any of you robotstans the idea that there are fewer crashes with the robot cars because THERE ARE FEWER ROBOT CARS ON THE ROAD

yes it did waterface

jfc

why post

me, i mean

why post

like LLMs and other AI models, it comes down to people who are resigned to it and those who are dead set against it, and imho those who are resigned to it have no spiritual backbone

did anyone read anything i posted

i might ask you the same question

didn't you incorrectly cite stats in a really obviously cherry-picked fashion, ignored everything i said about wishing cars didn't exist in the first place, and then called me a robotstan?

I don't have animosity to spare for z or hazel or anyone here. I save mine for people (I have heard at least one person say this) who say "I want driverless cars because I personally dislike driving and want to be driven." INSTEAD OF LIVING SOMEWHERE ELSE OR IMPROVING TRANSIT WITH YOUR CULTURAL CAPITAL AND MONEY. Once again we see the uptake of "AI" driven by people's desire to be served, to live and feel like whatever passes for our current aristocracy.

did anyone read anything i posted

i did, z, but some of your other posts betray being resigned to AVs and okay with then in some ways.

yeah! i'm into the idea of fewer people dying

it's because of my spiritual emptiness

seriously - i've seen you use that a few times, and you're usually "on my side" in ai bullshit - again, anyone with nuance isn't confused about where i stand - and the whole "spiritual" argument is complete bullshit

i'm sure most people could care less about that being brought into it, but yeah, that does just seriously piss me off and it's dumb

i didn't cherry pick statistics, not intentionally--i simply pointed out that fatalities were down but we're still high relative to other countries. if there's a way i cherry picked stats, i honestly didn't mean it that way.

it's hard for me to reconcile your idea of wishing cars don't exist with your acceptance of robotaxis as being ok because they will reduce fatalities (there's no proof of this) and not think you really want robocars to exist. again there are countless ways to solve this problem that don't involve cars.

autopilot on planes strikes me as a limited analogy, because for very good reasons, we still require multiple humans on board and ready to take over, and to do the crucial parts of the flight which cannot be entrusted to the autopilot. i think the driving equivalent is not a robot car, but cruise control.

I'm not firmly in either camp but...

When it comes to determining "fault" in a crash, there's already several biases. Insurance and police make their own determination of fault/cause.

Insurance companies try to avoid liability and blame the other company's driver and view the facts through a tinted lens. Cops otoh regularly give citations to the wrong person based on their own biases, or just grasping at straws based on competing narratives with no witnesses.

Manufacturers of self-driving cars already try to control the narrative regarding accidents involving their vehicles. When this industry gains an even bigger foothold in cities that don't currently have them, I imagine they are going to, to Ned's point, be throwing a lot of money at deflecting blame to the other driver.

Just like the creator of Taser insists that product doesn't kill anyone.

Given how the last guardrails are eroding just about everywhere, I have a bad feeling the data is going to be politically manipulated.

Doesn't mean I'm against it or that o don't trust statistics, but this is an industry still in its infancy so that's often when the data is at its worst

i'm sure most people could care less about that being brought into it, but yeah, that does just seriously piss me off and it's dumb

― z_tbd, Thursday, June 26, 2025 10:18 AM (twenty-eight minutes ago) bookmarkflaglink

sorry, honestly that wasn't directed at you, but point taken.

Given that:

- Automation of anything is hard
- Accidents will always happen
- The capability of human beings to adapt to new circumstances is possibly our greatest attribute
- Every technological advance we have ever had has been met with fatalistic doom by a vocal contingent

It is hard for me to believe that we won’t have widespread acceptance of driverless vehicles in the us within the next 50 years and there will be a lot of retroactive “not sure what all the fuss was about” as the baseline attitude towards the arguments we are having now

Having said that, we aren’t living 50 years from now; we are living right now. There are a lot of things I enjoy being on the bleeding edge of but this is not one of them. I fully expect an unforeseen catastrophe to crop up in the refinement/adaptation process, whether that manifests as direct injury due to machine error or malicious use, or societal fallout from eliminating a sector of work. So, while I feel that they are somewhat inevitable at this point, I am not at all looking forward to what I believe we will go through to get there.

do It’s ok. And I’m sorry to call what you said “dumb”, too , that’s a word I shouldn’t ever use. I have an idea of where you’re coming from, I do. I just don’t like to be mischaracterized (or feel like I am) and although it wasn’t pointed at me, I know that I’m not a stoic tech-advocate trying to pull wool over anyone’s eyes. I cry my guts out about all sorts of shit, all the time, and try my best, and that’s true of other people who hold other opinions too

don't really feel comfortable posting in this thread anymore, not sure why I deserve personal attacks for talking about autonomous vehicles which (as I've stated repeatedly) I don't even particularly like or approve of

do It’s ok.

this is my iphone autocorrecting "xp it's ok"

djp comes closer to my own view than what anyone else has said. i would guess that there is a strong possibility of a massive data breach, a big grid / communications shutdown, risk of a big hack, etc etc, there's all sorts of horrible things that could happen. it's hard to balance that against the current status quo, which is not hypothetically a tragedy for millions of people but is an active, ongoing one

Every technological advance we have ever had has been met with fatalistic doom by a vocal contingent

I would argue that one of the bad things about VC is that the failures and almost-successes get more press than the actual successes, because the things that are actually successful are either not flashy enough for anyone to care about and thus get no press, or become too successful and end up as targets for disruption by the churn machine. (The popularity of the failures should be self-evident; everyone lives a good hubris-driven faceplant.)

I'm not saying driverless cars will be safer. I'm saying they will implement a "discover weekly" feature where they drive you to other folks' favorite coffee, smoke, and burrito shops.

maybe it's accelerationist (hoho) wishful thinking but I do strongly suspect driverless cars will potentially make car driving in general so miserable that it's maybe the only realistic path to reclaiming city roads for bikes/pedestrians. i can think of a lot of dystopian ways to make this happen!

Busy roads will be more crowded than ever. Driverless cars will converge to already congested areas in order to win more rides. Google will read your email, find out you have an appointment at 8am, and sell that information to 28 robotaxis that will hover outside your building playing annoying music like a swarm of Lloyd Christmas Bumblebees

so you're saying you want people to be killed by drunk drivers?

one man's drink is booze. another man's drink is hubris.

people who are out and have to shit will book a 2 block ride and will leave more than 2 blocks behind them

probably just used it as a toilet and moved on

sometimes I feel
the need to move on
rent a robocab
and move on
move on

waiting for the first instance of a riderless car like, going to a destination 40 miles away from the one you input

wait will a driverless car pick up hitchhikers? asking for a friend

Will they spin doughnuts if asked?

Living with these things daily and ubiquitously around me, in a pretty chaotic and crowded urban city like SF, the doom posting around safety is just not true with Waymos. Like, at all. I think there's a lot of assumptions/anxiety around safety that spring from Tesla's handling of this tech (and its leadership, obviously), and associating that with what the rest of the industry is doing and that's just not true, and I can see it with my own eyes. My guess is those expressing the most anxiety have yet to see one of these vehicles or live with them on a regular basis.

maybe should save it for the doomposting thread, but the VC era (let's say 2009 onwards) has seen very few advances, everything feels like an unfulfilled promise, many things seem to actually be getting worse.

This is just not true with this tech, or ride sharing in general, it was a genuine need in places where cabs were terrible (like SF) and mass transit, while around us and effective, aren't exactly NYC subway quality. With driverless cars, it's pretty wild to see them, and their very presence and efficacy demonstrates to many, with their own eyes, that tech is changing and improving in dramatic ways. I often see tourists photographing their trip in or out of a Waymo.

I find it amusing this thread has a lot of people panicking around "safety" with this tech. It's only getting better in that area, and knowing how LLMs and AI dev works at a super high level, it will only get better. The safety is no longer my concern with these things.

The areas that I find concerning are more legal, ethical or economical, and they stem from non-tech related issues and more corporate or political. Ned brought up a great point wrt to liability. That's a big deal. With regards to mass transit usage, these issues are the exact same with ride sharing - if you look at AVs from the lense of simply usurping Lyft/Uber's market, the issues specific to the market then become a narrower, and it becomes more challenging in the face of the very real benefit of the safety the vehicles offer, which I can directly attest to, having to bike around them and humans daily now. In terms of privacy/tracking - Lyft/Uber et al already have this capability. Waymos basically are that but without the driver.

xp sideshows will be polite and orderly

In another angle - multiple female friends of mine expressed an open mind towards robotaxis recently because they don't trust taking Lyfts/Ubers late at night alone and have experienced incidents where they felt very uncomfortable or unsafe.

they're not allowed on freeways though, right? At least not yet

I live in the northeastern part of SF which is heavily frequented by tourists, and Waymos are ubiquitous. Sometimes it feels like every 3rd or 4th car is a Waymo. I have read that they now comprise 27% of rideshare services in SF. It is astonishing how quickly they have been accepted. At first they were very hesitant and too cautious, but lately the technology I think has improved, and they move in sync with the traffic.

They still drive safely but now at a more normal speed, and they don’t hold up traffic. I’m starting to trust them more than I trust most crazy SF drivers. It’s kind of funny to see people hesitant to step out at a crosswalk in front of them.

They are now testing them out on highways but are not offering that service to the public yet. I’m not sure at this point I would want to ride in a Waymo on an interstate highway

It’s kind of funny to see people hesitant to step out at a crosswalk in front of them.

Happened to me last week. Was crossing a quiet street, one pulled up and I still very much gave it a side-eye.

It's not like cars don't already have dashcams but it would be interesting to see companies offering to hand over footage to catch plate numbers of traffic scofflaws as incentives to bring them to a city...

xpost once again i will remind you that i agree with you that fewer people should die and that is not the argument i am having, but it seems to be the argument everyone wants to have because having robots drive cars is kind of indefensable but sure go ahead and continue to bring up the grim spectre of death and avoid the idea that you have to mount up a defense of why technology should be allowed to drive cars

― a (waterface), Thursday, 26 June 2025 18:05 (yesterday) bookmarkflaglink

See this is what I was getting at with the "you need a soul to drive"; you seem to have as an axiom that technology must not drive cars, even factoring out safety or politics, and (maybe I've missed something but) I don't really think you've dug into that?

That's already happening with regular people on reddit. Some dude posted a dashcam of a person doing a hit and run at the intersection of Geary and Masonic last week

Oops XP!

ha "you need a soul to drive" reminds me a lot of the arguments I had with people in the 90's telling me electronic music was soulless.

Souldriver average 6 monthly listens on spotify

As another SF dweller and heavy cyclist, I agree w/ Octo that Waymo so far has been considerate and safe.

Because of how Cruise was run out, O think Waymo will continue to work hard on keeping a good actor image.

With that said:

- way easier to recover your forgotton phone w/ waymo va uber/lyft. At least the one time, just had to go to the depot the depot.

- easier for me to pay for a robot than to pay for a human driver who is getting (in my view) screwed

- i try to bike or take public transit instead

The Zooxs riding around SF are cute, but they look like a tech person's whimsical idea of a cable car or tram on wheels. They are a delight to see, but are just too clunky and impracticable to be useful except as a tourism thing

sorry, I meant impractical

the thing about autopilot is that, in the case of that Air France plane that went down near Brazil... autopilot had been implemented to such a degree that the pilots didn't know wtf to do when things went wrong. major simplification, iirc it's a tricky balance in aeronautics to balance safety with flexibility or what have you.

i witness so much reckless bullshit every time i drive, i have a hard time thinking that some/many drivers shouldn't be forced to be driven by a robot.

it's not really logically sound to be all "machines don't get heartburn, machines don't get obsessed with tv shows"

it's really dumb how you get branded a collaborator/corroborator if you're not completely pissing your pants panicked about whatever you think AI is

it's not really logically sound to be all "machines don't get heartburn, machines don't get obsessed with tv shows"

not sure what that means, but I'm thinking of murderbot, who is obsessed with tv shows

i try to bike or take public transit instead

The biggest transport "revolution" for me has been the proliferation of e-bikes, honestly. Getting around is such a piece of cake and super cheap when you pay the yearly subscription. I think it cost me $120 a year and my longest rides, which tend to be upwards of 5 miles, are not much more than a bus ticket. But I'm not going to use them when I've had a half dozen Jamesons while yelling at the goddamn Giants in a bar

xp sorry, i just meant defending robocars by saying robots don't get "distracted", for example, like, isn't that a category error or something?

do robots get distracted? again not sure what you mean

I commuted in my car 60 miles every day, across the city and into San Mateo County and back for almost 3 decades, and the traffic nightmares, accidents I was involved in, endless delays that left me scrambling to find alternative routes, white-knuckle driving in the pouring rain up-and-down the curvy hills of 280 - those experiences I’m thankful to leave behind.

I still have a car for the occasional longer trips I have to take, but have come to hate driving and am thinking of giving up my car. I value living in a walkable neighborhood and would rather take an Uber/Lyft to visit friends than drive and have to try to find a place to park. I haven’t tried a Waymo yet.

Fwiw when I drive (only in the middle of the day) to visit friends in Oakland I don’t feel like that. The traffic there is reasonable and there are places to park

Yeah I also can't stand driving. It super stresses me out. Hell, owning a car stresses me out (haven't owned one in over a decade). Ideally we'd have a subway like NYC or a proper light rail. But this shit died in California in the 50s. Bart's original plan was to loop the bay area. Between how expensive it is build, NIMBYism, and the current anarcho-late-stage-capitalist society we're pushing towards, it's nearly impossible or prohibitively expensive to seemingly do anything positive towards public transport. If anything it feels like it's getting pulled back - Bart has fewer trains, and they take longer, Muni is losing stops and routes, etc.

I did have an interesting thought - it would be cool if municipal cities could leverage robo-vehicles to create dynamic public bus/shuttle systems, robo busses for main routes and smaller more dynamically routed shuttles to branch out and drop people off in smaller clusters as a public good, and charge appropriately for more specialized routes.

Automated public transport is not a new thing - in Copenhagen the subways are basically those driverless automated airport trams but also in the city center. A new subway line that opened up in Milan is doing something similar too. We're so unimaginative of what public goods can be in the US because we're so used to it either failing or never getting funding and just being cynical about it.

Sydney is pretty crap at public transport, but a new driverless metro line has just opened and it is unbelievably quick. A journey that would have taken 45 minutes before can now be done in 15. I think Paris has some driverless metro lines as well.

"do robots get distracted? again not sure what you mean"

lol are you fucking with me? i'm just talking about, like, anthropomorphism

if it helps, read my post in a voice in between grover and pauley shore

I am not sure I understand the point made upthread to "invest in public transport instead". Investment in robotaxis has no correlation with investment in public transport. A robobus sounds like an obvious application of the tech.

I also really enjoy driving. Roads here are generally safe, people generally disciplined. Under ideal circumstances though, I wouldn't own a car, since I don't need one on a daily basis. I looked into rentals / carsharing and was sad to find that it looked more expensive than buying and not worth the hassle. Robotaxis, in the long-term, with costs optimized, to order from home, looks like an interesting complement to buses and trains, and a reason to ditch a car. Obviously that's not for tomorrow. I have never even seen an AV.

With all that said, I wouldn't say I'm positive about AVs yet - they have to prove themselves, be accepted. I would not forgive the world for the rest of my life if my child was killed due to an algorithm error and became a "statistic" (even in an overall safer world). But I wouldn't rule them out.

I've taken a Waymo any time I would have taken an Uber, since October. Its a really weird feeling at first, and then it's weird how quickly you acclimate. I'd say the only downside is sometimes it drops you off in some random spot like a block away, maybe unlike Ubers etc they aren't willing to idle in a red zone, which for non-CA people is like no stopping anytime technically but is the default loading/waiting zone.

Every technological advance we have ever had has been met with fatalistic doom by a vocal contingent

It's interesting how fear of being in this contingent makes people champion any old crap (not the topic of this thread specifically but it's a major subtext with AI boosters imo).

It's also worth unpacking how much this contingent gets cast as just scaredy cat busybodies when historically it's often included groups whose livelihoods were erased by these advances, and how in a society that is less SHINY THING NOW a lot of advances we've had could still have happened, at a slower pace perhaps but at the cost of much less human misery.

^ ^ ^

True, though the fact that Luddites have been publicly synonymous with progress-denying fools (for as least as long as I've been alive) indicates that it's not just a 21st century phenomenon.

the slope is vertical

I think it's fair to ask "Do robotaxies improve the world?"

It's hard for me to see how this really enables or unlocks something. In urban areas, the transit problem is not the lack of drivers but the fact that single passenger cars take up too much space. While maybe we need less parking, in practice the empty capacity is often on the roads instead.

I definitely agree with the fact that any criticism of new technology immediately gets an anti-Luddite response. We only need to look to the blockchain fad to see how that can make a bubble.

With that said, to me:

- Who are the losers likely to be? If its pedestrians, cyclists, urban residents, the poor...what can we do about it?
- We clearly have actors who are acting in bad faith (Tesla, Cruise was doing this too). How do we contain that?

To me, the likely answers to this question lead me to believe
- We should tax these heavily and use the revenue to finance public transit, cycling, and pedestrian improvements
- We should demand safety transparency and accountability

- Who are the losers likely to be? If its pedestrians, cyclists, urban residents, the poor...what can we do about it?

Cyclists and pedestrians aren't losers with these things - remember we already have ride sharing. These things won't roll coal or drive aggressively around people on bikes or pedestrians. If you're saying we won't get more bike lanes or train lines, that's not an issue AVs are suppressing any more than current driver based ride sharing and politica/economic climate in the US, so it's not an AV specific concern.

Otherwise OTM. Safety, accountability and transparency should be legal mandates, right now it's not so much with these companies and they are "volunteering it" (at least that's what I'm seeing from Waymo and Zoox's public statements wrt accountability).

It shouldn't be a "competitive advantage" it should be table stakes.

I think the ultimate losers will be drivers, who are going to be bullied from occupational and recreational driving. The jobs will go away and driving will just be miserable, but I don't think it's foregone that pedestrians etc... will be winners from this. They really have to claim the streets and do the constant mini parades across unprotected crosswalks that self-drivers will be forbidden to plow through.

Yeah occupational for sure. Recreational I could see too. I've long envisioned a potential dystopian outcome of driverless vehicles in that they'd be normalized for safety purposes, but once saturated and the safety metrics prove out, it then becomes illegal to drive on your own, which then prevents people from freely traveling without being tracked and to prevent organized protests, etc. Ironically I see this happening in the UK or EU first, as the culture around car ownership and driving in the US is so strong, and extra strong in red states, that it would receive incredible push back.

xp I don't think drivers will be bullied off the roads, that is ridiculous

"Every technological advance we have ever had has been met with fatalistic doom by a vocal contingent

― my favorite herbs are fennel and Drake (DJP)"

This is true

"It's also worth unpacking how much this contingent gets cast as just scaredy cat busybodies when historically it's often included groups whose livelihoods were erased by these advances, and how in a society that is less SHINY THING NOW a lot of advances we've had could still have happened, at a slower pace perhaps but at the cost of much less human misery.

― a ZX spectrum is haunting Europe (Daniel_Rf), Friday, June 27, 2025"

I don't know about that. I think human-driven ride shares will continue to exist. What kind of slower pace do you imagine?