It’s wild that $goog is so undervalued (p/e 27) given Alphabet owns Waymo in addition to everything else, and yet Tesla is so overvalued (p/e 243!!!) despite zero Robotaxis in the near (or far) future and lackluster sales.
Goes to show empty promises and fraudulent showmanship sell better than actual working products that people use.
GOOGL is up like 25% over the last few weeks after they resolved the DoJ lawsuit about Search bundling. Clearly there were some investors who thought that was a material risk to the business.
Tesla is clearly a meme stock though, and an example of how the market can say irrational longer than you can stay solvent.
If you buy Alphabet stock you're betting on the whole company doing well.
Google makes around $300B a year. Uber's entire business makes around $50B and that took a decade. Waymo would have to become a major business to move Alphabet's stock price in the near term.
Considering Waymo is very likely losing money, experiment very slowly with scaling up, and still raising billions in private capital outside Google... idk. Doesn't seem as simple as buy $goog in 2025.
I think Waymo has huge potential for being much larger than Uber - people are willing to pay more compared to ordinary uber drive just to avoid dealing with taxi drivers and tech will only get cheaper.
More than that, I think the ride-hailing business is just the fist volley in the self driving vehicle space. It’s a short jump from there to self driving trucks, self driving package delivery, self driving private vehicles, and on and on.
Can any of those companies catch up on self-driving faster than Waymo can pivot to their niche? Cruise seemed to be a distant second, but did themselves in with an attempted cover-up.
> There are already self-driving trucks on the roads.
2 trucks?! I suppose that's the minimum number required to make your pluralization correct.
I will stand on my earlier statement regarding this particular outfit: they'll need to catch up because Waymo started class 8 variants in 2021 https://waymo.com/blog/search/?t=Waymo%20Via
And plenty have failed. Perhaps a smaller problem space but still really, really hard. Some self driving freight company failures: Starsky, TuSimple, Embark, Ghost, among others.
One promising self driving truck startup, Aurora, was forced to put a safety driver back in the driver's seat after testing in May.
Buy a Comma.ai and install it in a supported vehicle, and just try it out. It doesn't talk to GPS, but it handles left right gas brake on the freeway well enough, and that's with two fairly shit optical cameras and a radar system. Granted, geohot helped start the company, and he's no slouch, but if their system is that good, a couple things are true. A) Lidar isn't necessary b) Extensive mapping that Waymo does also isn't necessary c) that last 10% gonna take 500% of the time to get to L3/4/5 autonomous, and that last 1% is maybe never. The other day I was in a Waymo, and there was a semi totally blocking the street, backing into a loading dock. The Waymo correctly identified that there was an object in the way, and stopped and did not plow into it. At first it crept up to the semi, blocking it from making progress as well. It might have started backing up, I've seen them do that, but I was already on the customer support line as soon as I saw the semi blocking the road.
Comma.ai is probably the purchase I'm most happy with this year (to be fair though, I buy a lot of crap off Temu). Drives are now just "get on the freeway, and just chill." Pay enough attention because it's not collected to GPS and just in case something goes wrong. So to be clear, Comma.ai is not autonomous driving, it's classified as an ADAS, advanced driving assistance program. It just makes driving suck that much less, especially in stop and go traffic, for $1,000, and compatible with recent vehicles that have built-in lane guidance features. Waymo's got to be light years ahead of them, given how much money they've spent, so it's my belief that Waymo's taking it very slow and cautious, and that their technology is much more advanced than we've been told.
Cruise was nixed by GM execs, whom I believe were looking for whatever excuse they could find to shut the operation down. They simply couldn't afford to stay in the game for the long haul. Cruise was under pressure to appear more capable than they were, and they took risks.
Waymo is distinguished in that it doesn't need to pander to nervous investors to keep getting money. The company is Sergei and Larry's baby. Google's founders will ensure that Waymo is patronized until it can stand on it's own.
> ...I believe were looking for whatever excuse they could find to shut the operation down
Cruise's self driving license was suspended because humans displayed poor judgement by omitting from the official report details of their stopped car dragging a knocked-down accident victim under the car for dozens of feet. They took "risks" alright, and their harebrained cover-up was discovered by chance by the oversight body.
I believe any driver who covers up the details of injuries in an accident permanently lose their license, because they'll definitely do it again. What good is a self-driving subsidiary that can't operate on public roads?
There are several “last meters” delivery robots developed.
Short range drones are being used in Australia.
And I heard of at least one company working with apartment architects to standardize a “port” on the building exterior to which a truck/robot would connect to “inject” packages to the inside.
This was just an acquaintance some years ago in SF, but I recall it was fancier with conveyor belts and a protocol for the robot to communicate the size and weights of the packages being delivered.
But to be serious, there may be a way of doing it, it just seems very far off unless you're talking about Amazon hub or something like that, where it would be more feasible (but still difficult to achieve).
Think of Waymo Driver as the equivalent of Android for vehicles. It's an operating system and a suite of cloud services for both autonomy and ride hailing.
What about all the expensive hardware, gpus, lidars? That’s like having iOS on your phone and if you want android you need to buy extra things that are worth same price as your phone.
And costs should be lower in the long run if you don't have to share the ride fee with a driver (not case yet because seems like they still have alot of staff to manage the cars)
Statistically Waymos are more expensive than Uber rides, but practically as an individual they are often cheaper than Uber, its very easy for the stated price to be lower
That TC article doesn't substantiate its overly broad claim. "People" aren't paying more, in general, across its US markets; it only shows that a subset of its customers in what is already the top-5 most expensive cities (SF) in the world are prepared, and at that, only 10-27% are prepared to pay significantly more ($5-10). Still fewer than the 40% who would pay “the same or less.”
Quoting: "Perhaps even more striking is how people answered a question about whether they would be willing to pay more for a Waymo. Nearly 40% said they’d pay “the same or less.” But 16.3% said they’d pay less than $5 more per ride. Another 10.1% said they’d pay up to $5 more per ride. And 16.3% said they’d pay up to $10 more per ride."
There are going to be lots of causal factors: number of rider(s), time of day, safety, gender, wait time, price estimate, predictable arrival. Let's see an apples-to-apples comparison/regression breaking out each.
Anyone who's taken enough Ubers and/or has had bad enough luck to have gotten a terrible Uber driver. Pretty much everyone I know, along with myself have had multiple awful Uber driver experiences.
I think waymo actually has a better km/accident ratio than the average driver. Plus if you haven't done it before, it'll be a cool experience to ride in a car with no driver!
But in the long term I think the point of waymo is that it'll be cheaper: no need to pay the driver if there isn't one!
I don’t think Waymo is very likely a losing money experiment. I give them a 50% chance to be successful within the next 10 years. Successful being that self-driving cars are able to operate in 50% of the world/terrain types/region types, probably within another 10 years to scale up.
They have already spent an enormous amount of money. It’s hard to see how they could make it back quickly, if ever. I’d like to be wrong, but I expect they will continue to be a money losing experiment for a long time yet.
How much money they've spent in the past is irrelevant. That money all came from investors, in exchange for a stake in the company. It never needs to be "paid back". Besides which, those investors have earned all those funds back already, and then some (on paper).
All that matters at this point is how much money they'll lose/earn in the future. There are no shortage of investors willing to put money into this effort, and they're growing exponentially, so there won't be any pressure for them to turn overall profitable for several more years.
Waymo investors are Google investors, are they not? Will Google shareholders be happy continuing to put gigabucks into Waymo, or will they prefer that Google cut expenses before Waymo can make money?
How much money do they make off the average person in the value of ads shown per year?
Now compare to how much money the average person spends on driving per year.
If Waymo winds up running half the market in autonomous transportation over the next several decades, it'll make search look like peanuts in comparison.
When these are ubiquitous enough, the vast majority of people who currently own cars won't need to. It'll be so much cheaper and easier to use rideshare.
I can't really imagine the circumstances where I wouldn't want to still own my own vehicle even if it had an autonomous mode. I drive it places where I don't have cellular service. I keep lots of stuff in the vehicle. It's customized with accessories like roofracks. I can hop in my vehicle from my house immediately whenever I want to.
If I lived in a city and garaging a car were inconvenient/expensive? Maybe. But that's not me or a lot of other people.
But if it's half the price over the course of a year? And you can summon it in advance cheaply? And it basically never takes more than 5 min to arrive anyways, since they're everywhere?
You might decide it's worth it to keep the stuff you really need in a messenger bag or backpack or something, the way people in NYC do. And maybe roof racks don't matter if you can just summon a second autonomous van behind you to hold whatever you were going to put on your roof.
Obviously if you're a contractor or something you'll need your own vehicle. But the point is that for most people, sure they can't keep stuff in their trunk all the time, but that's a happy tradeoff for a total cost of driving that's 50% less.
Imagine if you could buy your own "Waymo-equipped car". No need for driving lessons. No aggravation. No road rage.
How many people would pay for such a luxury car? With the US population aging and public transit non-existent in most places, Waymo probably has a market for cars.
Uber took 14 years to make it to profitability. Money's frequently characterized as impatient, unable to look past the next quarter, but when it wants to be, it can wait.
Waymo's older than Uber, but they hold many key patents by this point. Now that they've started running a taxi service, it seems straightforwards to scale up, assuming that is the business they want to be in. Then it's just a matter of charging more than it costs to run the service, and wait.
The cost of the computers, LIDAR, special maintenance, vandalism, staffing humans for remote issue handling etc will probably costs the same as a year's income for an Uber driver. But after that it's mostly profit and they can run cars longer.
The most important thing for Waymo is scaling up production of LIDAR and maintaining them efficiently. They will have a massive fleet running very sophisticated radar+computers. That's a huge logistical investment when it's a million cars. Those sensors will break or be damaged.
They've been partnering with Uber to maintain the fleet in some cities haven't they since they already have regional infrastructure? I don't think they want to be in the fleet management business.
AFAIK Uber is doing app integrations + some local operational fleet management. Waymo is supplying the cars, radars, computers, remote service, the brand, etc. Waymo has to scale that production and maintenance up country wide and then globally.
Uber's CEO compared it to Marriot, people come in to run the hotels in the local region, but they actually don't own the hotels. It's like hired managers who take a cut.
It also makes sense to have people with local experience run them in each local region. But those businesses still involve margins and expenses that have to make sense.
Don’t forget that Waymo will always be a much lower margin business than search! Setting aside the decades of R&D expense, those cars require purchasing, maintenance, warehousing, etc.
But the market is so, so much bigger. And the margins will likely stay high for a long time while there are few competitors, and their main competition is human drivers.
Not having to pay drivers is an enormous source of profit.
Don’t forget Zip2, PayPal, Neuralink, OpenAI, and The Boring Company.
There are large swaths of people that accept headlines as fact and/or cannot or will not grapple with nuance and complexity (“I think Elon’s a jerk and he is a formidable engineer.”) Perhaps it’s a sign of these polarized times, or, as I believe, people have always been like this. We just have more time and resources to dedicate to outrage and flamewarring than we did in the past.
Yes, there has been nice geniuses (ie. people with extreme talent), Mozart was for example a good person. Da Vinci (if a little sycophantic when young) was not unhinged at all nor abusive and was appreciated.
But since romantism we have built this image of the genius as necessarily abusive.
I’m sure abusive genius are very visible (by definition?) and that abusive people tend to monopolize more ressources too. (Like these tenured professors that use their students to advance their own career)
He was ousted from Paypal before anything major happened, he was basically just a shareholder.
The Boring Company is an obvious bust. So is the Hyperloop. Neuralink is another likely bust. Tesla solar is going nowhere. The Cybertruck is a millstone around Tesla's neck. Etc, etc.
I think the real purpose of the Boring Company and Hyperloop were preventing/slowing expansion of public transit, and that by that measure they were successful.
I don't think it was a carefully calculated conspiracy (such as 1)
I think it was an engineer with found wealth starting to do stuff with it.
but nowadays I think he has evolved into something different, maybe some of it from the wild public feedback loop, some of it because some of the things he cares about are going wildly wrong.
Usually Elon's technical flaws aren't on display, or at least he covers them well. For example while it's true FSD hasn't worked out, but I don't know you could say at the time "most competent AI devs knew it wouldn't work out". However, when Elon attempted to move PayPal from Linux to Windows, most competent software engineers would have advised against it. Paypal isn't an example of Elon's genius in action - it's the opposite.
When Tesla introduced HW2 it was clear to people in the self-driving industry that it wouldn't work out. Elon was insistent on repeating mistakes that other companies had already learned from. Of course the other companies never considered some people's willingness to pay good money just to pretend that their cars can drive themselves.
Says who? I've tried it and the capabilities are amazing. If you told me 10 years ago that I would be able to buy this in 2025 I wouldn't have believed you.
I am "just a shareholder" in Paypal. Elon Musk had a > 10% stake inherited from his ownership of one of the companies that was the precursor to Paypal itself. It's not remotely the same thing. And listing failures is not meaningful at all. Failure is the default outcome in business.
Either go ague with Wikipedia, or put some argument in the comment when making claims you expect people to verify themselves. People are just going to look it up on Wiki.
> SpaceX was founded by Elon Musk in 2002 with a vision of decreasing the costs of space launches, paving the way to a self-sustaining colony on Mars.
but if they're google's products how would they cannabalize ads biz. would revenue not just shift? or do you believe ai search will be overly adopted but not as profitable?
Google was late to search, late to smartphones, late to internet email. I'm having a hard time thinking of any of their large markets where they were a first mover, maybe YouTube-ish, widespread user uploaded internet video wasn't meaningfully available before the rise of YouTube.
On topic, Waymo is clearly a first mover in self-driving, having the first legal commercial services.
But, being the first mover is usually more of a disadvantage than an advantage, IMHO.
I believe TSLA also represents their humanoid robot segment with some questionable addressable market definitions done by investment analysts. I believe it’s overvalued but they are a forcing function for the other tech companies to push ahead
TIL. I stand corrected. Though worth pointing out (as the article does) that on September 1st, new legislation in Texas was passed adding some restrictions to autonomous vehicles. So seems reasonably likely this is more regulatory than necessary.
My Model Y in Vancouver drives me to and from work daily. I cannot get a Waymo here -- and I certainly cannot purchase one privately. Which is more effective where I live?
I don’t doubt that Waymo car is more advanced than FSD, but that comparison isn’t as impressive as it sounds. The numbers of FSD equipped Teslas dwarfs that of Waymos, and they are available everywhere, not just selective cities. You have to take that into account.
Teslas is also much cheaper, and easier to scale. Tesla has better growth potential even if their tech is less impressive.
It's not that their FSD tech isn't less impressive, it's that it's not FSD tech.
Even worse (for Tesla) is that if they do try an make their non-FSD tech do FSD, and it decks little jimmy because the flashlight in his hand looked like a far off street light, Tesla is liable to face a knee-jerk federal law mandating lidar. And just like that the dream is dead.
This forces Tesla to be extremely paranoid, as it's one visual mistake away from being told to use lidar.
Overvalued by traditional (PE) means. I've ridden in Waymo (50+) and Austin Robotaxis (12). Tesla has Waymo beat in terms of human-like feel, interior features (sync to your own Spotify, Youtube, etc). When Tesla removes the passenger seat monitor, scaling will happen much faster than Waymo... Tesla just received the initial license for driverless Robotaxi in Nevada. Tesla also produces more Robotaxi-capable Model Ys in ~6 hours as Waymo has cars in service (in total).
Tesla's self-driving technology is a joke compared to Waymo's and the Tesla brand is extremely toxic now. I see from your other comments that you're big on Tesla (own several and have a son who works there) but as an unbiased observer I cannot fathom them winning this market.
I have 2 AI4 Teslas with FSD, and I don't find V13.2.9 lacking at all in the Vancouver area. V14 will be a 10x increase in parameters, too. Why do you feel it's a "joke"?
It's a "joke" (I wouldn't call it that, but it's a vastly different product) because you have to pay attention to the road at all times.
You don't live in a Waymo city, so I understand. A lot of people who don't live in a Waymo city don't really get it.
Waymo is a completely different product than FSD. It's a robot that comes and drives you from point A to point B. You can do whatever you want while it's driving, such as take a nap or work on your laptop.
Tesla was SAE level 2 in 2013, and they are still SAE level 2. Waymo's Robotaxis are SAE level 4, and they can drive on public roads empty with no human supervision, both technically and legally.
I have friends on the Autopilot team (and a son). Their goal is by end of year. I've been on HN for 15+ years, and seemingly the only downvotes I get are when I post my thoughts and opinions on Tesla.
1:1 is going to be ruinously expensive. You need three shifts of remote operators. Even in the Philippines or Vietnam, if you can make the latency work, that's prohibitive.
> How do Elon Musk's predictions relate to Tesla achieving a robotaxi service or not?
>
> Ignore his predictions and just... look at whether or not the Tesla FSD team is making progress.
I'm seriously baffled by this comment. How can Elons comments not be relevant? How are you proposing we assess the progress of the FSD team? And why should the assessment be different to the last 5 years where FSD was supposedly ready (according to someone with intimate insight into the work of the FSD team) by the end of the year?
> How are you proposing we assess the progress of the FSD team?
...any metric you want? Miles driven under FSD. Miles driven without intervention. Miles driven without accident. Anecdata from friends of yours who own a Tesla. Whether or not a partially supervised pilot program has been launched in some cities.
If Elon Musk said in 1999 "I think we will achieve self-driving next year", that also has no bearing on whether or not self-driving is achieved in 2025 (in either the positive or negative direction). It only means that Elon Musk's "predictions" can't be trusted as an accurate harbinger of success. Which is precisely why you look beyond his words and at the reality on the ground, which strongly indicates Tesla has made a huge amount of progress in the last 10 years, and could be very close to having unsupervised robotaxi service in various jurisdictions.
You're not "posting your opinions on Tesla", you're literally shoveling them into everybody's throats. You'd be "posting your opinions" when it was one, two comments, and not plenty, like under this news. You're a Tesla freak or fanboy, not an objective commenter.
This sentence was a bit cute: "Waymo has received our pilot permit allowing for commercial operations at San Francisco International Airport." Yeah, that kind of pilot.
I really had to read through it twice to make sure they were just talking about car taxis picking up travelers, rather than some kind of prototype pilotless commuter helicopter or something.
The hard part of automated driving is dealing with all the ground clutter that planes serenely fly over. If pedestrians could charge out in front of a 777 going 650 mph at 34,000 feet... well... we'd be living in pretty different world! And in that world, flying would be much more difficult. Not just for computers but for humans too.
Flying is obviously much harder than driving, but it's a sort of harder that is generally more amenable to automation, though I still think pilots are a good idea because when it goes wrong it goes wrong much worse.
Flying is almost always easier than driving. landing is hard. Bad weather is hard. But just flying - human pilots have napped many times over the years and it only rarely is an issue. Airplanes with primitive autopilot are very good.
Yeah, a primitive autopilot in a plane just needs an altimeter and compass, but a AoA sensor, speedometer, fuel level sensor, and pitch sensor help to detect unsafe conditions like runaway pitch, stalling, overspeed, low fuel, etc. Each of those sensors is providing a simple 1-dimensional data point. Redundancy is relatively inexpensive.
Automatic lane keeping in a car requires cameras that software needs to then analyze to find the lines in the road in real time. But if you want a "set it and read a book for an hour", then you have to respond to other traffic. No longer just some simple PID controllers, the software now needs to plan and execute based on surrounding traffic.
Taxiing is probably harder to automate than the rest. But you could have pilots on hand to taxi to the runway, and take a shuttle to the other end and hop on a just landed plane to taxi to the gate. Or you could use tugs for ground movement.
I'm not convinced - in a commercial airport taxiways are controlled by a ground control systems, not just pilots looking out the window. If the only airplanes around are also equips with the self taxi system they just report position to the central control and that tells them when to go. There needs to be emergency overrides for when that system fails, or a small plane without it is around, but that can be handled by stopping everyone else in the area until the hazard is gone.
There's also all of the service vehicles when you get closer to the gates. The likely damage from an incident during taxiing is much less than during take off or landing, but I think the risk of having an incident is higher and the situation is trickier to manage. And it's super doable to have a pilot come on to manage that, and drop off after the hard part; you couldn't reasonably have pilots do a takeoff and then jetpack over to an arriving plane to do the landing, but it wouldn't be unreasonable for ground moves... similar to canal pilots taking ships through canals.
"The [German pilots'] union said it had carried out a survey of more than 900 pilots in recent weeks, which found that 93% of them admitted to napping during a flight in the past few months."
-The Guardian, "Almost all German pilots admit to napping during flights in union survey"; 2025-09-10
Years as since humans have flown planes stable enough not to need constant attention. On a calm day you don't need autopilot, just set your trims correctly and some airplanes will hold course well enough for a short nap - though of course this is more likely to result in a crash (which likely has happened, though it is hard to guess why a plane crashed beyond pilot error)
Not to mention that almost all civilian planes in the US are required to broadcast a bunch of details that include their coordinates and altitude on a public channel (ADS-B). It's the kind of automated collision avoidance input that you'd probably dream of as a self-driving system engineer. Basically the only thing you'd need to avoid via more complex systems is the odd military traffic, small craft at low altitudes, and birds.
In the abstract yes but in practice the economic (ratio of cost of pilot to pax miles) and safety context of aviation mean fully autonomous flying has to be extremely robust before it has actual utility in industry.
In practice, you're also currently very reliant on infrastructure that is definitely not as solid as you want (eg: ILS and GPS can be interfered with quite nastily).
ILS being under maintenance and unavailable for certain runways is also far from unusual.
On the happy path, yes. Though I don’t think takeoff is automated yet.
Currently we rely very much on the problem solving abilities of human pilots to deal with troublesome situations. Autopilot will disengage in many scenarios.
Drones (both autonomous and remote piloted) have much higher mishap rates than crewed aircraft. Taking off is "easy" until something goes wrong, like a mechanical failure or runway incursion. It's impossible to anticipate and explicitly code for every possible failure mode, so developing autonomous flight control systems that would be safe enough for commercial passenger flights is extremely challenging.
Category IIIC ILS (full auto-land) does exist but requires special equipment for both the aircraft and airport. Human pilots have to actively monitor the system and take back control if anything goes wrong (which does happen).
Garmin also has the Autonomí auto-land system for certain general aviation aircraft which can attempt to land at the closest suitable airport. But this is only used for single pilot operation in case the pilot becomes incapacitated. It isn't suitable for regular flights.
OTOH takeoff and landing could in theory be operated by people on the ground, flying simulator style.
I still believe that having an actual pilot inside the plane that care for his own life is not a bad idea vs someone remote feeling a bit disconnected with the reality of a crash.
Remote piloting is how the military operates certain drones like the MQ-1 Predator. The mishap rate is very high relative to crewed aircraft due to network lag and sensor issues. The military is willing to accept some level of equipment loss in order to accomplish their mission but this would never be allowed for commercial airliners.
The pilot’s self-preservation instincts aren’t the most important reason to have them onboard. It’s that any loss of communication between the ground and the airplane at any point during either procedure would turn it into an uncontrolled cruise missile.
I am not sure why you were down voted. The original meaning of the word pilot is someone who comes aboard a ship for "the last mile" - getting in and out of the harbor and what you are talking about is kinda like that - a person associated with the airport rather than airplane to guide the planes in and out - perhaps using more reliable local communication technology vs what is used to control drones half way around the world.
I have no idea if that works but I thought you were making a good contribution to the conversation by proposing a potential solution to the exact problem everyone is talking about.
When everything is working correctly, no other pilots have emergencies, and no temporary restrictions are in place, and there are no clouds in the sky. Then yes, it /could/ be easier, but almost always it never actually is.
There's a reason the majority of accidents occur during take off and landing.
It depends a bit on your safety standards. There are already autonomous flying things delivering blood and blowing up oil depots where it doesn't matter so much if stuff goes wrong, but to be an airline pilot you have to know how to deal with a huge range of emergencies and systems packing up.
With a car if the engine fails you just pull over. With an airliner it's not so simple. As a result the training for a pilot is much longer than for a bus driver say.
Depends on the size of the plane, really. One of the reasons a few companies were investing in fully autonomous air taxis is because the math on a small piloted aircraft wasn't realistic for a low enough price point to be competitive.
The problem is actually safety. As automated systems get better, the pilot is left with not much to do, and has to maintain vigilance while being really really bored. It is almost better to have fewer automated systems and give the pilot more things to do during the flight so it is easier to keep them paying attention, or all automated with no human pilot to mess things up.
I'm not actually sure how hard landing is. Most airports that support autonomous landings do it by having ILS antennae that guide the airplane to within tens of feet of the runway, at which point the airplane switches to radar for altitude.
Automatic landings started in 1964. I think that it seems hard mostly because of how tightly regulated aviation is - modern technology could probably make things a lot better if people were more receptive to the idea of heavy automated aircraft over populated areas.
landing is easy. the hard part is landing with 20mph cross winds and one engine out (or other mechanical failures). we've had auto-land that is 99% reliable for a while now, but you need to get to 6 9s before you have a system safe enough to replace pilots
I think that as long as the autopilot is able to fly in a crosswind or with an engine failure, it can probably land with one. Autopilots are already able to do these things.
I doubt anyone has tested this in depth, but I'm not sure there are too many configurations of airplane these days where a human can safely land it and a computer can't. Maybe if a big chunk of wing or control surfaces were totally gone, but even a human pilot isn't getting 99% reliability in a situation like that.
In any case, I don't think that the first candidates for automation are gonna be passenger flights. It will probably be small cargo planes first - Cessna Caravans and other turboprop aircraft where the cost of paying pilots is roughly similar to the price of fuel.
Don’t have a ref but heard that it’s been safe for quite a while but they keep the pilots around due to consumer fear rather than actual improved performance. Curious if anyone can confirm.
No. Airliners can't even take off on their own yet, and are only allowed to auto-land with zero visibility at a few dozen airports when the pilots, plane, and runway are all current/recently checked.
Look up the Airbus ATTOL project's first automated takeoff a few years ago.
Also, there's virtually no automation when it comes to interacting with ATC.
An airplane will take off when it is properly configured and it hits a certain speed. It's simple aerodynamics/physics. Pilots are there to react to failures and unexpected events.
There's a bit more to it since you do need to do last bit of configuration (pull up the nose) just as you hit the target speed. But yeah, automatic take-off is quite a bit easier than automatic rejection of take-off.
Even manually pulling up the nose once you reach Vr isn't necessary if you just trim for a little extra nose-up. It'll eventually get off the ground with just enough speed.
Yeah temperature, wind, altitude, weight, runway slope all matter, and then there needs to be enough spare space for the aircraft to successfully take off even with engine failure in the worst possible moment. Then there's the question of fuel consumption too. Takeoff power isn't typically configured to get the aircraft off the ground as fast as possible, but to minimize fuel consumption, while still leaving enough margin in case of engine failure.
It wouldn't be that hard to fully automate a flight from gate to gate when everything works perfectly. But the various failure modes, human error like airport vehicles entering active runway, all that requires human backup. Self-driving car can just stop to the side of the road and turn on emergency lights if its engine fails, with a plane things get much more complicated.
And you don't need rudder input or any aileron input because of crosswind, and other bits that falls into "technically correct but not particularly relevant" territory.
It's fun to see/feel planes do stuff "on their own" (eg making them oscillate, or level on their own, or feeling ground effect, or even your own wake on steep turns) but it's not something you'd want to rely on (maybe with the exception of ground effect on short field takeoffs, but I digress).
> Also, there's virtually no automation when it comes to interacting with ATC.
Check out the Cirrus Autoland feature in their aircraft. They are all small personal aircraft, but the tech is pretty cool. Will talk to ATC and fully auto-land for you in the event of an emergency where the pilot is incapacitated.
"Talking to ATC" is a bit of a huge ask. The system basically just hops on 121.5 (and maybe the nearest/local unicom/tower frequency) and start an automated callout with its intentions that it will be doing. It operates on the assumption that all other airspace users will hear the radio calls and stay clear of the emergency aircraft.
If you can design the product and environment to fit automation, then automation can be quick and effective.
The less you can change about the product and environment, then automation run slower and less effectively.
Air liner operations could be automated, but the minimum equipment list would be more stringent, the destination airport would not be able to take any equipment out of service for maintenance, visibility minimums would increase, takeoff and landing operations would require more slack time.
Besides all of that, the owner of the airplane would still want to have some crew on board.
In short, it's not worth it yet.
===
There is also the paradox of automation: Automation generally makes the hard parts harder and the easy parts easier.
The current goal of autonomy for airliners is single-pilot operation more than full autonomy.
It's very cool stuff, technology wise, with potentially significant redesigns of cockpits, etc.
But the main thing is the plane basically needs to be able to operate just about entirely autonomously (especially during critical flight phases) in case the pilot is incapacitated.
In theory, once SPO is solved, autonomy is almost solved.
I'm skeptical that SPO will be allowed for commercial airliners in our lifetimes. Pilot workloads are fairly low during most routine flights. But when an emergency occurs then the workload suddenly gets extremely high, to the extent that even two pilots are sometimes overwhelmed. This isn't a problem that current automation technology can solve. There are an infinite number of possible emergency scenarios and engineers can't possibly code for and test every one.
Cargo flights over oceans and (mostly) unpopulated areas might be a valid use case for SPO. Cargo pilots have always been considered somewhat expendable.
That “just” is doing some heavy lifting! The car still has to deal with all the normal hazards of the road while pulling over, plus the hazards it is itself creating by acting abnormally.
Well if we're being picky, technically the car itself doesn't have to deal with the hazards it has created, rather everyone else does.
The point is you can't just "stop" a plane and wait for someone to figure things out (https://support.google.com/waymo/answer/9449023?hl=en). Whatever the difficulties in dealing with an abnormal situation in a car, it is strictly much more difficult to deal with them in a vehicle constantly fighting the homicidal urge to fall out of the sky.
Most carriers have a rule that on clear days you always hand fly the landing.
This is a competence you do not want to lose.
It's also the case that you can have a whole approach setup in your flight computer and at the last minute the controller gives you a runway change. You could drop your head down and start typing a bunch info the FMC but you're generally better off just disabling auto pilot and manually making the adjustment.
But two interesting data points from the Wikipedia article I linked are that the first aircraft certification for ILS Cat III was in 1968, and Cat IIIB in 1975.
And IIRC by the 1980s, autoland was already a pretty common feature.
Yes, but autopilot usually just keeps the plane flying in a straight line at some specified altitude, which have been around since 1912. It isn't full self-flying (although we definitely have drones that can fly themselves already, so that tech already exists).
Auto-landers are not simply classified with autopilots. An autoland system is an advanced function that is part of a modern aircraft's overall autopilot capabilities. A basic autopilot can control an aircraft's attitude and heading, but an autoland system can automatically execute the full landing procedure.
The setup at SFO is currently quite annoying (Lyft/Uber require you to walk 5 mins to the garage roof, and drivers need to park/wait 5-10 mins away, so there's always a substantial delay). Taxis get the privileged parking spot immediately outside arrivals, but if it's busy you might still need to wait a bit.
I've been wondering for a while why Waymo can't offer a semi-managed solution to SFO to dynamically manage load, have just the right volume of cars inbound, maximize parking utilization, etc. with all of the nice intelligence that an app-based system would enable.
It feels like you should be able to have a buffer of cars waiting right at the curbside, and automatically refill that buffer on short notice depending on observed or predicted demand.
As a European, I can’t help but feel a bit sad that we’re missing out on the driverless side of things. It seems like most of the meaningful deployments are happening in the US (Waymo, Cruise).
I’d really like to see either a Waymo competitor emerge in Europe, or even Waymo themselves operating here. The regulatory environment is obviously more complex, but it’d be great if we didn’t end up years behind on something this transformative.
Cars of any sort, self-driving or otherwise, do not solve traffic any more than Uber does because you need to have enough of them to get everyone to and from work at basically the same time. Trains are the only way to address traffic. Trains are self-driving. Europe already has the better self-driving system. It's just boring because self-driving is much easier when you build the road to support it instead of removing all constraints and adding GPUs, lidar sensors, cameras and an army of fall-back operators in overseas call centers.
They replace taxis and potentially postal and trucking applications in future.
It’s certainly not a replacement for mass transit. US is sparsely populated compared to Europe and mass transit don’t work as well in the suburbia. That said, I do see many transit oriented development in SF Bay Area where high density buildings are being built near transit stations.
Can you imagine how much traffic there would be if NYC didn't have the MTA? The principle of induced demand tells us that as long as there are roads they will have roughly constant traffic because people are willing to spend some roughly constant amount of time getting to and from destinations by road each day. More roads speeds up everyone's commute which brings in more drivers, which brings traffic right back to the baseline terribleness.
The question is how shitty it would be if they also had everyone on them who's currently on public transit.
So basically, it is a traffic-free panacea for everyone who chooses to use it. It's not a goal of trains to eliminate traffic for everyone who insists on driving.
The induced demand argument works for trains too. If NYC didn't have MTA (no subway, no LIRR, no MNR) then the population of NYC would probably be 1% of what it currently is. Building more train tracks and having better train services also encourages more people to move to NYC so that these new train services become more utilized.
Neither roads or train tracks solve the traffic problem.
Yep, this is a good point. There are appropriate technologies for each situation. It's not a winner-takes-all contest.
For another example, can you imagine trains replacing school buses in a large, rural school district? Sometimes (not always), buses are better than trains.
Any one part would have the about same amount of traffic it does now. It would just sprawl out bigger across adjacent counties and the highest density parts would be lower density.
This is what bikes and busses are for, or just walking because the metro system is comprehensive enough you are at most four blocks away from a station.
You could maybe have something like Zermatt Switzerland which is car free but you can get around in human driven golf cart like taxis. It's pretty pleasant but expensive. If the carts were self driving it could be cheaper.
Zermatt is fundamentally a pedestrian town. There are a limited number of permits for electric vehicles available for companies that have an objective need for a vehicle. That limited availability makes the electric taxis expensive.
The total number of permits seems to be around 500 in a town of 5k permanent residents. And the population grows to 30k or 40k during the peak tourist season.
Trains will fairly unreliably take you from one place that is not your home, to another place, which is not where you want to go, at a time that is probably not exactly when you wanted to arrive. Freedom of movement is incredibly important, and trains are very rigid in this aspect.
Well That’s certainly not been my experience when visiting Europe. In fact, it many cases it’s been the opposite - having a car would have been restrictive in any major city and a source of friction.
Well to the extent it draws people from public transit, yes because traffic makes being a pedestrian more unpleasant and waymos still are traffic. And increased traffic adds friction to crossing streets and they park obnoxiously, among other things.
So yes, they would be obnoxious at any significant quantity and also not really help with getting across the city since transit is pretty good
Well yes if we’re arbitrary limiting our choice to car based transportation that makes sense for mild climate cities. But why are we insisting on cars being the backbone?
No limits. Each option should be evaluated on its merits.
My contention is that in US cities the high cost of existing rail makes it uncompetitive for most uses, and there is no justification for building new rail.
This is quite the "I have never lived anywhere else other than North America" take.
Rail and other public transport in pretty much everywhere in the world are designed to serve commute first, tourist stuff second or third.
Public transport isn't just having some trains, or having only trains between major cities. It is designing whole commute routes from various urban and suburban areas to workplace. There needs to be regional and suburban links that arrive to metro and tram stations. Metro and tram have to operate very frequently to handle commuters. The frequency of the trains should adapt to the commuters in the morning and evening. They need to be convenient, clean and safe too.
Cities around the world are also much better balanced than NA ones. The workplaces and living areas are almost always mixed rather than having a "downtown" area where every office worker travels to. My area has many buildings with a supermarket, apartments and small offices in the same building. There are two car factories in the city next to one of the biggest urban parks.
> Trains are great when going to tourist attractions, especially in the center of old cities. When you live and work in a city, they're much less practical.
This is the most "tell me you live in America without telling me you live in America" thing I've seen in a long time...
America basically the only place in the world where in its cities, trains and other public transport aren't a major part of people's lives. In other places (Seoul, Tokyo, many European cities, etc.), even people who own a car will sometimes commute via train due to the convenience.
Try a bicycle or a stroll instead of embracing the WALL-E.
If you feel that way about transit you may not have tried a good transit option like Hong Kong MTR with 90 second headways and travel from and to substantially everywhere you want to be.
This is based on my personal experience, I used to ride trains for travel a lot. I grew up in Europe and lived there for 31 years so this is not based on ignorance.
Buddy the tube seldom fails for that reason either. Plus some self-driving sauce would reduce their hours to 0. Certain lines in London like the DLR are already driverless (Grade of Automation 3). Most of the other lines are GoA2.
Trains still don't solve last mile transport for most people (even in places with robust transit systems)
Self driving cars might not solve traffic problems but they could greatly reduce them. Problems like traffic waves and gridlock go away when all cars are driving themselves.
The last mile problem is only a problem because of poor layout. Build homes and work near transit nodes (instead of in the middle of nowhere) and there isn't a problem in the first place.
> Problems like traffic waves and gridlock go away when all cars are driving themselves.
How would that make those problems go away? It could probably slightly alleviate them in marginal cases, but any given road has a finite throughput limitation, and once it is reached, it wouldn't matter even if every robo-driver were perfectly synchronized.
> Trains still don't solve last mile transport for most people
This has not been my experience since moving to Manhattan last January. Subways, alone, close the gap between regional rail and most destinations astoundingly well. I haven't yet needed to use a bus (but they seem abundant, too), and I haven't even thought of taking a taxi yet.
Here, robust transit has solved the last mile problem for most people.
Here, robust transit has solved the last mile problem for most people.
There are huge gaps in subway coverage in New York. Manhattan, especially Lower Manhattan, is the exception here. Go to the outer reaches of Queens and see where the subway gets you. Try to go between (or sometimes within) boroughs.
The last mile is a solved problem. Most people can walk (and many of those who can't would need human assistance anyway). And then there are bikes, electric scooters, and other light vehicles that use space much more efficiently than a car.
Self-driving cars may help with the actual weakness of transit, which is the long tail of trips. Trips on routes with too few passengers to justify good transit service, and with the trips too long for the last-mile solutions.
Walking a mile with groceries or a baby is common. People in less car-oriented neighborhoods typically do quick visits to a grocery store when it's convenient for them several times a week, rather getting a week's haul of groceries in a single visit.
With furniture, you usually pay for delivery. Especially because the furniture store probably doesn't have the items you bought on site anyway.
Ideally: there's a train close enough to walk, or a bus or tram that's nearby that runs frequently, is clean, and doesn't get stuck in traffic because there's not much car traffic.
Slightly more realistic: enough people can and do cycle or walk to the train that pressure is relieved on the roads for those who cannot cycle or walk.
For the consumer, maybe not, other than a delay of some years.
In terms of having the industry? Absolutely. How many other areas of "tech" has Europe basically punted on and ceded to Americans? Currently there's some gnashing of teeth across the pond for how there's no real European equivalent to the big US cloud providers (AWS, Azure, GCP).
There doesn't have to be an equivalent of everything, I wouldn't want to use US cloud because of price and governance. At most I use the "cloudy" services and rent "capacity" from a European provider, companies are fleeing the cloud. They're done subsidizing Amazon deliveries.
MobilEye and Mercedes works on self-driving, so does BMW. It's probably not Waymo quality, but just because there aren't cars on the (wide and car friendly) roads doesn't mean nothing is happening.
Meanwhile Europe has solid infrastructure for electricity (esp France), ASML has no competition, Carl Zeizz is world leading in optics, there's probably a Leica LIDAR in the Waymo cars... I mean while we're throwing pies and bringing up other markets..
My old boss was working on a project with Leica where he was working with some partner on self-driving industrial machines, they we're using Leica gear for collosion avoidance and such.
Europe doesn't need self-driving cars, we have alternative modes of transportation. Where it's needed (mines and industry) it's already there. And whatever modern car you're driving here has ADAS which helps make driving comforable.
USA is huge.
This is happening in a small part of the USA in a very limited fashion.
It's not like the USA has driverless cars everywhere, 99.9% of the population never saw one.
I'd guess Waymo covers 5% now. San Francisco, San Jose, Los Angeles, Austin, and Phoenix are ~10% of US population. Waymo service areas don't cover all of those cities.
Considering tourism and people living just outside service areas who see them but don't get to use them (which includes me sadly) I would not be surprised if 10% of population had seen at least one.
>pilots of self-driving taxi- and bus-like services will be brought forward by a year to spring 2026, attracting investment and making the UK one of the world leaders in this technology
I'm wondering how self-driving cars will solve the priority problem of narrow streets of UK towns where drivers need to let each other pass all the time.
I've wondered that myself. It seems quite challenging for human drivers at times. Around Ladbroke Grove you quite often get some complicated jam with two busses and about ten cars stuck.
Yeah I'll believe this when I see it. Most UK roads are significantly harder to drive on than anything in the US. That's why they always test these things in Milton Keynes.
Also a lot of UK driving requires communication with other drivers (letting people out, etc.) in a way that US roads don't. I'm not sure how driverless cars can handle that.
I really wish we could get them, because they're great. But I'd say we're talking 10 years behind the US simply because of the extra engineering challenge.
> I’d really like to see either a Waymo competitor emerge in Europe, or even Waymo themselves operating here
I think you’ll see American and Chinese self-driving kit in Europe once it matures. It’s just easier to iterate at home, so while the technology advances that’s where it will be.
Maybe there just not enough interest? After all there is good public transportation (especially rail), increasing biking habits and just loving the driving experience.
European cities have lots of taxis. Same with Asian cities. They will obviously have AVs in the future. I'm not sure why you think they should be mutually exclusive with transit.
Many American cities don't have the population density to make metros and trams economically viable. And those few cities that do have comparable density (New York, Chicago, namely) do have metros.
Public infrastructure has high overhead costs, and low population density means there isn't enough ridership to make it viable.
People in the thread are asking why people ride Waymo to SFO, which is well outside San Francisco proper. Thus, the whole peninsula's density is what's relevant.
The US does not have many metro areas with population densities above 3,000/km2. And those that do, like Washington D.C, NYC, Boston, Chicago, do have metro systems.
American public transit construction costs are now ridiculous in terms of both money and political capital. Even somewhere as sprawled as San Jose now requires well over 1b/mi to build a subway under; BART could've acquired an entire autonomous driving company for the cost of the Silicon Valley extension.
…What? What sort of terminally online strawman would be spending his free time “virtue-signaling with Europe” to some anonymous bozos on a tech forum? What a dull and intellectually uncurious reply.
I think self-driving cars may eventually become common in areas where cars are currently common. I think public transit will continue to dominate in parts of the world where it currently dominates, because it is simply a superior user experience for the majority of people when the government cares to invest in it. (Not to mention far cheaper and more egalitarian.)
I am conveying my lived experience in most European cities I've been to.
A superior user experience is going exactly from where I am to where I want to be safely, quickly, and affordably. Self-driving cars are looking really good for those criteria.
$20+ per ride is affordable? Waiting 10m+ for your ride and slowly sifting through traffic is quick?
In London, Paris, or St. Petersburg, I pay a few bucks to hop on a train that runs every few minutes and rapidly end up across town, roughly in the area I need to be. It's literally the cheapest and fastest way to get from point A to point B, not to mention tested at scale and thoroughly battle-hardened over the course of a century.
Not every city has this privilege, of course, but surface trams are 80% of the way there, especially if they have right-of-way. And they don't make pedestrians' lives a living hell: https://www.youtube.com/watch?v=bNTg9EX7MLw
The core for a good experience is a good structure.
In many regions of the U.S. people live too far apart, shops and businesses are zoned apart into wide spread business areas. Public transport won't provide a good experience.
In a notable part of European cities people live in denser quarters, where a "third place" is reachable in walking distance, some degree of shipping, doctor visits, work are close by. There public transport can fill the gaps for the remaining trips in an (space) efficient way. Self driving cars however would clog the area.
Adapting US settlement structure to allow public transport won't happen. However a self-driving car can turn the dial for individuals to move out of the urban European area into more rural areas. Question is how big that group is.
Good. Cars ruin walkable cities, and the last-mile problem can be solved in other ways.
And it's not just the EU. I'm sure that e.g. China and Japan will continue to invest in their excellent public transit infrastructure even when there are more self-driving cars on the road.
Much of Japan's transit infrastructure is private. There's nothing special about transit that means the government has to own it; being a government, it can regulate things without owning them.
Americans have this idea that transit is for poor people, which translates to "it's not important for transit to make money", which translates to "we need to make it illegal for transit to possibly try to make money", so there aren't even vending machines at the platforms. Whereas in Asia they do profitable land development at the transit stations.
> Much of Japan's transit infrastructure is private. There's nothing special about transit that means the government has to own it; being a government, it can regulate things without owning them.
Japan's private transit infrastructure is only private in high-very high density environments (inner-city) and subsidized in low-density environments (rural, cross-country). Ultimately private group transit requires population density above a certain threshold to be viable.
Don't worry, we're missing out on a lot of "progress" on this side of the ocean thanks to Trump's dislike of wind farms and RFK Jr's whole anti-vaxxer thing
One thing you are missing out on: mandatory loud (97 to 112 db) 1000 Hz audible beep when the vehicle reversing, oh so slowly, such as at the recharging station. Also, constant shop vac five horsepower vacuum cleaner sound. BEEP BEEP BEEP BEEP BEEP. VROOOOOOOOOOOOOOOOOOOOOOOOOOOOM
Oh wait, you thought those would be in the middle of nowhere? Nope.
These backup warning systems operate at approximately 1,000 Hz, producing sound levels between 97 and 112 decibels.
Santa Monica’s municipal code adds another layer of complexity, prescribing exterior noise limits of approximately 50 decibels during the day and 40 decibels at night.
The continuous operation—with vehicles reversing dozens of times hourly, including during late-night hours—continues to challenge community peace.
So, constant car screaming BACKUP BACKUP BACKUP I'M BACKING UP HERE right outside your residential window. Kinda sucks. A whole lot.
I'm surprised and incredibly impressed at this announcement. It seems trivial, but the general feeling in the industry has been that SF would fight tooth and nail against robotaxis at SFO.
What likely happened now is that SFO got a kick up their backside from the Mayor after the press started asking why it was still dragging its feet, while SJC approved Waymo swiftly.
Recent changes in the composition of the San Francisco Board of Supervisors (i.e. Peskin being out of government) may have something to do with it being easier than expected.
Waymo got approval for SJC last week. That probably accelerated approval for SFO, which had been stalling. Nice.
When they get clearance to drop people off at the main terminals, that will be more convenient. Pickup at the terminals is harder. There will be a need for a staging area somewhere in the parking structures.
Few major airports I've been to allow Uber/Lyft anywhere near the pickup area, so many fliers are already accustomed to walking a quarter mile or so to their rideshare. But their inability to use the drop-off area is a new inconvenience, and I can see it limiting the appeal.
Waymo will probably get access to the drop-off area after a while. One step at a time seems to be the Waymo way.
Waymo at airports could work really well with automatic dispatching. They already have an app running in the customer's phone. It should be aware of when someone with a reservation gets off an airplane, and how close they're getting to the pickup point. With good coordination, as the customer heads to the arrival lanes, a Waymo pulls out of short-term parking and heads for the meeting point.
A few more years, and humanoid robots will put the luggage in the trunk.
I'm talking about Uber/Lyft drivers being required by many airports to pick up away from the normal pick-up area, usually down the road a bit or in a parking garage.
I drove for Uber/Lyft back in 2020 and let me tell you, SFO is a nightmare. I missed a turn once and had a passenger trying to make a flight furious at me. I quickly figured out there were a group of drivers who specialized in SFO and amatuers like me should avoid the place. When Waymo announced San Jose I thought ok, that makes sense because SJC is easy, but SFO? Wow, I'm impressed. I hope it goes to plan.
Nothing more rewarding than a company working hard and seeing real-world, first of its kind results in action. Makes me feel giddy about a company again like peak tech back in the 2010 era.
Congrats to the Waymo team, I’m sure this was a huge milestone internally.
I'd hope so. As an aside, I wish Waymo was more transparent on the app that their cars are not allowed to take passengers on the freeway. I was unaware of this restriction when I booked a ride from SF to Burlingame last month and I was stuck in a Waymo for an hour going down residential streets!
I wonder how that'd feel. I took a Waymo in SF last fall and I was pretty impressed. But it was also slow city speeds. I wonder if it feels different going at freeway speeds with "no one" at the wheel.
While the margin of error is much lower on a freeway due to the speeds, other drivers are generally a lot more predictable (also in part due to the speeds).
Sure - a good freeway is actually a lot more predictable in most circumstances than city driving, so as a problem to solve it's likely a little bit less complicated. What I wonder about is what it feels like as a passenger. I wonder if it would be more or less frightening than being a passenger when my 17 year old is driving.
I use adaptive cruise control a lot, where I rely on the car for keeping a safe distance.
I have a limited version of SuperCruise which means it operates hands-free on freeways but nowhere else. My wife's Equinox EV has the regular version, which operates on a lot of arterials near us and has more capabilities. The first time that the Equinox signaled, changed lanes to pass, signaled, then changed lanes back was shocking.
We moved to a small town and drive a lot more than we used to and I find that having those capabilities really helps relieve the stress.
I will say that I move to the center lane when going through a notorious set of curves on I-5 in Portland because my Bolt doesn't steer as smoothly as I'd like near the concrete barricades. I wanted SuperCruise because it has a fantastic safety record. There are lots of times it's not available but when it is, I have near-total confidence in it.
I took a Waymo that drove on an 'expressway' which had a speed limit of 40mph and it was definitely a different feeling. I did feel a bit scared, at 25mph it feels like a gentle theme park ride, at 40mph it's beyond that and feels dangerous.
There was a good overview on here a while ago about the challenges[1]. You need to plan longer in the future and your sensors need to reach further. It's also a much bigger challenge to collect sensor data as fewer diversions happen per mile (but those that do have higher stakes).
Roads that get used more collect more debris. They also break and require maintenance more often. That maintenance is exceptionally disruptive to the normal operation of the road.
Other drivers aren't your only challenge out there.
From the article “ Pickups and dropoffs will initially start at SFO’s Kiss & Fly area – a short AirTrain ride from the terminals – with the intention to explore other locations at the airport in the future.”
Oh, this makes a bit of sense. The Avis/Budget fleet team will be part of managing the vehicles, so they can be quickly cleaned and fueled up when they slide into the airport, too.
Same. I go to the rectal car center at least 4 times each year. I just was there on Saturday and had no idea either. Still don't know what it is other than Waymo pickup.
I did always find the term kiss and fly confusing and weirdly intimate, as if everyone is getting a ride to the airport from a spouse or parent. Definitely a throwback to another era.
Hopefully Waymo does a better job than SF Uber drivers. I can't tell you how many times I've had drivers make a wrong turn trying to find their way to the pickup point.
There's the big sign there telling you to go to arrivals for drop-off. This is probably a stupid question but can Waymo cars interpret those temporary display signs and follow them? Would it?
It seems to handle the standardized ones (think "construction ahead, detour left") perfectly well from the rides I've taken, but there's all sorts of ways they could be 'cheating' on that.
I'll be honest, I think LAX's traffic is better than SFO's.
It feels like there's a lot less spaghetti at LAX, the shortcuts are reasonable, and you don't have separate international and domestic loops.
The Big Dig, for all the digs it rightfully got for taking forever and costing a shitton, actually does the job it's supposed to (mostly). I'm generally pleasantly surprised how few problems it had when I lived there.
I'll take JFK over LAX. The construction going on right now at JFK sucks, but LAX is comically bad. Just last week I was on a rental car shuttle at LAX and watched 3 separate groups of people at different terminals miss their flights because traffic just wasn't moving.
Maybe, but this approvals only allows them to go to the rental car center, which is quite far from the terminals. The passenger will need to take the air train to the terminals.
Can you handle parking structures? I heard a lot of the autonomous cars were using 2D maps and couldn't handle multiple levels. Haha! This was just a year or two ago.
Google maps has been able to figure out parking structures for me recently. Not sure what technology is involved (gps isn't great for vertical) but it's clearly possible.
Google has been collecting data on building interiors for several years now. Not just parking structures. This data is currently used in streetview. Google's geospatial data is unequaled and maybe a bigger advantage than is readily visible.
Do they need a "map" of a parking tower though, just like how humans don't exactly need Google Maps inside of one? I feel like this is something self driving + vision (exit signs and arrows) can handle
Not sure if you have a recent side-by-side example with Uber, but this seems like it would have to happen if the demand is there. How else can you offer a quality product (i.e., car shows up in a reasonable amount of time) if you don't have enough cars to satisfy the demand? Pricing is the primary demand lever.
There's so much polarizing opinion on Tesla's offering and whether they'll get to Waymo's level sooner than later, but this seems like it's going to be or already is a huge issue for Waymo where they can't manufacture the vehicles fast enough to satisfy the demand as they expand both locally (because they capture more of the market) and into new geographies. Will they try and acquire a manufacturer? I don't think that's economically feasible for Waymo (Geely market cap is $25b, per Google snippet fwiw), and obviously being in the car business is different than autonomous, but I'm sure Google would bankroll a purchase if they thought it was the right growth strategy.
I guess Tesla, even if their autonomous is on par with Waymo tomorrow, also has to manufacture the fleet, but it seems extremely beneficial to have that capacity in house vs. relying on partners. Maybe I'm wrong and it's not that much of an advantage, but at first glance it would seem to be.
The CEO of Uber was quoted as saying Waymos complete more rides per day than 99% of Uber drivers. He didn't give a precise ratio but this makes me think that hundreds of Waymos can replace thousands of Uber drivers and their cars.
CMs like Magna have the flexibility to manufacture, at the low end, hundreds of vehicles, and at the high end thousands. I doubt Waymo will ever make their own vehicles. They are already working with Toyota on adapting Waymo technology to privately owned cars. That implies mass production. That would be a supply of vehicles that are probably simple to adapt to robotaxi use.
That's a crazy statistic and an interesting one for him to actually say out loud. Was that in the context of Uber partnering with Waymo in Austin? And thanks for the insight on the manufacturing side. Sounds like it might actually be to their advantage to use third parties because you can spread the demand around and since auto margins are not high the added cost for that benefit is minimal.
Private cars have a ridiculously low duty cycle. They mainly sit around waiting for their owners to use them. I suppose at some point in the future there might be a traffic jam of autonomous vehicles, but only if the providers are antisocial and don't coordinate ride destinations and routes.
@harmmonica I do. we prefer to use Waymos in SF but Uber has been a lot cheaper in the last six months or so regardless of time of day...
Also saw some Zoox self driving boxes on the las vegas strip last week but no one seemed to be using them.
Thanks for sharing, oliver. For anything local I've almost entirely switched and I guess I haven't been doing much price comparison between the two. One thing I have noticed here in LA, albeit only a couple of times, is that during rush hour the waiting time is significantly longer for Waymo. I've taken some Ubers because the wait for a Waymo has been way too long.
Not sure why you're downvoted. I've tried Robotaxi a few times and has been great. They still have a safety driver these days and wait time is a big high though.
It will remain higher for a while. From reporting I have seen, they are close to maxing out their vehicles, and many people prefer it to other options, so are willing to pay a premium. As long as that is true, it's going to be priced as a premium product. It won't be until fleets grow significantly in size and/or another driverless taxi service enters the market that we will maybe start to see prices driven down closer to marginal cost of a ride.
-edit- multiple other comments apparently disagree with this. I'll defer to people who actually use them over the reporting. Odd that there is that disconnect though.
Yeah they need scaling and competition before the prices get lower. As long as supply is saturated with demand and nobody else is on their level, there's little reason to lower prices.
Yeah, and just to add even though it's implied in your comment, there's plenty of reason to keep prices where they are independent of a desire to increase revenue. Customers will not wait forever for the car and so if the demand is high you have to keep the price high to discourage people from using it so wait times remain in check. Tricky tightrope they're going to be walking while they optimize the fleet size for local adoption and geographic expansion.
It's also higher right now because it is a novely. Plenty of people are booking it just to say they rode in a Waymo and take pictures. When that wears off they will have to start competing strictily on price and wait/ride time.
To be fair, Waymo claims to not record or transmit audio without you either manually engaging such (by requesting support), or a very unambiguous announcement (presumably when the car gets into some sort of emergency situation). And lying about that claim would probably run afoul of California's 2 party consent law. So still a step up in privacy versus having someone in the car listening in on your conversation.
That said, even if they were listening to you, there's a lot of things that are completely inconsequential from a perspective of an anonymous call center employee far away listening in on, that I probably wouldn't want to talk about in front of a taxi driver.
I know this is somewhat besides the point of the discussion, but.. many Ubers have recording devices inside the car too. The drivers have gotten savvy and protect themselves from false claims or even harassment.
This is just me, but maybe helps explain it. It's not that the presence of a driver is bothersome, but in the pre-Waymo world your interaction with the outside world starts when you step out the door of your house. Now the interaction with the outside world starts when you get to your destination and step out of the Waymo. I really enjoy the outside world, mind you. But it just feels easier to traverse my local area in solitude and with a consistent and comfortable vehicle, and non-erratic driving style.
I imagine how nice it could (will?) be when you can hop into a self-driving car for a longer ride or even a road trip. I think you'll feel like it's an extension of your living room vs. being in a car.
If you step back do you really think that's indicative of a mental disease? Does it make any difference to you that many times I'm taking a Waymo to go and hang out with friends? Not much of a stretch to say it's allowing me to socialize more because I don't have to worry about my meter running dry, or having one too many drinks to drive myself home, or being able to move around from area to area in comfort. And if you say "you can do that with an Uber too!" it's true! But does it really surprise you that someone would want a car that drives calmly, obeys all traffic laws and gives you a little downtime from the outside world pre or post the activity you'd been doing before stepping in the car? Does that really rise to the level of mental disease?
It seems like a huge catastrophizing stretch to get there based solely on preferring to be in a Waymo rather than a taxi or Uber.
Well on second reading your comment reads like an ad, given advertising isn't natural either, you can understand the confusion
But I was referring to the wanting the outside world to resemble your house and to have little interaction with humans. No, that's not normal, despite any sophistry or ad speak
The problem isn't when they don't talk and just drive, the problem is when it's late at night and the passenger is a woman who is inebriated. Not having a driver entirely makes that much harder.
On other threads I've seen conflicting anecdata regarding pricing being higher or lower than an Uber ride. That's not too surprising since the supply and demand variables are going to be different for Waymo.
In my experience so far, Waymo costs about the same as an Uber when you take into account tipping, but takes longer (they're not yet doing freeways). With the addition of SFO to their zone, I can't imagine freeways are far behind, because getting from the city to SFO without using the freeways would be... a novelty.
That's not been my experience... 90% of the time when I check, Waymo is still a good 20-50% more expensive in SF, when comparing to a tip-included Uber or Lyft price.
I've used Waymo countless times in SF. It's typically 15% cheaper than an Uber/Lyft and trip time/wait are generally the same. I much prefer the Waymo.
They still have to compete with alternate modes of transportation such as buses, bikes, trains, e-scooter rentals, self-owned cars, Uber with human drivers.
If it would be "too much", then there's no reason why taxis (incl uber/lyft) wouldn't be too much today.
Direct competitors are uber and Lyft which they can undercut since they don’t pay drivers.
The people who want to take buses and trains will continue to do so although Waymo might sway some with their ease and if pricing is reasonable.
Bikes and e-scooters only get you so far. Last time I was in SF I didn’t see too many bikes but I saw a ton of e-scooters. Are you really taking an e-scooter further than a few blocks? And when it rains?
Self owned cars make sense for longer trips out of the city but parking is a pain and driving is stressful so this is an easy win for Waymo.
It’s cheaper now so they can take market share. And their cost will certainly be lower than Ubers so they can win the pricing battle. But long term monopoly gonna monopoly. Perfect pricing is a given with the wealth in SF and how many rides will be on a business CC.
from what I heard, the intention is to make it much more affordable than it is now. I don't remember the source right now but I did think it was a blog post or something like that.
I think if it's affordable then people will easily take that. instead of drinking and driving at night or other unsafe activities. if it's affordable then people can just take a waymo home and then back again to get their car when it's safe again.
Certainly they aim to make it affordable now in order to undercut Lyft and uber. Long term they will own the market and jack up prices as monopolies do.
L5 means the car can drive everywhere a human can. Waymo's refuse to drive outside of a constrained area, and occasionally stop to ask for assistance, so that makes them L4.
This whole autonomous driving levels kinda muddies the waters. Some would argue this isn't full L4 even. But it is a self driving car in the places it offers its services.
I think there's an implicit "where a decent human driver could drive safely" for L5, otherwise you get increasingly ridiculous scenarios like, "can Waymo drive safely in a whiteout blizzard?" or "can Waymo safely escape an erupting volcano??"
haven ridden in both a few times, yes, Waymo is head and shoulders better. It's smooth and I don't think I've ever seen any false alarms or behavior that made me feel unsafe in a Waymo, while I've had a few scary or annoying situations in the Teslas. I took a 6-minute robotaxi in drizzling weather where it parked in intersections twice because the cameras were obscured. Meanwhile Waymo can drive perfectly in heavy fog.
Both the Waymos and Teslas have that central display that shows you what the car sees (pedestrians, dogs, traffic cones, other cars, etc). The Waymo representation of the world reaches pretty far is is pretty much perfect from what I've seen. Meanwhile the Tesla one until recently had objects popping in and out.
Neither is perfect, of course; both will hesitate sometimes and creep along when (IMO) they should commit. But they're both still way better in that regard compared to the zoox autonomous cars I see in SF.
Tesla doesn't have a real robotaxi yet, they're still in the testing/prototyping phase where they need a safety driver or safety monitor in the car.
They might be close to a real robotaxi in some areas, but it's hard to say until they actually pull the trigger on removing any employees from the car.
An interesting thing about this is that there are fewer than 1000 Waymos in the SF service area. I don't know today's total, but I'm pretty certain that there are fewer than 5000 Waymos in existence. Maybe as few as half that.
Some months ago Waymo claimed to be providing 250,000 rides per week. If the fleet size was 2500 at the time, that would be 100 rides per vehicle per week.
What’s special about the airport is that the City of San Francisco owns and regulates it (as opposed to the streets that are regulated by the state CPUC), and the Board of Supervisors previously were regulatory captured by taxi medallion owners and Teamsters union (https://missionlocal.org/2024/12/waymo-rolls-toward-san-fran...). Specifically, Aaron Peskin (BoS supervisor from 2001–2009, 2015–2025, and board president for the last 2 years) said, “Their entire M.O. is, ‘The state regulates us; we don’t have to work with you, we don’t have to partner with you.’ My response is: There are things we do control. Including where you charge your cars. And the airport. What I intend to do, is condition their deployment and use of the airport property on their meeting a number of conditions around meeting this city’s minimum standards for public safety and transit.” https://missionlocal.org/2023/11/waymo-rebuffed-by-sfo-sf-gu...
I’d say it puts a lot of Uber (and similar) drivers at risk because airport rides are a good source of income. Waymo undercutting them will reduce the amount of passengers available for pick up.
Not saying it’s a bad or good thing. Just that it has real world impact on people and the economy.
Usually you'd have to take the BART one stop then the waymo, which seems to be a common tourist attraction for fresh deplaners. Perhaps the airport was afraid without that step of friction, too many people would try this and cause a waymo-jam
Inside SF, my experience is that Uber and Lyft are ~10-15% cheaper than Waymo, but that's before tipping. I don't have to tip a robot, so they work out to nearly identical prices.
The inherent problem there is the edges, most food delivery isn't the trip, it's the person getting out of the vehicle and putting it on your doorstep or going through the building. Zipline and their droneports for buildings seem to have the better solution, at least until waymo has some sort of legged robot that can bring the bag the last meter(s)
I think the frustration with tips is so prevalent that the advertising could just be "Skip the tip, simply walk to the street to pick up your order!"
Would work great in suburbs where a robot car could pull in front of home for a minute or two, your food will be bid to another customer if you don't pick it up in 5 minutes. maybe the little robots in NYC are better.
I would argue that the sidewalk robots are too hard to coordinate and not strong enough to hold up against crime, the solution is somewhat closer to my other comment below, a vehicle with maybe 4 or 8 food cells that can fill up at various locations then make its journey around the city. At that point the problem would be idle timeouts and how to handle disgruntled consumers that lost their window for pickup
Aren't the "first meters" also pretty problematic? Are Waymos going to double park in front of a restaurant waiting for someone to come out and put the right order in the right vehicle?
That's easier to do with training, and the business is usually more willing than a consumer as it increases their business. Anecdotally, see how many of them (at least locally to myself) have adopted the doordash/grubhub tablets in their kitchen ordering system. I imagine it would be a co-packing situation with lockers on wheels similar to the vehicle KFC uses in China: https://www.mashed.com/284555/the-futuristic-way-kfc-is-sell...
My anecdotal evidence also has so many incorrect orders that I'm a wee bit less optimistic than you about restaurant-side human handling of the first meters. :)
Nuro is an independent company from Uber, the latter just has a partnership with and some investment in the former. Uber has similar relationships with more than half the industry at this point.
I have a relative in Texas who is looking into leasing a drone to operate for food delivery. Apparently, that's already a thing there? If we could get food/small packages delivered to our building's roof instead of the front door, it would be a huge win for everyone in the building.
The longer the route, the harder it is for the food to stay fresh and warm/cold/frozen. It's a trade-off between efficiency, price, and customer satisfaction.
Self-serve ordering terminals already often ask for tips. Presumably to be legal they're being paid to the kitchen staff, but I think sticking to "tips are for workers who have to pretend to like me" is a pretty firm boundary to stick to.
(Also, arbitrarily reclassifying things as tips is hard, because legally 100% of tip revenue has to go to workers, not management, and certainly not the company's investors or coffers).
Tax-free tips paid to robots go to the hardworking AI engineers -> AI engineers voluntarily donate part of their tips to a 501(c)(3) that helps support struggling venture capitalists.
Something like that. We'll work it out the details once the right PAC donations are in place.
They aren't the only autonomous vehicle for hire service. Zoox is operating in Vegas.
Even if they were the only one, it would be odd to classify autonomous rideshare as a distinct market given they compete directly with other vehicle for hire services where they have nothing close to monopoly-like power.
Okay what're the odds on how long it is until there's a stray Waymo on the tarmac. Hopefully with enough warning to divert any planes about to land on it.
Cause what this country needs is to automate away even the gig economy jobs that are out there. Let's keep making a few people rich and screw all the normal people out there.
Why the downvotes? That jobs will be lost is fact. Does this represent an increase in wealth concentration? Obviously. Is that a net bad? I don't know, let's discuss instead of silencing people.
Waymo are toast, Tesla will out scale them in months if not sooner. They can't compete on costs, 100k plus for an ugly Waymo vs < 40k for Tesla model Y or cyber cab.
It’s wild that $goog is so undervalued (p/e 27) given Alphabet owns Waymo in addition to everything else, and yet Tesla is so overvalued (p/e 243!!!) despite zero Robotaxis in the near (or far) future and lackluster sales.
Goes to show empty promises and fraudulent showmanship sell better than actual working products that people use.
GOOGL is up like 25% over the last few weeks after they resolved the DoJ lawsuit about Search bundling. Clearly there were some investors who thought that was a material risk to the business.
Tesla is clearly a meme stock though, and an example of how the market can say irrational longer than you can stay solvent.
I finally capitulated and bought a few shares of TSLA, shorting wasn't working.
Err... that sounds like gambling sir
So you think it's going to raise further?
No, they were just scared of missing out.
If you buy Alphabet stock you're betting on the whole company doing well.
Google makes around $300B a year. Uber's entire business makes around $50B and that took a decade. Waymo would have to become a major business to move Alphabet's stock price in the near term.
Considering Waymo is very likely losing money, experiment very slowly with scaling up, and still raising billions in private capital outside Google... idk. Doesn't seem as simple as buy $goog in 2025.
Otherwise I agree Tesla is a bit of a meme stock.
I think Waymo has huge potential for being much larger than Uber - people are willing to pay more compared to ordinary uber drive just to avoid dealing with taxi drivers and tech will only get cheaper.
More than that, I think the ride-hailing business is just the fist volley in the self driving vehicle space. It’s a short jump from there to self driving trucks, self driving package delivery, self driving private vehicles, and on and on.
All of those spaces are actively being explored by various companies.
Can any of those companies catch up on self-driving faster than Waymo can pivot to their niche? Cruise seemed to be a distant second, but did themselves in with an attempted cover-up.
There are already self-driving trucks on the roads. Their pilots came earlier, because the problem space is much smaller.
They don't need to "catch up" to Waymo, because of the niche.
https://bigrigs.com.au/2024/04/18/driverless-trucks-trial-be...
> There are already self-driving trucks on the roads.
2 trucks?! I suppose that's the minimum number required to make your pluralization correct.
I will stand on my earlier statement regarding this particular outfit: they'll need to catch up because Waymo started class 8 variants in 2021 https://waymo.com/blog/search/?t=Waymo%20Via
And plenty have failed. Perhaps a smaller problem space but still really, really hard. Some self driving freight company failures: Starsky, TuSimple, Embark, Ghost, among others.
One promising self driving truck startup, Aurora, was forced to put a safety driver back in the driver's seat after testing in May.
https://www.ttnews.com/articles/aurora-driver-back-in-seat
Buy a Comma.ai and install it in a supported vehicle, and just try it out. It doesn't talk to GPS, but it handles left right gas brake on the freeway well enough, and that's with two fairly shit optical cameras and a radar system. Granted, geohot helped start the company, and he's no slouch, but if their system is that good, a couple things are true. A) Lidar isn't necessary b) Extensive mapping that Waymo does also isn't necessary c) that last 10% gonna take 500% of the time to get to L3/4/5 autonomous, and that last 1% is maybe never. The other day I was in a Waymo, and there was a semi totally blocking the street, backing into a loading dock. The Waymo correctly identified that there was an object in the way, and stopped and did not plow into it. At first it crept up to the semi, blocking it from making progress as well. It might have started backing up, I've seen them do that, but I was already on the customer support line as soon as I saw the semi blocking the road.
Comma.ai is probably the purchase I'm most happy with this year (to be fair though, I buy a lot of crap off Temu). Drives are now just "get on the freeway, and just chill." Pay enough attention because it's not collected to GPS and just in case something goes wrong. So to be clear, Comma.ai is not autonomous driving, it's classified as an ADAS, advanced driving assistance program. It just makes driving suck that much less, especially in stop and go traffic, for $1,000, and compatible with recent vehicles that have built-in lane guidance features. Waymo's got to be light years ahead of them, given how much money they've spent, so it's my belief that Waymo's taking it very slow and cautious, and that their technology is much more advanced than we've been told.
Probably not.
Cruise was nixed by GM execs, whom I believe were looking for whatever excuse they could find to shut the operation down. They simply couldn't afford to stay in the game for the long haul. Cruise was under pressure to appear more capable than they were, and they took risks.
Waymo is distinguished in that it doesn't need to pander to nervous investors to keep getting money. The company is Sergei and Larry's baby. Google's founders will ensure that Waymo is patronized until it can stand on it's own.
> ...I believe were looking for whatever excuse they could find to shut the operation down
Cruise's self driving license was suspended because humans displayed poor judgement by omitting from the official report details of their stopped car dragging a knocked-down accident victim under the car for dozens of feet. They took "risks" alright, and their harebrained cover-up was discovered by chance by the oversight body.
I believe any driver who covers up the details of injuries in an accident permanently lose their license, because they'll definitely do it again. What good is a self-driving subsidiary that can't operate on public roads?
How does self-driving package delivery work? Who delivers the package?
There are several “last meters” delivery robots developed.
Short range drones are being used in Australia.
And I heard of at least one company working with apartment architects to standardize a “port” on the building exterior to which a truck/robot would connect to “inject” packages to the inside.
Like some sort of "mail chute"?
This was just an acquaintance some years ago in SF, but I recall it was fancier with conveyor belts and a protocol for the robot to communicate the size and weights of the packages being delivered.
Tiny catapults. It's the only correct answer.
Sadly, this would still be an improvement on many smaller delivery services that especially Amazon is fond of using.
The slaves obviously.
But to be serious, there may be a way of doing it, it just seems very far off unless you're talking about Amazon hub or something like that, where it would be more feasible (but still difficult to achieve).
Think of Waymo Driver as the equivalent of Android for vehicles. It's an operating system and a suite of cloud services for both autonomy and ride hailing.
What about all the expensive hardware, gpus, lidars? That’s like having iOS on your phone and if you want android you need to buy extra things that are worth same price as your phone.
The long history of "First mover advantage" being a myth implies they are more likely Nokia or Blackberry than Android
And costs should be lower in the long run if you don't have to share the ride fee with a driver (not case yet because seems like they still have alot of staff to manage the cars)
Statistically Waymos are more expensive than Uber rides, but practically as an individual they are often cheaper than Uber, its very easy for the stated price to be lower
So its not even about willingness to pay more
Gig drivers are cooked
A lot of times the Waymo is only a few bucks more, so if you were going to tip the uber driver it balances out anyway.
Driving in a car that doesn’t smell like driver just farted right before picking you up is worth the premium.
Who are these people?
There is no downside to having someone drive you Uber has homogenised the experience.
Checkout this thread for who those people are: https://news.ycombinator.com/item?id=44258139
That TC article doesn't substantiate its overly broad claim. "People" aren't paying more, in general, across its US markets; it only shows that a subset of its customers in what is already the top-5 most expensive cities (SF) in the world are prepared, and at that, only 10-27% are prepared to pay significantly more ($5-10). Still fewer than the 40% who would pay “the same or less.”
Quoting: "Perhaps even more striking is how people answered a question about whether they would be willing to pay more for a Waymo. Nearly 40% said they’d pay “the same or less.” But 16.3% said they’d pay less than $5 more per ride. Another 10.1% said they’d pay up to $5 more per ride. And 16.3% said they’d pay up to $10 more per ride."
There are going to be lots of causal factors: number of rider(s), time of day, safety, gender, wait time, price estimate, predictable arrival. Let's see an apples-to-apples comparison/regression breaking out each.
Anyone who's taken enough Ubers and/or has had bad enough luck to have gotten a terrible Uber driver. Pretty much everyone I know, along with myself have had multiple awful Uber driver experiences.
I think waymo actually has a better km/accident ratio than the average driver. Plus if you haven't done it before, it'll be a cool experience to ride in a car with no driver!
But in the long term I think the point of waymo is that it'll be cheaper: no need to pay the driver if there isn't one!
Women. Turns out, Uber/Lyft can't really do anything about drivers assaulting passengers.
Waymo doesn’t own manufacturing of vehicles.
I don’t think Waymo is very likely a losing money experiment. I give them a 50% chance to be successful within the next 10 years. Successful being that self-driving cars are able to operate in 50% of the world/terrain types/region types, probably within another 10 years to scale up.
They have already spent an enormous amount of money. It’s hard to see how they could make it back quickly, if ever. I’d like to be wrong, but I expect they will continue to be a money losing experiment for a long time yet.
How much money they've spent in the past is irrelevant. That money all came from investors, in exchange for a stake in the company. It never needs to be "paid back". Besides which, those investors have earned all those funds back already, and then some (on paper).
All that matters at this point is how much money they'll lose/earn in the future. There are no shortage of investors willing to put money into this effort, and they're growing exponentially, so there won't be any pressure for them to turn overall profitable for several more years.
Waymo investors are Google investors, are they not? Will Google shareholders be happy continuing to put gigabucks into Waymo, or will they prefer that Google cut expenses before Waymo can make money?
How much money do they make off the average person in the value of ads shown per year?
Now compare to how much money the average person spends on driving per year.
If Waymo winds up running half the market in autonomous transportation over the next several decades, it'll make search look like peanuts in comparison.
The entire global taxi market is ~$250B a year.
Google made ~$265B from its ads last year.
Not the global taxi market.
The global driving market.
When these are ubiquitous enough, the vast majority of people who currently own cars won't need to. It'll be so much cheaper and easier to use rideshare.
I can't really imagine the circumstances where I wouldn't want to still own my own vehicle even if it had an autonomous mode. I drive it places where I don't have cellular service. I keep lots of stuff in the vehicle. It's customized with accessories like roofracks. I can hop in my vehicle from my house immediately whenever I want to.
If I lived in a city and garaging a car were inconvenient/expensive? Maybe. But that's not me or a lot of other people.
It's not going to be for everybody.
But if it's half the price over the course of a year? And you can summon it in advance cheaply? And it basically never takes more than 5 min to arrive anyways, since they're everywhere?
You might decide it's worth it to keep the stuff you really need in a messenger bag or backpack or something, the way people in NYC do. And maybe roof racks don't matter if you can just summon a second autonomous van behind you to hold whatever you were going to put on your roof.
Obviously if you're a contractor or something you'll need your own vehicle. But the point is that for most people, sure they can't keep stuff in their trunk all the time, but that's a happy tradeoff for a total cost of driving that's 50% less.
Imagine if you could buy your own "Waymo-equipped car". No need for driving lessons. No aggravation. No road rage.
How many people would pay for such a luxury car? With the US population aging and public transit non-existent in most places, Waymo probably has a market for cars.
I am I really hate driving long road trips.. So yes! Or they could even sell private taxi between states so I don't even have to own a car :)
Uber took 14 years to make it to profitability. Money's frequently characterized as impatient, unable to look past the next quarter, but when it wants to be, it can wait.
Waymo's older than Uber, but they hold many key patents by this point. Now that they've started running a taxi service, it seems straightforwards to scale up, assuming that is the business they want to be in. Then it's just a matter of charging more than it costs to run the service, and wait.
Tesla has 1/3rd the market cap.
If Waymo is a rounding error to GOOG, it's basically a rounding error to Tesla's implied valuation.
So what is Tesla valued in then?
Clearly not car sales, profit, and especially growth in either of those segments.
xAI is supposed to be where all the AI is.
Where is it?
Faith in the fact that Elon has never lost investors money.
Robotics I think.
Uber also has to pay drivers. How much of that $50B goes to the operator?
Meanwhile, for Waymo, a good chunk of it is profit (after the fixed cost of the vehicle, of course).
The cost of the computers, LIDAR, special maintenance, vandalism, staffing humans for remote issue handling etc will probably costs the same as a year's income for an Uber driver. But after that it's mostly profit and they can run cars longer.
The most important thing for Waymo is scaling up production of LIDAR and maintaining them efficiently. They will have a massive fleet running very sophisticated radar+computers. That's a huge logistical investment when it's a million cars. Those sensors will break or be damaged.
They've been partnering with Uber to maintain the fleet in some cities haven't they since they already have regional infrastructure? I don't think they want to be in the fleet management business.
AFAIK Uber is doing app integrations + some local operational fleet management. Waymo is supplying the cars, radars, computers, remote service, the brand, etc. Waymo has to scale that production and maintenance up country wide and then globally.
Uber's CEO compared it to Marriot, people come in to run the hotels in the local region, but they actually don't own the hotels. It's like hired managers who take a cut.
It also makes sense to have people with local experience run them in each local region. But those businesses still involve margins and expenses that have to make sense.
Don’t forget that Waymo will always be a much lower margin business than search! Setting aside the decades of R&D expense, those cars require purchasing, maintenance, warehousing, etc.
Autonomous cars won't sue you, never sleep, don't go on strike, don't sleep 8 hours a day, keep driving when the car needs obvious repairs.
>Autonomous cars won't sue you
but the companies that own them will or their insurance carriers.
I think the plan is that other entities will own and maintain the cars. That's why they're working with partners like Uber and Avis.
But the market is so, so much bigger. And the margins will likely stay high for a long time while there are few competitors, and their main competition is human drivers.
Not having to pay drivers is an enormous source of profit.
Indeed. The richest showman that ever lived and successfully duped both liberal and conservative population and politicians. Well deserved I say.
Wild that people will call the founder of SpaceX a "showman"
Let's settle on calling the founder of Hyperloop a "showman".
That is a real, important accomplishment, but he's also a showman.
Don’t forget Zip2, PayPal, Neuralink, OpenAI, and The Boring Company.
There are large swaths of people that accept headlines as fact and/or cannot or will not grapple with nuance and complexity (“I think Elon’s a jerk and he is a formidable engineer.”) Perhaps it’s a sign of these polarized times, or, as I believe, people have always been like this. We just have more time and resources to dedicate to outrage and flamewarring than we did in the past.
I don't deny his accomplishments. On the contrary, I think he is a genius. It's just that he is an extremely, damagingly biased genius.
https://x.com/elonmusk/status/1240754657263144960
Genuine question - are there (or have there been ) any geniuses that are not unhinged?
Yes, there has been nice geniuses (ie. people with extreme talent), Mozart was for example a good person. Da Vinci (if a little sycophantic when young) was not unhinged at all nor abusive and was appreciated.
But since romantism we have built this image of the genius as necessarily abusive.
I’m sure abusive genius are very visible (by definition?) and that abusive people tend to monopolize more ressources too. (Like these tenured professors that use their students to advance their own career)
Einstein, Euler, and Darwin were also nice people by many accounts.
He was ousted from Paypal before anything major happened, he was basically just a shareholder.
The Boring Company is an obvious bust. So is the Hyperloop. Neuralink is another likely bust. Tesla solar is going nowhere. The Cybertruck is a millstone around Tesla's neck. Etc, etc.
He wasn't even a fonder of Tesla. He was just a investor that became the CEO.
And the tweet below makes me question a lot about him. Doesn't sound like a genius to me.
"Lidar and radar reduce safety due to sensor contention. If lidars/radars disagree with cameras, which one wins?
This sensor ambiguity causes increased, not decreased, risk. That’s why Waymos can’t drive on highways.
We turned off the radars in Teslas to increase safety. Cameras ftw."
Waymo is driving on the 101 in LA
To be fair Tesla was in a very primitive state when he took over.
I think the real purpose of the Boring Company and Hyperloop were preventing/slowing expansion of public transit, and that by that measure they were successful.
I think the purpose was to extract money from governments, like most of Elon's businesses.
I don't think it was a carefully calculated conspiracy (such as 1)
I think it was an engineer with found wealth starting to do stuff with it.
but nowadays I think he has evolved into something different, maybe some of it from the wild public feedback loop, some of it because some of the things he cares about are going wildly wrong.
[1] https://en.wikipedia.org/wiki/General_Motors_streetcar_consp...
Usually Elon's technical flaws aren't on display, or at least he covers them well. For example while it's true FSD hasn't worked out, but I don't know you could say at the time "most competent AI devs knew it wouldn't work out". However, when Elon attempted to move PayPal from Linux to Windows, most competent software engineers would have advised against it. Paypal isn't an example of Elon's genius in action - it's the opposite.
When Tesla introduced HW2 it was clear to people in the self-driving industry that it wouldn't work out. Elon was insistent on repeating mistakes that other companies had already learned from. Of course the other companies never considered some people's willingness to pay good money just to pretend that their cars can drive themselves.
> FSD hasn't worked out
Says who? I've tried it and the capabilities are amazing. If you told me 10 years ago that I would be able to buy this in 2025 I wouldn't have believed you.
I am "just a shareholder" in Paypal. Elon Musk had a > 10% stake inherited from his ownership of one of the companies that was the precursor to Paypal itself. It's not remotely the same thing. And listing failures is not meaningful at all. Failure is the default outcome in business.
Wild that people will call a guy who bought SpaceX the founder of SpaceX.
Either go ague with Wikipedia, or put some argument in the comment when making claims you expect people to verify themselves. People are just going to look it up on Wiki.
> SpaceX was founded by Elon Musk in 2002 with a vision of decreasing the costs of space launches, paving the way to a self-sustaining colony on Mars.
https://en.wikipedia.org/wiki/SpaceX 2nd paragraph
Deceiving people doesn't mean you deserve your gains.
"Deserve" is a human construction.
Largely because investors fear that Google's new products (especially AI) will cannibalize its massively lucrative ads business.
Fear is a bit of an understatement
but if they're google's products how would they cannabalize ads biz. would revenue not just shift? or do you believe ai search will be overly adopted but not as profitable?
I think its the later. And also the fact that they are not the firstmover in AI search. More people know about chatgpt than they know about gemini
Google was late to search, late to smartphones, late to internet email. I'm having a hard time thinking of any of their large markets where they were a first mover, maybe YouTube-ish, widespread user uploaded internet video wasn't meaningfully available before the rise of YouTube.
On topic, Waymo is clearly a first mover in self-driving, having the first legal commercial services.
But, being the first mover is usually more of a disadvantage than an advantage, IMHO.
I believe TSLA also represents their humanoid robot segment with some questionable addressable market definitions done by investment analysts. I believe it’s overvalued but they are a forcing function for the other tech companies to push ahead
Stocks are narrative-driven, and sometimes this aspect swamps the "fundamentals." Keynesian beauty contests all the way down.
Tesla is literally operating a robotaxi service.
They're operating a Robotaxi service, not a robotaxi service.
If I create a shuttle bus service for my neighborhood and call it the "Space Shuttle", I am not operating a space shuttle.
A whole 15 cars, with "supervisors" in the drivers seat!
And only last week did they even open up the waitlist to non-influencers.
The supervisors are not in the driver's seat.
https://electrek.co/2025/09/03/tesla-moves-robotaxi-safety-m...
The day this news was released, Elon released the video of him talking to the Optimus bot to overshadow the news. Showman gonna showman.
TIL. I stand corrected. Though worth pointing out (as the article does) that on September 1st, new legislation in Texas was passed adding some restrictions to autonomous vehicles. So seems reasonably likely this is more regulatory than necessary.
Unsafe at any speed
Google is just not a risk taker these days. You don't risk you don't get rewarded.
There are about 1,500 Waymo cars in existence, versus about 7,000,000 Teslas in the last seven years.
But there are 0 Teslas that are as effective at self-driving as Waymo, so they're still ahead.
My Model Y in Vancouver drives me to and from work daily. I cannot get a Waymo here -- and I certainly cannot purchase one privately. Which is more effective where I live?
Teslas have a ~about 500 miles between interventions (they don't release actual data, no surprise), whereas Waymo is at around 17,000 miles.
That's a 34x divide. At full scale that's something like 30% of Teslas having an intervention every day.
I don’t doubt that Waymo car is more advanced than FSD, but that comparison isn’t as impressive as it sounds. The numbers of FSD equipped Teslas dwarfs that of Waymos, and they are available everywhere, not just selective cities. You have to take that into account.
Teslas is also much cheaper, and easier to scale. Tesla has better growth potential even if their tech is less impressive.
It's not that their FSD tech isn't less impressive, it's that it's not FSD tech.
Even worse (for Tesla) is that if they do try an make their non-FSD tech do FSD, and it decks little jimmy because the flashlight in his hand looked like a far off street light, Tesla is liable to face a knee-jerk federal law mandating lidar. And just like that the dream is dead.
This forces Tesla to be extremely paranoid, as it's one visual mistake away from being told to use lidar.
You are supposed to supervise Tesla FSD. Waymo doesn't require someone in the driver's seat at all. They aren't the same thing.
Market says “as effective” doesn’t matter. Needs to be “good enough”.
I mean FSD is pretty good and useful. But yes, not unsupervised.
The Coca-Cola company sells even more units than Tesla, but if those units don't drive themselves they're moot to this discussion.
Same could be said about Tesla when it started.
Overvalued by traditional (PE) means. I've ridden in Waymo (50+) and Austin Robotaxis (12). Tesla has Waymo beat in terms of human-like feel, interior features (sync to your own Spotify, Youtube, etc). When Tesla removes the passenger seat monitor, scaling will happen much faster than Waymo... Tesla just received the initial license for driverless Robotaxi in Nevada. Tesla also produces more Robotaxi-capable Model Ys in ~6 hours as Waymo has cars in service (in total).
Tesla's self-driving technology is a joke compared to Waymo's and the Tesla brand is extremely toxic now. I see from your other comments that you're big on Tesla (own several and have a son who works there) but as an unbiased observer I cannot fathom them winning this market.
I have 2 AI4 Teslas with FSD, and I don't find V13.2.9 lacking at all in the Vancouver area. V14 will be a 10x increase in parameters, too. Why do you feel it's a "joke"?
It's a "joke" (I wouldn't call it that, but it's a vastly different product) because you have to pay attention to the road at all times.
You don't live in a Waymo city, so I understand. A lot of people who don't live in a Waymo city don't really get it.
Waymo is a completely different product than FSD. It's a robot that comes and drives you from point A to point B. You can do whatever you want while it's driving, such as take a nap or work on your laptop.
Tesla was SAE level 2 in 2013, and they are still SAE level 2. Waymo's Robotaxis are SAE level 4, and they can drive on public roads empty with no human supervision, both technically and legally.
> When Tesla removes the passenger seat monitor
This is a huge jump, possibly still 5+ years away.
I have friends on the Autopilot team (and a son). Their goal is by end of year. I've been on HN for 15+ years, and seemingly the only downvotes I get are when I post my thoughts and opinions on Tesla.
Tesla FSD has been autonomous by the end of the year for 8+ years now. Don't believe people desperate to make Elon's lies seem plausible.
FSD is widely considered to be off its originally-stated goal by at least 5 to 6 years.
It would be great if Tesla figured out how to do safe autonomous driving with their very limited sensor suite, but there's a lot of reason to be skeptical: https://en.wikipedia.org/wiki/List_of_predictions_for_autono...
I wouldn't not be surprised if they figure out some very narrow way to have no safety driver in the car (1:1 remote ops?) by the end of the year.
1:1 is going to be ruinously expensive. You need three shifts of remote operators. Even in the Philippines or Vietnam, if you can make the latency work, that's prohibitive.
> 1:1 is going to be ruinously expensive.
I agree, but this is how taxis/Ubers work.
How do Elon Musk's predictions relate to Tesla achieving a robotaxi service or not?
Ignore his predictions and just... look at whether or not the Tesla FSD team is making progress.
> How do Elon Musk's predictions relate to Tesla achieving a robotaxi service or not? > > Ignore his predictions and just... look at whether or not the Tesla FSD team is making progress.
I'm seriously baffled by this comment. How can Elons comments not be relevant? How are you proposing we assess the progress of the FSD team? And why should the assessment be different to the last 5 years where FSD was supposedly ready (according to someone with intimate insight into the work of the FSD team) by the end of the year?
> How are you proposing we assess the progress of the FSD team?
...any metric you want? Miles driven under FSD. Miles driven without intervention. Miles driven without accident. Anecdata from friends of yours who own a Tesla. Whether or not a partially supervised pilot program has been launched in some cities.
If Elon Musk said in 1999 "I think we will achieve self-driving next year", that also has no bearing on whether or not self-driving is achieved in 2025 (in either the positive or negative direction). It only means that Elon Musk's "predictions" can't be trusted as an accurate harbinger of success. Which is precisely why you look beyond his words and at the reality on the ground, which strongly indicates Tesla has made a huge amount of progress in the last 10 years, and could be very close to having unsupervised robotaxi service in various jurisdictions.
Can we expect you to come back on Jan 1, 2026 and provide an update?
You're not "posting your opinions on Tesla", you're literally shoveling them into everybody's throats. You'd be "posting your opinions" when it was one, two comments, and not plenty, like under this news. You're a Tesla freak or fanboy, not an objective commenter.
> Their goal is by end of year.
It's like what 6-7 years since the goal was "end of the year".
> Their goal is by end of year.
Ummm.
Waymo does not have YouTube sync, but they do have Apotify sync.
> When Tesla removes the passenger seat monitor
They literally moved that monitor to the driver's seat! Progress, indeed.
This sentence was a bit cute: "Waymo has received our pilot permit allowing for commercial operations at San Francisco International Airport." Yeah, that kind of pilot.
I really had to read through it twice to make sure they were just talking about car taxis picking up travelers, rather than some kind of prototype pilotless commuter helicopter or something.
That was my first interpretation, and I was very surprised and kind of afraid. Glad to know they aren't trying for autonomous flight yet.
I have zero expertise for my claim, but I feel like autonomous flight is easier than autonomous driving.
The hard part of automated driving is dealing with all the ground clutter that planes serenely fly over. If pedestrians could charge out in front of a 777 going 650 mph at 34,000 feet... well... we'd be living in pretty different world! And in that world, flying would be much more difficult. Not just for computers but for humans too.
Flying is obviously much harder than driving, but it's a sort of harder that is generally more amenable to automation, though I still think pilots are a good idea because when it goes wrong it goes wrong much worse.
Flying is almost always easier than driving. landing is hard. Bad weather is hard. But just flying - human pilots have napped many times over the years and it only rarely is an issue. Airplanes with primitive autopilot are very good.
Yeah, a primitive autopilot in a plane just needs an altimeter and compass, but a AoA sensor, speedometer, fuel level sensor, and pitch sensor help to detect unsafe conditions like runaway pitch, stalling, overspeed, low fuel, etc. Each of those sensors is providing a simple 1-dimensional data point. Redundancy is relatively inexpensive.
Automatic lane keeping in a car requires cameras that software needs to then analyze to find the lines in the road in real time. But if you want a "set it and read a book for an hour", then you have to respond to other traffic. No longer just some simple PID controllers, the software now needs to plan and execute based on surrounding traffic.
Yep. 0ft-1000ft AGL Takeoff, Climb, Approach, and Landing are the tough bits. The rest (Cruise) is very low demand and much easier than driving.
Taxiing is probably harder to automate than the rest. But you could have pilots on hand to taxi to the runway, and take a shuttle to the other end and hop on a just landed plane to taxi to the gate. Or you could use tugs for ground movement.
I'm not convinced - in a commercial airport taxiways are controlled by a ground control systems, not just pilots looking out the window. If the only airplanes around are also equips with the self taxi system they just report position to the central control and that tells them when to go. There needs to be emergency overrides for when that system fails, or a small plane without it is around, but that can be handled by stopping everyone else in the area until the hazard is gone.
Time will tell...
There's also all of the service vehicles when you get closer to the gates. The likely damage from an incident during taxiing is much less than during take off or landing, but I think the risk of having an incident is higher and the situation is trickier to manage. And it's super doable to have a pilot come on to manage that, and drop off after the hard part; you couldn't reasonably have pilots do a takeoff and then jetpack over to an arriving plane to do the landing, but it wouldn't be unreasonable for ground moves... similar to canal pilots taking ships through canals.
>human pilots have napped many times over the
months?! :)
"The [German pilots'] union said it had carried out a survey of more than 900 pilots in recent weeks, which found that 93% of them admitted to napping during a flight in the past few months."
-The Guardian, "Almost all German pilots admit to napping during flights in union survey"; 2025-09-10
Years as since humans have flown planes stable enough not to need constant attention. On a calm day you don't need autopilot, just set your trims correctly and some airplanes will hold course well enough for a short nap - though of course this is more likely to result in a crash (which likely has happened, though it is hard to guess why a plane crashed beyond pilot error)
Not to mention that almost all civilian planes in the US are required to broadcast a bunch of details that include their coordinates and altitude on a public channel (ADS-B). It's the kind of automated collision avoidance input that you'd probably dream of as a self-driving system engineer. Basically the only thing you'd need to avoid via more complex systems is the odd military traffic, small craft at low altitudes, and birds.
In the abstract yes but in practice the economic (ratio of cost of pilot to pax miles) and safety context of aviation mean fully autonomous flying has to be extremely robust before it has actual utility in industry.
In practice, you're also currently very reliant on infrastructure that is definitely not as solid as you want (eg: ILS and GPS can be interfered with quite nastily).
ILS being under maintenance and unavailable for certain runways is also far from unusual.
On the happy path, yes. Though I don’t think takeoff is automated yet.
Currently we rely very much on the problem solving abilities of human pilots to deal with troublesome situations. Autopilot will disengage in many scenarios.
I'm pretty sure drones can already take off on their own. Taking off is a lot easier than landing, and planes have auto-landing tech already.
Drones (both autonomous and remote piloted) have much higher mishap rates than crewed aircraft. Taking off is "easy" until something goes wrong, like a mechanical failure or runway incursion. It's impossible to anticipate and explicitly code for every possible failure mode, so developing autonomous flight control systems that would be safe enough for commercial passenger flights is extremely challenging.
Category IIIC ILS (full auto-land) does exist but requires special equipment for both the aircraft and airport. Human pilots have to actively monitor the system and take back control if anything goes wrong (which does happen).
Garmin also has the Autonomí auto-land system for certain general aviation aircraft which can attempt to land at the closest suitable airport. But this is only used for single pilot operation in case the pilot becomes incapacitated. It isn't suitable for regular flights.
Takeoff at a commercial airport is a very challenging and potentially dangerous situation. There’s way more margin to abort a landing than a takeoff.
OTOH takeoff and landing could in theory be operated by people on the ground, flying simulator style.
I still believe that having an actual pilot inside the plane that care for his own life is not a bad idea vs someone remote feeling a bit disconnected with the reality of a crash.
Remote piloting is how the military operates certain drones like the MQ-1 Predator. The mishap rate is very high relative to crewed aircraft due to network lag and sensor issues. The military is willing to accept some level of equipment loss in order to accomplish their mission but this would never be allowed for commercial airliners.
The pilot’s self-preservation instincts aren’t the most important reason to have them onboard. It’s that any loss of communication between the ground and the airplane at any point during either procedure would turn it into an uncontrolled cruise missile.
I am not sure why you were down voted. The original meaning of the word pilot is someone who comes aboard a ship for "the last mile" - getting in and out of the harbor and what you are talking about is kinda like that - a person associated with the airport rather than airplane to guide the planes in and out - perhaps using more reliable local communication technology vs what is used to control drones half way around the world.
I have no idea if that works but I thought you were making a good contribution to the conversation by proposing a potential solution to the exact problem everyone is talking about.
It's kind of funny how you can both be right.
Drones crash on takeoff all the time. Worth noting that drones are more than just quadcopters and serious drones are often winged aircraft.
It's the failed takeoffs that lead more often to jets leaving the run way and crashing into buildings or trees.
When everything is working correctly, no other pilots have emergencies, and no temporary restrictions are in place, and there are no clouds in the sky. Then yes, it /could/ be easier, but almost always it never actually is.
There's a reason the majority of accidents occur during take off and landing.
Spend some time listening: https://www.liveatc.net/search/?icao=ksfo
It depends a bit on your safety standards. There are already autonomous flying things delivering blood and blowing up oil depots where it doesn't matter so much if stuff goes wrong, but to be an airline pilot you have to know how to deal with a huge range of emergencies and systems packing up.
With a car if the engine fails you just pull over. With an airliner it's not so simple. As a result the training for a pilot is much longer than for a bus driver say.
I also feel like the demand is way, way lower. A pilot can't be that large a % of the cost of a flight. Maybe if we lived in the jetsons era.
Depends on the size of the plane, really. One of the reasons a few companies were investing in fully autonomous air taxis is because the math on a small piloted aircraft wasn't realistic for a low enough price point to be competitive.
The problem is actually safety. As automated systems get better, the pilot is left with not much to do, and has to maintain vigilance while being really really bored. It is almost better to have fewer automated systems and give the pilot more things to do during the flight so it is easier to keep them paying attention, or all automated with no human pilot to mess things up.
0d (parked) - null program easy
1d (train) - easy. just one lever
2d (car) - hard. super hard. why is it so crowded? who thought this was a good idea? you let teenagers do this?
2.5d (plane at takeoff or landing) - almost as hard as car. fewer pedestrians.
3d (plane flying) - easy even with all those extra levers
I'm not actually sure how hard landing is. Most airports that support autonomous landings do it by having ILS antennae that guide the airplane to within tens of feet of the runway, at which point the airplane switches to radar for altitude.
Automatic landings started in 1964. I think that it seems hard mostly because of how tightly regulated aviation is - modern technology could probably make things a lot better if people were more receptive to the idea of heavy automated aircraft over populated areas.
landing is easy. the hard part is landing with 20mph cross winds and one engine out (or other mechanical failures). we've had auto-land that is 99% reliable for a while now, but you need to get to 6 9s before you have a system safe enough to replace pilots
I think that as long as the autopilot is able to fly in a crosswind or with an engine failure, it can probably land with one. Autopilots are already able to do these things.
I doubt anyone has tested this in depth, but I'm not sure there are too many configurations of airplane these days where a human can safely land it and a computer can't. Maybe if a big chunk of wing or control surfaces were totally gone, but even a human pilot isn't getting 99% reliability in a situation like that.
In any case, I don't think that the first candidates for automation are gonna be passenger flights. It will probably be small cargo planes first - Cessna Caravans and other turboprop aircraft where the cost of paying pilots is roughly similar to the price of fuel.
Don’t have a ref but heard that it’s been safe for quite a while but they keep the pilots around due to consumer fear rather than actual improved performance. Curious if anyone can confirm.
No. Airliners can't even take off on their own yet, and are only allowed to auto-land with zero visibility at a few dozen airports when the pilots, plane, and runway are all current/recently checked.
Look up the Airbus ATTOL project's first automated takeoff a few years ago.
Also, there's virtually no automation when it comes to interacting with ATC.
An airplane will take off when it is properly configured and it hits a certain speed. It's simple aerodynamics/physics. Pilots are there to react to failures and unexpected events.
There's a bit more to it since you do need to do last bit of configuration (pull up the nose) just as you hit the target speed. But yeah, automatic take-off is quite a bit easier than automatic rejection of take-off.
Even manually pulling up the nose once you reach Vr isn't necessary if you just trim for a little extra nose-up. It'll eventually get off the ground with just enough speed.
There's no lack of online arguments about whether or not Vr is "real" or should exist.
I just followed what my CFI and Cessna's manual for the C172 said (which iirc was giving input to rotate at 55kts).
Sure. It'll also land if you don't care about anyone surviving.
And the air is within acceptable temperature and pressure ranges. I assume configuration takes weight into account as well.
Yeah temperature, wind, altitude, weight, runway slope all matter, and then there needs to be enough spare space for the aircraft to successfully take off even with engine failure in the worst possible moment. Then there's the question of fuel consumption too. Takeoff power isn't typically configured to get the aircraft off the ground as fast as possible, but to minimize fuel consumption, while still leaving enough margin in case of engine failure.
It wouldn't be that hard to fully automate a flight from gate to gate when everything works perfectly. But the various failure modes, human error like airport vehicles entering active runway, all that requires human backup. Self-driving car can just stop to the side of the road and turn on emergency lights if its engine fails, with a plane things get much more complicated.
And you don't need rudder input or any aileron input because of crosswind, and other bits that falls into "technically correct but not particularly relevant" territory.
It's fun to see/feel planes do stuff "on their own" (eg making them oscillate, or level on their own, or feeling ground effect, or even your own wake on steep turns) but it's not something you'd want to rely on (maybe with the exception of ground effect on short field takeoffs, but I digress).
> Also, there's virtually no automation when it comes to interacting with ATC.
Check out the Cirrus Autoland feature in their aircraft. They are all small personal aircraft, but the tech is pretty cool. Will talk to ATC and fully auto-land for you in the event of an emergency where the pilot is incapacitated.
"Talking to ATC" is a bit of a huge ask. The system basically just hops on 121.5 (and maybe the nearest/local unicom/tower frequency) and start an automated callout with its intentions that it will be doing. It operates on the assumption that all other airspace users will hear the radio calls and stay clear of the emergency aircraft.
I'm aware of it, though I've never flown a Cirrus. But AFAIK, it announces what it's doing. It's not communicating.
If you can design the product and environment to fit automation, then automation can be quick and effective.
The less you can change about the product and environment, then automation run slower and less effectively.
Air liner operations could be automated, but the minimum equipment list would be more stringent, the destination airport would not be able to take any equipment out of service for maintenance, visibility minimums would increase, takeoff and landing operations would require more slack time.
Besides all of that, the owner of the airplane would still want to have some crew on board.
In short, it's not worth it yet.
===
There is also the paradox of automation: Automation generally makes the hard parts harder and the easy parts easier.
The current goal of autonomy for airliners is single-pilot operation more than full autonomy.
It's very cool stuff, technology wise, with potentially significant redesigns of cockpits, etc.
But the main thing is the plane basically needs to be able to operate just about entirely autonomously (especially during critical flight phases) in case the pilot is incapacitated.
In theory, once SPO is solved, autonomy is almost solved.
I'm skeptical that SPO will be allowed for commercial airliners in our lifetimes. Pilot workloads are fairly low during most routine flights. But when an emergency occurs then the workload suddenly gets extremely high, to the extent that even two pilots are sometimes overwhelmed. This isn't a problem that current automation technology can solve. There are an infinite number of possible emergency scenarios and engineers can't possibly code for and test every one.
Cargo flights over oceans and (mostly) unpopulated areas might be a valid use case for SPO. Cargo pilots have always been considered somewhat expendable.
I'm maybe less skeptical than you but still not super positive.
At the very least, I'd say it's at least two clean-sheet designs away (which I'd guesstimate at 30 years).
I'm a bit partial to it because I did a brief stint in the Airbus realm. Autonomy for airliners is an interesting set of challenges.
Navigation might be easier. The battery and safety tech isn't there yet to make it practical.
In a pinch, a car can just put on its hazards and pull over
That “just” is doing some heavy lifting! The car still has to deal with all the normal hazards of the road while pulling over, plus the hazards it is itself creating by acting abnormally.
Well if we're being picky, technically the car itself doesn't have to deal with the hazards it has created, rather everyone else does.
The point is you can't just "stop" a plane and wait for someone to figure things out (https://support.google.com/waymo/answer/9449023?hl=en). Whatever the difficulties in dealing with an abnormal situation in a car, it is strictly much more difficult to deal with them in a vehicle constantly fighting the homicidal urge to fall out of the sky.
"Autopilot" already exists when it comes to flying.
Sure but it's not autonomous in the sense of Waymo (ie, driverless)
Landing can be: https://en.m.wikipedia.org/wiki/Autoland
In fact, it's pretty routine. Don't have the source at hand, but somewhere around 1% of all landings (at airports with ILS) are autolands.
I think it was Boeing that even requires at least 1 autoland per plane every 30 days or so.
You can find videos of this on YouTube. Completely hands-off.
Most carriers have a rule that on clear days you always hand fly the landing.
This is a competence you do not want to lose.
It's also the case that you can have a whole approach setup in your flight computer and at the last minute the controller gives you a runway change. You could drop your head down and start typing a bunch info the FMC but you're generally better off just disabling auto pilot and manually making the adjustment.
I'm curious, what is harder to implement: autoland for airplanes, or autoland for rockets (spaceX)?
I don't know if these are comparable.
But two interesting data points from the Wikipedia article I linked are that the first aircraft certification for ILS Cat III was in 1968, and Cat IIIB in 1975.
And IIRC by the 1980s, autoland was already a pretty common feature.
Yes but it should have been obvious that in the context of Waymo + SFO, the implication was autonomous flying of commercial airlines.
Yes, but autopilot usually just keeps the plane flying in a straight line at some specified altitude, which have been around since 1912. It isn't full self-flying (although we definitely have drones that can fly themselves already, so that tech already exists).
That's an oversimplification of autopilot systems. They can follow flight routes, avoid traffic (TCAS), even auto land to name a few.
Auto-landers are not simply classified with autopilots. An autoland system is an advanced function that is part of a modern aircraft's overall autopilot capabilities. A basic autopilot can control an aircraft's attitude and heading, but an autoland system can automatically execute the full landing procedure.
See also: garden-path sentence (https://en.wikipedia.org/wiki/Garden-path_sentence)
This is actually why you have "Naval aviators". To maritime people a pilot means something else. https://en.wikipedia.org/wiki/Maritime_pilot
The flying kind is a license, not a permit.
There are driver's licenses and learner's permits. This could be the flying equivalent.
lol deniable demand-gauging :)
The setup at SFO is currently quite annoying (Lyft/Uber require you to walk 5 mins to the garage roof, and drivers need to park/wait 5-10 mins away, so there's always a substantial delay). Taxis get the privileged parking spot immediately outside arrivals, but if it's busy you might still need to wait a bit.
I've been wondering for a while why Waymo can't offer a semi-managed solution to SFO to dynamically manage load, have just the right volume of cars inbound, maximize parking utilization, etc. with all of the nice intelligence that an app-based system would enable.
It feels like you should be able to have a buffer of cars waiting right at the curbside, and automatically refill that buffer on short notice depending on observed or predicted demand.
Taxis have a powerful local lobby; Google/Waymo doesn't.
As a European, I can’t help but feel a bit sad that we’re missing out on the driverless side of things. It seems like most of the meaningful deployments are happening in the US (Waymo, Cruise).
I’d really like to see either a Waymo competitor emerge in Europe, or even Waymo themselves operating here. The regulatory environment is obviously more complex, but it’d be great if we didn’t end up years behind on something this transformative.
Cars of any sort, self-driving or otherwise, do not solve traffic any more than Uber does because you need to have enough of them to get everyone to and from work at basically the same time. Trains are the only way to address traffic. Trains are self-driving. Europe already has the better self-driving system. It's just boring because self-driving is much easier when you build the road to support it instead of removing all constraints and adding GPUs, lidar sensors, cameras and an army of fall-back operators in overseas call centers.
They replace taxis and potentially postal and trucking applications in future.
It’s certainly not a replacement for mass transit. US is sparsely populated compared to Europe and mass transit don’t work as well in the suburbia. That said, I do see many transit oriented development in SF Bay Area where high density buildings are being built near transit stations.
Self driving cars could work with trains to do the desired location to the station bit that has always been a bit awkward.
Trains are all very well but they've been around nearly 200 years and have yet to bring on a traffic free utopia.
Can you imagine how much traffic there would be if NYC didn't have the MTA? The principle of induced demand tells us that as long as there are roads they will have roughly constant traffic because people are willing to spend some roughly constant amount of time getting to and from destinations by road each day. More roads speeds up everyone's commute which brings in more drivers, which brings traffic right back to the baseline terribleness.
The question is how shitty it would be if they also had everyone on them who's currently on public transit.
So basically, it is a traffic-free panacea for everyone who chooses to use it. It's not a goal of trains to eliminate traffic for everyone who insists on driving.
https://www.tomtom.com/newsroom/explainers-and-insights/indu...
The induced demand argument works for trains too. If NYC didn't have MTA (no subway, no LIRR, no MNR) then the population of NYC would probably be 1% of what it currently is. Building more train tracks and having better train services also encourages more people to move to NYC so that these new train services become more utilized.
Neither roads or train tracks solve the traffic problem.
Yep, this is a good point. There are appropriate technologies for each situation. It's not a winner-takes-all contest.
For another example, can you imagine trains replacing school buses in a large, rural school district? Sometimes (not always), buses are better than trains.
Any one part would have the about same amount of traffic it does now. It would just sprawl out bigger across adjacent counties and the highest density parts would be lower density.
See also: LA
This is what bikes and busses are for, or just walking because the metro system is comprehensive enough you are at most four blocks away from a station.
They help to remove some congestion in the Netherlands. That’s my everyday experience. Traffic would be way worse otherwise
> have yet to bring on a traffic free utopia
This is a silly expectation to have. As long as there are roads for cars people will put cars on them.
Trains solve traffic for the people who get on them, not for drivers. The more people taking the train, the fewer people impacted by the traffic.
You could maybe have something like Zermatt Switzerland which is car free but you can get around in human driven golf cart like taxis. It's pretty pleasant but expensive. If the carts were self driving it could be cheaper.
(Zermatt pics https://www.traveladventuregurus.com/zermatt)
Zermatt is fundamentally a pedestrian town. There are a limited number of permits for electric vehicles available for companies that have an objective need for a vehicle. That limited availability makes the electric taxis expensive.
The total number of permits seems to be around 500 in a town of 5k permanent residents. And the population grows to 30k or 40k during the peak tourist season.
Human beings naturally take advantage of new conveniences.
If public transportation just encourages people to move to the suburbs and commute in every day you've actually just displaced the problem.
Trains will fairly unreliably take you from one place that is not your home, to another place, which is not where you want to go, at a time that is probably not exactly when you wanted to arrive. Freedom of movement is incredibly important, and trains are very rigid in this aspect.
Well That’s certainly not been my experience when visiting Europe. In fact, it many cases it’s been the opposite - having a car would have been restrictive in any major city and a source of friction.
> having a car would have been restrictive in any major city and a source of friction.
Would a Waymo that you don't have to store, park, fuel, or maintain have been restrictive?
Well to the extent it draws people from public transit, yes because traffic makes being a pedestrian more unpleasant and waymos still are traffic. And increased traffic adds friction to crossing streets and they park obnoxiously, among other things.
So yes, they would be obnoxious at any significant quantity and also not really help with getting across the city since transit is pretty good
Human driven vehicles are a menace: dangerous, loud, dirty. Self-driving vehicles are entirely different: safe, quiet, no tailpipe emissions.
I'd easily take extra self-driving vehicles if it reduced human driven ones.
Well yes if we’re arbitrary limiting our choice to car based transportation that makes sense for mild climate cities. But why are we insisting on cars being the backbone?
No limits. Each option should be evaluated on its merits.
My contention is that in US cities the high cost of existing rail makes it uncompetitive for most uses, and there is no justification for building new rail.
Trains are great when going to tourist attractions, especially in the center of old cities.
When you live and work in a city, they're much less practical.
This is quite the "I have never lived anywhere else other than North America" take.
Rail and other public transport in pretty much everywhere in the world are designed to serve commute first, tourist stuff second or third.
Public transport isn't just having some trains, or having only trains between major cities. It is designing whole commute routes from various urban and suburban areas to workplace. There needs to be regional and suburban links that arrive to metro and tram stations. Metro and tram have to operate very frequently to handle commuters. The frequency of the trains should adapt to the commuters in the morning and evening. They need to be convenient, clean and safe too.
Cities around the world are also much better balanced than NA ones. The workplaces and living areas are almost always mixed rather than having a "downtown" area where every office worker travels to. My area has many buildings with a supermarket, apartments and small offices in the same building. There are two car factories in the city next to one of the biggest urban parks.
Is that why the trains and trams are crowded around commute? Because people find them impractical?
> Trains are great when going to tourist attractions, especially in the center of old cities. When you live and work in a city, they're much less practical.
This is the most "tell me you live in America without telling me you live in America" thing I've seen in a long time...
America basically the only place in the world where in its cities, trains and other public transport aren't a major part of people's lives. In other places (Seoul, Tokyo, many European cities, etc.), even people who own a car will sometimes commute via train due to the convenience.
Try a bicycle or a stroll instead of embracing the WALL-E.
If you feel that way about transit you may not have tried a good transit option like Hong Kong MTR with 90 second headways and travel from and to substantially everywhere you want to be.
>Try a bicycle or a stroll instead of embracing the WALL-E.
You see a robot driving around in a pile of trash.
I see a robot with nobody micromanaging him telling him how to live his life, etc, etc.
<we are not the same meme dot jpeg>
I think the answer to this is microbility bike/scooter sharing (ex: lime)
Trains to cover the longer distance and micro mobility options to get to exactly where you need to go
Fairly unreliably? Unlike cars, trains do not typically suffer from traffic jams.
This is based on my personal experience, I used to ride trains for travel a lot. I grew up in Europe and lived there for 31 years so this is not based on ignorance.
Private cars seldom fail to work because the drivers are striking to reduce their hours to 32 hours a week like London last week.
Buddy the tube seldom fails for that reason either. Plus some self-driving sauce would reduce their hours to 0. Certain lines in London like the DLR are already driverless (Grade of Automation 3). Most of the other lines are GoA2.
Yeah, first time for that particular one but the transport not working for various reasons is not that unheard of.
Trains still don't solve last mile transport for most people (even in places with robust transit systems)
Self driving cars might not solve traffic problems but they could greatly reduce them. Problems like traffic waves and gridlock go away when all cars are driving themselves.
The last mile problem is only a problem because of poor layout. Build homes and work near transit nodes (instead of in the middle of nowhere) and there isn't a problem in the first place.
> Problems like traffic waves and gridlock go away when all cars are driving themselves.
How would that make those problems go away? It could probably slightly alleviate them in marginal cases, but any given road has a finite throughput limitation, and once it is reached, it wouldn't matter even if every robo-driver were perfectly synchronized.
> Trains still don't solve last mile transport for most people
This has not been my experience since moving to Manhattan last January. Subways, alone, close the gap between regional rail and most destinations astoundingly well. I haven't yet needed to use a bus (but they seem abundant, too), and I haven't even thought of taking a taxi yet.
Here, robust transit has solved the last mile problem for most people.
The last mile is a solved problem. Most people can walk (and many of those who can't would need human assistance anyway). And then there are bikes, electric scooters, and other light vehicles that use space much more efficiently than a car.
Self-driving cars may help with the actual weakness of transit, which is the long tail of trips. Trips on routes with too few passengers to justify good transit service, and with the trips too long for the last-mile solutions.
Walking a mile with groceries, a baby, furniture, etc. is not really a solution.
I'm not saying self driving cars are the solution, but they are a piece of the solution.
Walking a mile with groceries or a baby is common. People in less car-oriented neighborhoods typically do quick visits to a grocery store when it's convenient for them several times a week, rather getting a week's haul of groceries in a single visit.
With furniture, you usually pay for delivery. Especially because the furniture store probably doesn't have the items you bought on site anyway.
> Trains are the only way to address traffic.
And how do you get to the train when it's too far to walk and you're not a cyclist?
Ideally: there's a train close enough to walk, or a bus or tram that's nearby that runs frequently, is clean, and doesn't get stuck in traffic because there's not much car traffic.
Slightly more realistic: enough people can and do cycle or walk to the train that pressure is relieved on the roads for those who cannot cycle or walk.
> Trains are self-driving. Europe already has the better self-driving system.
Well, I'm in Europe and it ain't here. Waymo can't get here fast enough.
Mobileeye in Munich https://www.mobileye.com/blog/self-driving-robotaxi-sixt-ger...
Moia (Volkswagen) in Hamburg https://www.moia.io/en
Mercedes autonomous driving https://group.mercedes-benz.com/innovations/product-innovati...
Yes, but they said "meaningful".
There's some self driving tech being developed in Europe, but AFAIK nothing is at the current deployment level of Zoox or Aurora, let alone Waymo.
Does it matter where it's developed though? Once it's good enough to expand into all major US cities they could look into deploying in Europe too.
Im happy to let Americans be the beta testers
For the consumer, maybe not, other than a delay of some years.
In terms of having the industry? Absolutely. How many other areas of "tech" has Europe basically punted on and ceded to Americans? Currently there's some gnashing of teeth across the pond for how there's no real European equivalent to the big US cloud providers (AWS, Azure, GCP).
There doesn't have to be an equivalent of everything, I wouldn't want to use US cloud because of price and governance. At most I use the "cloudy" services and rent "capacity" from a European provider, companies are fleeing the cloud. They're done subsidizing Amazon deliveries.
MobilEye and Mercedes works on self-driving, so does BMW. It's probably not Waymo quality, but just because there aren't cars on the (wide and car friendly) roads doesn't mean nothing is happening.
Meanwhile Europe has solid infrastructure for electricity (esp France), ASML has no competition, Carl Zeizz is world leading in optics, there's probably a Leica LIDAR in the Waymo cars... I mean while we're throwing pies and bringing up other markets..
My old boss was working on a project with Leica where he was working with some partner on self-driving industrial machines, they we're using Leica gear for collosion avoidance and such.
Europe doesn't need self-driving cars, we have alternative modes of transportation. Where it's needed (mines and industry) it's already there. And whatever modern car you're driving here has ADAS which helps make driving comforable.
USA is huge. This is happening in a small part of the USA in a very limited fashion. It's not like the USA has driverless cars everywhere, 99.9% of the population never saw one.
I'd guess Waymo covers 5% now. San Francisco, San Jose, Los Angeles, Austin, and Phoenix are ~10% of US population. Waymo service areas don't cover all of those cities.
Considering tourism and people living just outside service areas who see them but don't get to use them (which includes me sadly) I would not be surprised if 10% of population had seen at least one.
I saw one in New York the other day!
About 43% of the US population lives in 25 metro areas so Waymo doesn't have to be in a lot of places to have a big impact.
“More than 50 cities across China have introduced testing-friendly policies for autonomous vehicles.”[0]
Europe could do the same but they have other priorities.
0. https://restofworld.org/2025/robotaxi-waymo-apollo-go/
UK:
>pilots of self-driving taxi- and bus-like services will be brought forward by a year to spring 2026, attracting investment and making the UK one of the world leaders in this technology
I'm wondering how self-driving cars will solve the priority problem of narrow streets of UK towns where drivers need to let each other pass all the time.
I've wondered that myself. It seems quite challenging for human drivers at times. Around Ladbroke Grove you quite often get some complicated jam with two busses and about ten cars stuck.
Yeah I'll believe this when I see it. Most UK roads are significantly harder to drive on than anything in the US. That's why they always test these things in Milton Keynes.
Also a lot of UK driving requires communication with other drivers (letting people out, etc.) in a way that US roads don't. I'm not sure how driverless cars can handle that.
I really wish we could get them, because they're great. But I'd say we're talking 10 years behind the US simply because of the extra engineering challenge.
> I’d really like to see either a Waymo competitor emerge in Europe, or even Waymo themselves operating here
I think you’ll see American and Chinese self-driving kit in Europe once it matures. It’s just easier to iterate at home, so while the technology advances that’s where it will be.
Maybe there just not enough interest? After all there is good public transportation (especially rail), increasing biking habits and just loving the driving experience.
> As a European, I can’t help but feel a bit sad that we’re missing out on the driverless side of things
I don't know about other countries, but Spain will probably be one of the last ones to get it, thanks to the Uber-powerful (heh) taxi driver lobby
Cruise is basically winding down. Tesla is the other major competitor
I don't remember any plans Waymo has announced for Europe, but they are testing in Japan.
As an American with extensive time spent in Europe, I’d much, much rather have European-style metros and tramways than self-driving cars.
Waymo (though a technical marvel) is a bandaid over our inability to build and maintain public infrastructure. Be sure to cherish what you’ve got.
European cities have lots of taxis. Same with Asian cities. They will obviously have AVs in the future. I'm not sure why you think they should be mutually exclusive with transit.
Many American cities don't have the population density to make metros and trams economically viable. And those few cities that do have comparable density (New York, Chicago, namely) do have metros.
Public infrastructure has high overhead costs, and low population density means there isn't enough ridership to make it viable.
Rotterdam — a city with a population of around 650,000 — has both a metro and a tram system. Extraordinary density is not a prerequisite.
And in any case, there's no reason that public transit needs to be self-funded. We don't expect the same of most of our other public services.
Rotterdam has 3,000 people per square kilometer. Contrast that with the San Francisco bay peninsula's ~1,100 people per square kilometer.
This is demonstrating my point about population density and transit.
The peninsula might not be dense, but San Francisco has a density of 7,194/km2 and the transit situation pales in comparison to Rotterdam's.
There are many urban areas in the US with population density of 3,000/km2 or higher that do not have any public transit at all.
People in the thread are asking why people ride Waymo to SFO, which is well outside San Francisco proper. Thus, the whole peninsula's density is what's relevant.
The US does not have many metro areas with population densities above 3,000/km2. And those that do, like Washington D.C, NYC, Boston, Chicago, do have metro systems.
American public transit construction costs are now ridiculous in terms of both money and political capital. Even somewhere as sprawled as San Jose now requires well over 1b/mi to build a subway under; BART could've acquired an entire autonomous driving company for the cost of the Silicon Valley extension.
As an American, I think you’re naive and short-sighted.
You must realize that, at some point, self-driving cars will be ubiquitous. Maybe not for 50 years, but they will be.
What you’re actually saying is “I’m virtue-signaling with Europe because that’s what the cool kids do”
…What? What sort of terminally online strawman would be spending his free time “virtue-signaling with Europe” to some anonymous bozos on a tech forum? What a dull and intellectually uncurious reply.
I think self-driving cars may eventually become common in areas where cars are currently common. I think public transit will continue to dominate in parts of the world where it currently dominates, because it is simply a superior user experience for the majority of people when the government cares to invest in it. (Not to mention far cheaper and more egalitarian.)
I am conveying my lived experience in most European cities I've been to.
> a superior user experience
A superior user experience is going exactly from where I am to where I want to be safely, quickly, and affordably. Self-driving cars are looking really good for those criteria.
$20+ per ride is affordable? Waiting 10m+ for your ride and slowly sifting through traffic is quick?
In London, Paris, or St. Petersburg, I pay a few bucks to hop on a train that runs every few minutes and rapidly end up across town, roughly in the area I need to be. It's literally the cheapest and fastest way to get from point A to point B, not to mention tested at scale and thoroughly battle-hardened over the course of a century.
Not every city has this privilege, of course, but surface trams are 80% of the way there, especially if they have right-of-way. And they don't make pedestrians' lives a living hell: https://www.youtube.com/watch?v=bNTg9EX7MLw
> $20+ per ride is affordable?
In the US, the unsubsidized price of a ticket is close to this amount.
> Waiting 10m+ for your ride and slowly sifting through traffic is quick?
In my city, it's difficult to pick any 2 points that are faster to get between by public transit vs. taxi.
Every city is different, but trains rarely make sense in the US (outside of NYC).
Right of way is the huge advantage of trains, it would be great if self-driving vehicles could have that same advantage.
The core for a good experience is a good structure.
In many regions of the U.S. people live too far apart, shops and businesses are zoned apart into wide spread business areas. Public transport won't provide a good experience.
In a notable part of European cities people live in denser quarters, where a "third place" is reachable in walking distance, some degree of shipping, doctor visits, work are close by. There public transport can fill the gaps for the remaining trips in an (space) efficient way. Self driving cars however would clog the area.
Adapting US settlement structure to allow public transport won't happen. However a self-driving car can turn the dial for individuals to move out of the urban European area into more rural areas. Question is how big that group is.
Try moving a few bags of sod and mulch via public transit. Condescending tone is condescending.
Designing our urban transit around the needs of the mulch-bearing 0.1% seems like a bad idea.
Throwing out bullshit statistics like 0.1% is an ignorant take.
See, it’s super easy to be a jerk.
EU’s amazing infrastructure is the Minitel that will prevent it from getting the internet of self-driving.
Subways don’t solve last-mile problems or trucking.
Good. Cars ruin walkable cities, and the last-mile problem can be solved in other ways.
And it's not just the EU. I'm sure that e.g. China and Japan will continue to invest in their excellent public transit infrastructure even when there are more self-driving cars on the road.
Much of Japan's transit infrastructure is private. There's nothing special about transit that means the government has to own it; being a government, it can regulate things without owning them.
Americans have this idea that transit is for poor people, which translates to "it's not important for transit to make money", which translates to "we need to make it illegal for transit to possibly try to make money", so there aren't even vending machines at the platforms. Whereas in Asia they do profitable land development at the transit stations.
> Much of Japan's transit infrastructure is private. There's nothing special about transit that means the government has to own it; being a government, it can regulate things without owning them.
Japan's private transit infrastructure is only private in high-very high density environments (inner-city) and subsidized in low-density environments (rural, cross-country). Ultimately private group transit requires population density above a certain threshold to be viable.
Apollo Go (the Chinese Waymo owned by Baidu) is planning to start road testing in Germany and the UK in 2026, in partnership with Lyft.
Mercedes is quite close. They have demonstrated commercially viable Level 3 ADAS systems.
Cruise has been out of business for almost a year I think.
US and China basically.
Regulation and under investment
Those darn regulators, don't they realise companies just want what's best for us?
We can’t even use Waymo when we land at SFO for a visit
Isn't that what this article is about?
Wayve?
Don't worry, we're missing out on a lot of "progress" on this side of the ocean thanks to Trump's dislike of wind farms and RFK Jr's whole anti-vaxxer thing
One thing you are missing out on: mandatory loud (97 to 112 db) 1000 Hz audible beep when the vehicle reversing, oh so slowly, such as at the recharging station. Also, constant shop vac five horsepower vacuum cleaner sound. BEEP BEEP BEEP BEEP BEEP. VROOOOOOOOOOOOOOOOOOOOOOOOOOOOM
Oh wait, you thought those would be in the middle of nowhere? Nope.
https://www.karmactive.com/waymo-charging-noise-blasts-112-d...
Unless and until those noises that you mention are as annoying as those made by present time ICE vehicles, your point will remain irrelevant.
From the link:
These backup warning systems operate at approximately 1,000 Hz, producing sound levels between 97 and 112 decibels.
Santa Monica’s municipal code adds another layer of complexity, prescribing exterior noise limits of approximately 50 decibels during the day and 40 decibels at night.
The continuous operation—with vehicles reversing dozens of times hourly, including during late-night hours—continues to challenge community peace.
So, constant car screaming BACKUP BACKUP BACKUP I'M BACKING UP HERE right outside your residential window. Kinda sucks. A whole lot.
Ironically the car is virtually purpose built to not run over people, and likewise has an extensive sensor suite to detect people around it.
I suppose regulations don't care if you can see no one is behind you.
I'm surprised and incredibly impressed at this announcement. It seems trivial, but the general feeling in the industry has been that SF would fight tooth and nail against robotaxis at SFO.
Probably because SFO felt the heat after Waymo acquired SJC approval quickly: https://sfstandard.com/2025/09/05/phoenix-has-airport-robota...
the major difference being that SJC is easily accessible by surface streets whereas SFO mostly isn't.
Yes, but SFO (and the city of SF) has a long history of being hostile towards Waymo: https://missionlocal.org/2023/11/waymo-rebuffed-by-sfo-sf-gu...
What likely happened now is that SFO got a kick up their backside from the Mayor after the press started asking why it was still dragging its feet, while SJC approved Waymo swiftly.
I genuinely think things have changed with Lurie as mayor and 6 growsf endorsed people on the board.
It's going to take a long time for SF to overcome the reputation it built for itself in the 2010s.
Recent changes in the composition of the San Francisco Board of Supervisors (i.e. Peskin being out of government) may have something to do with it being easier than expected.
The NIMBY/landlord supervisors who controlled SF, such as Aaron Peskin and Dean Preston, are now a thing of the past.
Peskin is now reduced to showing up at protests with signs saying the rent is too low.
https://x.com/agarwal/status/1966365908085125384
Waymo got approval for SJC last week. That probably accelerated approval for SFO, which had been stalling. Nice.
When they get clearance to drop people off at the main terminals, that will be more convenient. Pickup at the terminals is harder. There will be a need for a staging area somewhere in the parking structures.
Few major airports I've been to allow Uber/Lyft anywhere near the pickup area, so many fliers are already accustomed to walking a quarter mile or so to their rideshare. But their inability to use the drop-off area is a new inconvenience, and I can see it limiting the appeal.
Waymo will probably get access to the drop-off area after a while. One step at a time seems to be the Waymo way.
Waymo at airports could work really well with automatic dispatching. They already have an app running in the customer's phone. It should be aware of when someone with a reservation gets off an airplane, and how close they're getting to the pickup point. With good coordination, as the customer heads to the arrival lanes, a Waymo pulls out of short-term parking and heads for the meeting point.
A few more years, and humanoid robots will put the luggage in the trunk.
this can be done today with humans. this is just dumb
> Few major airports I've been to allow Uber/Lyft anywhere near the pickup area
Few major airports have Waymo at all. Phoenix has allowed pick-up at the airport for ages. (EDIT: Never mind.)
I'm talking about Uber/Lyft drivers being required by many airports to pick up away from the normal pick-up area, usually down the road a bit or in a parking garage.
Harry Reid International Airport (LAS) is in a parking lot.
What airports are you flying out of? Every major airport i have been to in the last year has a dedicated rideshare pickup lane.
Uber black and at least lyft extra room have no problem picking you up at the arrivals
I drove for Uber/Lyft back in 2020 and let me tell you, SFO is a nightmare. I missed a turn once and had a passenger trying to make a flight furious at me. I quickly figured out there were a group of drivers who specialized in SFO and amatuers like me should avoid the place. When Waymo announced San Jose I thought ok, that makes sense because SJC is easy, but SFO? Wow, I'm impressed. I hope it goes to plan.
Those turn offs for specific terminals are very small and easy to miss.
Nothing more rewarding than a company working hard and seeing real-world, first of its kind results in action. Makes me feel giddy about a company again like peak tech back in the 2010 era.
Congrats to the Waymo team, I’m sure this was a huge milestone internally.
Does this mean they'll be able to take the freeways to get there? Surface streets from SF to SFO would be pretty slow.
I'd hope so. As an aside, I wish Waymo was more transparent on the app that their cars are not allowed to take passengers on the freeway. I was unaware of this restriction when I booked a ride from SF to Burlingame last month and I was stuck in a Waymo for an hour going down residential streets!
Doesnt it show the route and the ETA before your book the ride?
No, they need to add a pop up with even more text users will not read.
They have had permission to be on freeways for a while [0], although so far they have only done employee testing (I believe)
[0] https://sfstandard.com/2024/03/01/waymo-san-francisco-cpuc-e...
I wonder how that'd feel. I took a Waymo in SF last fall and I was pretty impressed. But it was also slow city speeds. I wonder if it feels different going at freeway speeds with "no one" at the wheel.
While the margin of error is much lower on a freeway due to the speeds, other drivers are generally a lot more predictable (also in part due to the speeds).
Sure - a good freeway is actually a lot more predictable in most circumstances than city driving, so as a problem to solve it's likely a little bit less complicated. What I wonder about is what it feels like as a passenger. I wonder if it would be more or less frightening than being a passenger when my 17 year old is driving.
I use adaptive cruise control a lot, where I rely on the car for keeping a safe distance.
I have a limited version of SuperCruise which means it operates hands-free on freeways but nowhere else. My wife's Equinox EV has the regular version, which operates on a lot of arterials near us and has more capabilities. The first time that the Equinox signaled, changed lanes to pass, signaled, then changed lanes back was shocking.
We moved to a small town and drive a lot more than we used to and I find that having those capabilities really helps relieve the stress.
I will say that I move to the center lane when going through a notorious set of curves on I-5 in Portland because my Bolt doesn't steer as smoothly as I'd like near the concrete barricades. I wanted SuperCruise because it has a fantastic safety record. There are lots of times it's not available but when it is, I have near-total confidence in it.
I took a Waymo that drove on an 'expressway' which had a speed limit of 40mph and it was definitely a different feeling. I did feel a bit scared, at 25mph it feels like a gentle theme park ride, at 40mph it's beyond that and feels dangerous.
There was a good overview on here a while ago about the challenges[1]. You need to plan longer in the future and your sensors need to reach further. It's also a much bigger challenge to collect sensor data as fewer diversions happen per mile (but those that do have higher stakes).
[1] https://news.ycombinator.com/item?id=38973404
Roads that get used more collect more debris. They also break and require maintenance more often. That maintenance is exceptionally disruptive to the normal operation of the road.
Other drivers aren't your only challenge out there.
From the article “ Pickups and dropoffs will initially start at SFO’s Kiss & Fly area – a short AirTrain ride from the terminals – with the intention to explore other locations at the airport in the future.”
They're still working on freeways, doing employee riding testing.
Waymo already had the permit, but they're just being (overly) cautious.
The surface street route that bypasses 101 near Brisbane is surprisingly often faster than 101.
People love crashing there.
Looks like this Kiss & Fly area where pickup will be is at the car rental center.
Oh, this makes a bit of sense. The Avis/Budget fleet team will be part of managing the vehicles, so they can be quickly cleaned and fueled up when they slide into the airport, too.
https://www.avisbudgetgroup.com/avis-budget-group-announces-...
(Dallas, but they do this in other cities, too.)
huh. I didn't even know this existed.
Same. I go to the rectal car center at least 4 times each year. I just was there on Saturday and had no idea either. Still don't know what it is other than Waymo pickup.
>rectal car center
Known nickname or typo?
Definitely phone autocorrect issue. I'm gonna leave it though
How often do you type "rectal" for that to become an autocorrect default for you??
If you are over 50, and serious about not getting colon cancer, maybe a little bit more than one would expect.
OP is a urologist
I use Google keyboard without customized auto correct.
It really likes to change random words to inappropriate things.
But I guess that's the people who are typing on phones a lot are typing about.
Otherwise known by its popular name “Cloaca-Rent—A-Car”
Or both :-D
It - along with cell phone waiting lots - are ways for people to drop others off and avoid the traffic around the terminals themselves.
Which can be bad - I often find it easier to just pay for a few minutes parking on dropoff/pickup.
Yeah, but the SFO cell phone lot is particularly bad, and traffic to the terminals is bad.
On the flip side, there are airports like Cleveland where people just park their car at arrivals and disappear for 20 minutes.
I did always find the term kiss and fly confusing and weirdly intimate, as if everyone is getting a ride to the airport from a spouse or parent. Definitely a throwback to another era.
I think it's also a regional thing; I'd never heard of it.
> rectal car center
That's way mo' information than needed thanks.
But seriously. I wonder why they have a designated pickup point if it would make sense to spread the cars out to alleviate traffic bottlenecks.
What's even better is the variety of names this thing has. I'm my area, it's the "cell phone lane"
Hopefully Waymo does a better job than SF Uber drivers. I can't tell you how many times I've had drivers make a wrong turn trying to find their way to the pickup point.
Boy, if they could actually navigate terminal traffic, I’d give ‘em true self driving.
SFO traffic is not bad at all. Send them to LAX and we're talking.
Its not true self driving until the Waymo asks if dropping you off at arrivals is ok.
There's the big sign there telling you to go to arrivals for drop-off. This is probably a stupid question but can Waymo cars interpret those temporary display signs and follow them? Would it?
It seems to handle the standardized ones (think "construction ahead, detour left") perfectly well from the rides I've taken, but there's all sorts of ways they could be 'cheating' on that.
Thats AGI
I'll be honest, I think LAX's traffic is better than SFO's. It feels like there's a lot less spaghetti at LAX, the shortcuts are reasonable, and you don't have separate international and domestic loops.
LAX's many parking lots with left lane entrances definitely threw me for a loop the first couple of times.
Overall though, I think I agree with you.
JFK is probably the 10th circle of hell
I don't touch JFK with 10mi pole. I've always found EWR to be much more consistent and easier to get to
Send them to BOS and we're talking
Is the mark of intelligence being able to navigate to BOS, or refusing to drive through the big dig in the first place?
The Big Dig, for all the digs it rightfully got for taking forever and costing a shitton, actually does the job it's supposed to (mostly). I'm generally pleasantly surprised how few problems it had when I lived there.
I didn't go to Logan a ton though.
JFK has entered the chat
I'll take JFK over LAX. The construction going on right now at JFK sucks, but LAX is comically bad. Just last week I was on a rental car shuttle at LAX and watched 3 separate groups of people at different terminals miss their flights because traffic just wasn't moving.
Maybe, but this approvals only allows them to go to the rental car center, which is quite far from the terminals. The passenger will need to take the air train to the terminals.
They already do in PHX.
Can't wait to argue over outsourced call centers to a non human while being stuck in traffic in a dangerous situation
Seems like Tesla keeps talking big, while waymo conquers city by city.
I think this is a really exciting development for Waymo and the future of autonomous transportation. Imagine the next steps; waymo flights
Can you handle parking structures? I heard a lot of the autonomous cars were using 2D maps and couldn't handle multiple levels. Haha! This was just a year or two ago.
Google maps has been able to figure out parking structures for me recently. Not sure what technology is involved (gps isn't great for vertical) but it's clearly possible.
Google has been collecting data on building interiors for several years now. Not just parking structures. This data is currently used in streetview. Google's geospatial data is unequaled and maybe a bigger advantage than is readily visible.
Do they need a "map" of a parking tower though, just like how humans don't exactly need Google Maps inside of one? I feel like this is something self driving + vision (exit signs and arrows) can handle
Do taxis need to park tho?
I mean, depending on the situation, of course. Do taxi drivers in US drop people right in the middle of a busy street?
Kinda yeah? They certainly don't navigate into long-term multi-storey parking structures.
one of the waymo depots in SF is a multi level parking building
Waymo ride costs are getting really expensive in SF.
Not sure if you have a recent side-by-side example with Uber, but this seems like it would have to happen if the demand is there. How else can you offer a quality product (i.e., car shows up in a reasonable amount of time) if you don't have enough cars to satisfy the demand? Pricing is the primary demand lever.
There's so much polarizing opinion on Tesla's offering and whether they'll get to Waymo's level sooner than later, but this seems like it's going to be or already is a huge issue for Waymo where they can't manufacture the vehicles fast enough to satisfy the demand as they expand both locally (because they capture more of the market) and into new geographies. Will they try and acquire a manufacturer? I don't think that's economically feasible for Waymo (Geely market cap is $25b, per Google snippet fwiw), and obviously being in the car business is different than autonomous, but I'm sure Google would bankroll a purchase if they thought it was the right growth strategy.
I guess Tesla, even if their autonomous is on par with Waymo tomorrow, also has to manufacture the fleet, but it seems extremely beneficial to have that capacity in house vs. relying on partners. Maybe I'm wrong and it's not that much of an advantage, but at first glance it would seem to be.
The CEO of Uber was quoted as saying Waymos complete more rides per day than 99% of Uber drivers. He didn't give a precise ratio but this makes me think that hundreds of Waymos can replace thousands of Uber drivers and their cars.
CMs like Magna have the flexibility to manufacture, at the low end, hundreds of vehicles, and at the high end thousands. I doubt Waymo will ever make their own vehicles. They are already working with Toyota on adapting Waymo technology to privately owned cars. That implies mass production. That would be a supply of vehicles that are probably simple to adapt to robotaxi use.
That's a crazy statistic and an interesting one for him to actually say out loud. Was that in the context of Uber partnering with Waymo in Austin? And thanks for the insight on the manufacturing side. Sounds like it might actually be to their advantage to use third parties because you can spread the demand around and since auto margins are not high the added cost for that benefit is minimal.
Private cars have a ridiculously low duty cycle. They mainly sit around waiting for their owners to use them. I suppose at some point in the future there might be a traffic jam of autonomous vehicles, but only if the providers are antisocial and don't coordinate ride destinations and routes.
@harmmonica I do. we prefer to use Waymos in SF but Uber has been a lot cheaper in the last six months or so regardless of time of day... Also saw some Zoox self driving boxes on the las vegas strip last week but no one seemed to be using them.
Thanks for sharing, oliver. For anything local I've almost entirely switched and I guess I haven't been doing much price comparison between the two. One thing I have noticed here in LA, albeit only a couple of times, is that during rush hour the waiting time is significantly longer for Waymo. I've taken some Ubers because the wait for a Waymo has been way too long.
Zoox just started allowing rides in Vegas 6 days ago.
I've seen them driving around SF as well, but they're not yet available here.
Waymo is a premium ride product that happens to be self driving.
just use Robotaxi. 1/3 of the price, sometimes less
How much per mile? For some recent example rides, let's say. One I took was exactly $2 / mile but not in SF.
What's that? An app? I see a Chinese app of that name in the android play store, but it only has about 1k downloads
Only for iPhone at the moment
https://apps.apple.com/us/app/tesla-robotaxi/id6744257048
The Tesla service is colloquially called "Robotaxi".
Not just colloquially but officially...
https://www.tesla.com/robotaxi
For now...
Not sure why you're downvoted. I've tried Robotaxi a few times and has been great. They still have a safety driver these days and wait time is a big high though.
But still no Waymo app in the European App Store for iOS.
I wonder what the ultimate price of this service will be compared to alternatives.
It will remain higher for a while. From reporting I have seen, they are close to maxing out their vehicles, and many people prefer it to other options, so are willing to pay a premium. As long as that is true, it's going to be priced as a premium product. It won't be until fleets grow significantly in size and/or another driverless taxi service enters the market that we will maybe start to see prices driven down closer to marginal cost of a ride.
-edit- multiple other comments apparently disagree with this. I'll defer to people who actually use them over the reporting. Odd that there is that disconnect though.
Yeah they need scaling and competition before the prices get lower. As long as supply is saturated with demand and nobody else is on their level, there's little reason to lower prices.
Yeah, and just to add even though it's implied in your comment, there's plenty of reason to keep prices where they are independent of a desire to increase revenue. Customers will not wait forever for the car and so if the demand is high you have to keep the price high to discourage people from using it so wait times remain in check. Tricky tightrope they're going to be walking while they optimize the fleet size for local adoption and geographic expansion.
It's also higher right now because it is a novely. Plenty of people are booking it just to say they rode in a Waymo and take pictures. When that wears off they will have to start competing strictily on price and wait/ride time.
Lots of people, myself included, pay a premium to not have a human at the steering wheel; it's nice to have the car to yourself.
Yourself and three dozen recording devices and call centers full of people tracking the car and reviewing the footage, yes.
To be fair, Waymo claims to not record or transmit audio without you either manually engaging such (by requesting support), or a very unambiguous announcement (presumably when the car gets into some sort of emergency situation). And lying about that claim would probably run afoul of California's 2 party consent law. So still a step up in privacy versus having someone in the car listening in on your conversation.
That said, even if they were listening to you, there's a lot of things that are completely inconsequential from a perspective of an anonymous call center employee far away listening in on, that I probably wouldn't want to talk about in front of a taxi driver.
I know this is somewhat besides the point of the discussion, but.. many Ubers have recording devices inside the car too. The drivers have gotten savvy and protect themselves from false claims or even harassment.
I still count that as a win.
Like, the driver's presence bothers you? Even if they don't talk?
This is just me, but maybe helps explain it. It's not that the presence of a driver is bothersome, but in the pre-Waymo world your interaction with the outside world starts when you step out the door of your house. Now the interaction with the outside world starts when you get to your destination and step out of the Waymo. I really enjoy the outside world, mind you. But it just feels easier to traverse my local area in solitude and with a consistent and comfortable vehicle, and non-erratic driving style.
I imagine how nice it could (will?) be when you can hop into a self-driving car for a longer ride or even a road trip. I think you'll feel like it's an extension of your living room vs. being in a car.
Christ that's depressing. A 'perfect' world where we never interact with anyone.
We need to stop normalising mental diseases
If you step back do you really think that's indicative of a mental disease? Does it make any difference to you that many times I'm taking a Waymo to go and hang out with friends? Not much of a stretch to say it's allowing me to socialize more because I don't have to worry about my meter running dry, or having one too many drinks to drive myself home, or being able to move around from area to area in comfort. And if you say "you can do that with an Uber too!" it's true! But does it really surprise you that someone would want a car that drives calmly, obeys all traffic laws and gives you a little downtime from the outside world pre or post the activity you'd been doing before stepping in the car? Does that really rise to the level of mental disease?
It seems like a huge catastrophizing stretch to get there based solely on preferring to be in a Waymo rather than a taxi or Uber.
edit: grammar
Well on second reading your comment reads like an ad, given advertising isn't natural either, you can understand the confusion
But I was referring to the wanting the outside world to resemble your house and to have little interaction with humans. No, that's not normal, despite any sophistry or ad speak
The problem isn't when they don't talk and just drive, the problem is when it's late at night and the passenger is a woman who is inebriated. Not having a driver entirely makes that much harder.
I know this is only a single data point, but I recently took one in Hollywood. Uber Lyft quoted $33 and Waymo was $20
On other threads I've seen conflicting anecdata regarding pricing being higher or lower than an Uber ride. That's not too surprising since the supply and demand variables are going to be different for Waymo.
Their goal is to have lower cost Hyundai models hit the market though, right? So the Jags probably remain the premium/higher cost option.
In my experience so far, Waymo costs about the same as an Uber when you take into account tipping, but takes longer (they're not yet doing freeways). With the addition of SFO to their zone, I can't imagine freeways are far behind, because getting from the city to SFO without using the freeways would be... a novelty.
That's not been my experience... 90% of the time when I check, Waymo is still a good 20-50% more expensive in SF, when comparing to a tip-included Uber or Lyft price.
I've used Waymo countless times in SF. It's typically 15% cheaper than an Uber/Lyft and trip time/wait are generally the same. I much prefer the Waymo.
I've never encountered it being cheaper, what hours do you generally use it?
Generally between 11a and 7p. Going to lunch/dinner.
Self driving taxis are fundamentally much cheaper to provide.
- No driver to pay. - Smaller simpler car. - Can drive 24h a day. - Needs much less parking space.
But fully realizing these benefits is probably a decade away.
Cheaper than uber rn. Long term once they own the market? Too much.
They still have to compete with alternate modes of transportation such as buses, bikes, trains, e-scooter rentals, self-owned cars, Uber with human drivers.
If it would be "too much", then there's no reason why taxis (incl uber/lyft) wouldn't be too much today.
I don’t really think they have to compete much.
Direct competitors are uber and Lyft which they can undercut since they don’t pay drivers.
The people who want to take buses and trains will continue to do so although Waymo might sway some with their ease and if pricing is reasonable.
Bikes and e-scooters only get you so far. Last time I was in SF I didn’t see too many bikes but I saw a ton of e-scooters. Are you really taking an e-scooter further than a few blocks? And when it rains?
Self owned cars make sense for longer trips out of the city but parking is a pain and driving is stressful so this is an easy win for Waymo.
It’s cheaper now so they can take market share. And their cost will certainly be lower than Ubers so they can win the pricing battle. But long term monopoly gonna monopoly. Perfect pricing is a given with the wealth in SF and how many rides will be on a business CC.
from what I heard, the intention is to make it much more affordable than it is now. I don't remember the source right now but I did think it was a blog post or something like that.
I think if it's affordable then people will easily take that. instead of drinking and driving at night or other unsafe activities. if it's affordable then people can just take a waymo home and then back again to get their car when it's safe again.
Certainly they aim to make it affordable now in order to undercut Lyft and uber. Long term they will own the market and jack up prices as monopolies do.
That's great to hear
The title makes it sound like GA but it's still in testing
at first I thought they were doing those cargo quad copter things...
Looks like this would (eventually) include service to not just San Francisco, but also the Peninsula (Silicon Valley) via freeways.
I certainly hope so. And if yes, I imagine people are gonna start using it as a transfer point, to take a Waymo from the peninsula to SF.
Is Waymo L5?
L5 means the car can drive everywhere a human can. Waymo's refuse to drive outside of a constrained area, and occasionally stop to ask for assistance, so that makes them L4.
This whole autonomous driving levels kinda muddies the waters. Some would argue this isn't full L4 even. But it is a self driving car in the places it offers its services.
What would be the argument that Waymo is anything except L4?
No, L5 is a car that can drive itself anywhere in any conditions.
I think there's an implicit "where a decent human driver could drive safely" for L5, otherwise you get increasingly ridiculous scenarios like, "can Waymo drive safely in a whiteout blizzard?" or "can Waymo safely escape an erupting volcano??"
Too bad waymo is more expensive than uber most of the time
Why is that too bad?
is waymo really that good???
how good it compared to Tesla FSD/Robotaxi ???
haven ridden in both a few times, yes, Waymo is head and shoulders better. It's smooth and I don't think I've ever seen any false alarms or behavior that made me feel unsafe in a Waymo, while I've had a few scary or annoying situations in the Teslas. I took a 6-minute robotaxi in drizzling weather where it parked in intersections twice because the cameras were obscured. Meanwhile Waymo can drive perfectly in heavy fog.
Both the Waymos and Teslas have that central display that shows you what the car sees (pedestrians, dogs, traffic cones, other cars, etc). The Waymo representation of the world reaches pretty far is is pretty much perfect from what I've seen. Meanwhile the Tesla one until recently had objects popping in and out.
Neither is perfect, of course; both will hesitate sometimes and creep along when (IMO) they should commit. But they're both still way better in that regard compared to the zoox autonomous cars I see in SF.
Tesla doesn't have a real robotaxi yet, they're still in the testing/prototyping phase where they need a safety driver or safety monitor in the car.
They might be close to a real robotaxi in some areas, but it's hard to say until they actually pull the trigger on removing any employees from the car.
tesla robotaxi is worse than waymo was 3 years ago when I was a tester
Until just now I had no idea Tesla had a taxi service. Otoh I've seen hundreds of Waymos in SF and the west side of LA.
An interesting thing about this is that there are fewer than 1000 Waymos in the SF service area. I don't know today's total, but I'm pretty certain that there are fewer than 5000 Waymos in existence. Maybe as few as half that.
Some months ago Waymo claimed to be providing 250,000 rides per week. If the fleet size was 2500 at the time, that would be 100 rides per vehicle per week.
Wait, what is special about driving to/from airports?
What’s special about the airport is that the City of San Francisco owns and regulates it (as opposed to the streets that are regulated by the state CPUC), and the Board of Supervisors previously were regulatory captured by taxi medallion owners and Teamsters union (https://missionlocal.org/2024/12/waymo-rolls-toward-san-fran...). Specifically, Aaron Peskin (BoS supervisor from 2001–2009, 2015–2025, and board president for the last 2 years) said, “Their entire M.O. is, ‘The state regulates us; we don’t have to work with you, we don’t have to partner with you.’ My response is: There are things we do control. Including where you charge your cars. And the airport. What I intend to do, is condition their deployment and use of the airport property on their meeting a number of conditions around meeting this city’s minimum standards for public safety and transit.” https://missionlocal.org/2023/11/waymo-rebuffed-by-sfo-sf-gu...
I’d say it puts a lot of Uber (and similar) drivers at risk because airport rides are a good source of income. Waymo undercutting them will reduce the amount of passengers available for pick up. Not saying it’s a bad or good thing. Just that it has real world impact on people and the economy.
Usually you'd have to take the BART one stop then the waymo, which seems to be a common tourist attraction for fresh deplaners. Perhaps the airport was afraid without that step of friction, too many people would try this and cause a waymo-jam
Isn't it by far and wide the most common use of taxi services? It certainly is basically the only time I ever use one.
Waymo getting into that space seems like a pretty big step up in market penetration.
Hope you like traffic!
Waymo can deliver as many rides as Uber with a small fraction of the number of vehicles.
And they're all electric.
How would this reduce or increase traffic? The demand is staying the same.
Presumably the increased supply of "drivers" going to SFO will lower rideshare prices for everyone and make public transit less appealing
Public transit connections to SFO are shit-tier. Rideshare prices have little to do with it.
You can BART from there to downtown, that beats a lot of cities. (yes, yes, BART quality can vary, but still.)
surge pricing FTW!
Have you priced this out compared to a regular taxi or Lyft?
It’s waaaay mo’
Inside SF, my experience is that Uber and Lyft are ~10-15% cheaper than Waymo, but that's before tipping. I don't have to tip a robot, so they work out to nearly identical prices.
You don't have to tip an Uber or Lyft driver either.
Preach.
You also don't have to tip a taxi driver. They get paid for giving rides, it's not an extra service.
Yup, but it is kinda culturally expected to tip. You're right, you dont need to, but then again we do a lot of things just because it is .. polite.
Waymo food delivery will be incredible. You won't even get your food if you don't tip your ubereats.
The inherent problem there is the edges, most food delivery isn't the trip, it's the person getting out of the vehicle and putting it on your doorstep or going through the building. Zipline and their droneports for buildings seem to have the better solution, at least until waymo has some sort of legged robot that can bring the bag the last meter(s)
I think the frustration with tips is so prevalent that the advertising could just be "Skip the tip, simply walk to the street to pick up your order!"
Would work great in suburbs where a robot car could pull in front of home for a minute or two, your food will be bid to another customer if you don't pick it up in 5 minutes. maybe the little robots in NYC are better.
I would argue that the sidewalk robots are too hard to coordinate and not strong enough to hold up against crime, the solution is somewhat closer to my other comment below, a vehicle with maybe 4 or 8 food cells that can fill up at various locations then make its journey around the city. At that point the problem would be idle timeouts and how to handle disgruntled consumers that lost their window for pickup
Aren't the "first meters" also pretty problematic? Are Waymos going to double park in front of a restaurant waiting for someone to come out and put the right order in the right vehicle?
That's easier to do with training, and the business is usually more willing than a consumer as it increases their business. Anecdotally, see how many of them (at least locally to myself) have adopted the doordash/grubhub tablets in their kitchen ordering system. I imagine it would be a co-packing situation with lockers on wheels similar to the vehicle KFC uses in China: https://www.mashed.com/284555/the-futuristic-way-kfc-is-sell...
Uber's NURO seems to be developing a vehicle with a similar form factor as seen on this page: https://www.nuro.ai/first-responders
EDIT: see comment below, uber does not own NURO
My anecdotal evidence also has so many incorrect orders that I'm a wee bit less optimistic than you about restaurant-side human handling of the first meters. :)
Nuro is an independent company from Uber, the latter just has a partnership with and some investment in the former. Uber has similar relationships with more than half the industry at this point.
A lot of restaurants already have dedicated parking spots where they'll bring the food out.
I feel like that's already quite special-case-y.
Certainly wouldn't apply to anywhere near where I currently live (then again, neither does Waymo).
I have a relative in Texas who is looking into leasing a drone to operate for food delivery. Apparently, that's already a thing there? If we could get food/small packages delivered to our building's roof instead of the front door, it would be a huge win for everyone in the building.
Using an otherwise empty 5 person vehicle to move a grocery bag worth of food is pretty stupid though
If it's a robot, why can't it have 10 lockers, and the right one pops open when it arrives to your place?
The longer the route, the harder it is for the food to stay fresh and warm/cold/frozen. It's a trade-off between efficiency, price, and customer satisfaction.
>I don't have to tip a robot
Now that tips are tax free, it's only a matter of time before some clever SV accountants figure out how make everything a tip.
Self-serve ordering terminals already often ask for tips. Presumably to be legal they're being paid to the kitchen staff, but I think sticking to "tips are for workers who have to pretend to like me" is a pretty firm boundary to stick to.
(Also, arbitrarily reclassifying things as tips is hard, because legally 100% of tip revenue has to go to workers, not management, and certainly not the company's investors or coffers).
That's why they're clever accountants.
Tax-free tips paid to robots go to the hardworking AI engineers -> AI engineers voluntarily donate part of their tips to a 501(c)(3) that helps support struggling venture capitalists.
Something like that. We'll work it out the details once the right PAC donations are in place.
Nobody needs this. This is, in all likelihood, just private enterprise developing technologies for automated warfare.
I see a monopoly about to take shape. DOJ/FTC is sleeping on breakup schemes. USDOT should start government/private ventures in this space.
What kind of monopoly?
They are cornering rideshare and automatic vehicles. They are the sole provider of an automatic vehicle.
Fair enough, I suppose they have a monopoly. Good thing that isn't illegal if they're not anti-competitive about it.
They aren't the only autonomous vehicle for hire service. Zoox is operating in Vegas.
Even if they were the only one, it would be odd to classify autonomous rideshare as a distinct market given they compete directly with other vehicle for hire services where they have nothing close to monopoly-like power.
Okay what're the odds on how long it is until there's a stray Waymo on the tarmac. Hopefully with enough warning to divert any planes about to land on it.
Zero. For the same reason there are never any stray civilian cars on the tarmac.
https://www.msn.com/en-us/news/crime/surveillance-video-show...
https://www.cbsnews.com/chicago/news/delivery-driver-secured...
Lower than a random human doing that.
Cause what this country needs is to automate away even the gig economy jobs that are out there. Let's keep making a few people rich and screw all the normal people out there.
Why the downvotes? That jobs will be lost is fact. Does this represent an increase in wealth concentration? Obviously. Is that a net bad? I don't know, let's discuss instead of silencing people.
Waymo are toast, Tesla will out scale them in months if not sooner. They can't compete on costs, 100k plus for an ugly Waymo vs < 40k for Tesla model Y or cyber cab.