GM exits robotaxi market, will bring Cruise operations in house
cnbc.com
cnbc.com
Isn't there basically Google/Waymo and then, seemingly much further behind, Tesla Cybertaxi, Amazon/Zoox, and Uber/Yandex? Cruise allegedly has one of the most sophisticated autonomous driving platforms, and GM's Super Cruise (if they share any tech) is comparable to Tesla FSD. Strange that they would bow out.
Small anecdote: I visited a GM dealership this week and the salesperson told me Super Cruise was not enabled for test drives. The excuse was pretty weak, like the dealership would have to pay for the service or something. GM might have the technology but they are completely bungling the strategy.
Ford just lowered the cost of its BlueCruise subscription by 1/3rd. In an earnings call eight months prior they remarked they made a 70% margin on the service. It seems like drivers did not find the feature compelling and were not renewing. Interest in autonomous driving appears to be cooling across the board.
IMO, it's largely because BlueCruise (and all of the similar services) is not sufficiently better than the non-subscription driver assists.
I have the option of BlueCruise on my vehicle and it's just not a compelling offer.
* It only works on major highways. Beyond actual coverage, major highways are places where the standard driver assists already work extremely well.
* It does nothing about traffic around you. You have to be prepared to react. Not much different than driver assist.
* It's wickedly expensive.
Right now, radar assisted cruise and camera based lane centering eases 95% of my driving fatigue.
I drive (and have driven) a fair number of rental cars due to travel, and I have to say I feel that way about many new vehicles in their entirety. So much vehicle tech is at best unimpressive and at worst positively in my way as a driver.
> It's wickedly expensive
We own two cars, 6 and 10 years old respectively, yet I've never felt less motivated to look at new cars than I do now.
Guess our part towards saving the planet may well turn out to be 'driving that existing ICE vehicle for just a little bit longer'...
So if you drive an average amount with those cars you're not saving the planet by continuing to drive them.
In a well functioning system you'd be selling them on to people who drive very little and are old and set in their ways and replacing them with an EV.
In the specific case of electric vehicles (and many plug in hybrids) luckily they're such an immense improvement that my guess is that they still have a net benefit. But far from the simple efficiency math, and more likely the user will increase their consumption by some meaningful percent.
(Genuine question, not whataboutism...) what about the environmental aspects of manufacturing and distributing a vehicle other than CO2 emissions?
> old and set in their ways
I fear I may be approaching this stage myself! :)
I don't follow your take on the current EV gen either. A car at the end of the day derives most of its value from taking you one place to another. The current gen of EVs do this and do it well. "current EV gen is kind of doomed" that is just hyperbole, nothing is changing in the near future that would doom current EV vehicles on the road.
Like what?
Honestly, I'm not surprised that new cars aren't breathtakingly awesome. It's been a really rough 5 years for everybody and we're not seeing much genuine light at the end of the tunnel. Everyone's just trying to survive. Hard to do your best work under those circumstances.
Yep, or selling it and buying a more efficient one, so the person buying yours upgrades their less efficient car as well.
> So much vehicle tech is at best unimpressive and at worst positively in my way as a driver.
So you are basically paying to take legal responsibility for the crashes of a not-quite-ready autopilot whilst still having to keep alert.
The moment there is a true self-driving capability, where you can sit back and let the car drive even if you're not sober etc, then it would be wildly popular.
That said, when trying to BUY a supercruise vehicle, the sales guys were clueless, I had to review the stat sheets of each car to see which ones did or did not have it. GM is treating this technology as “surveys say 2% of the market wants a self driving vehicle” (incorrect question) compared to “install on every vehicle as an available subscription, let the sales team earn their way.” I’m sure the tech is expensive, but it can’t be that much more at the OEM level.
I also have a Tesla with FSD. FSD is truly amazing but still struggles with edge cases like, my office has two entrances, one is blocked by barrels, it tries to turn into the barrels every time.
I really don't want to buy a Tesla, but from what I can tell, nobody has anything close to what FSD offers.
From what I read about it, you're absolutely right. Down to the point of unwanted veering towards road dividers and other vehicles.
SuperCruise does what it claims to do, keep it up the middle on listed corridors. Most highways and interstates are eligible. I love both systems.
FSD is truly almost there for unsupervised, point to point driving. People want to virtue signal and complain about Tesla but it is an amazing piece of hardware being mass produced today. My car is 3 years old.
Going from 99.9% to 99.99999% is what makes a system truly driverless and where most of the work is. Waymo is way way ahead of FSD for this.
Sadly this is believable, I've asked to check out things like remote start and control of heating/cooling, and the sales people cannot show those features off because they require an app + subscription tied to the car.
I worked on a feature for new vehicles and the company failed in part because buyers simply didn’t know those features were part of the vehicle. Dealers never set it up for the buyer and it wasn’t something many people would think to do on their own
A salesperson isn’t going to jeopardize an easy sale by bungling some fancy new feature they can’t control
Salespeople are always trained on the latest features in the auto industry.
Even if they don't get in person training, all carmakers release 1 hour long videos showing the features of their cars for training purposes.
As it turned out, the feature was still pretty janky and I only tried it a few times before reverting to just regular speakerphone. And when the battery died and the head unit lost all its config, I never bothered figuring out how to set it up a second time.
Instead of transparently selling a product for fixed price, the dealer system appears based on information asymmetry, haggling, upcharges, finance bullshit, warranty bullshit, subscription bullshit, and many decades of entrenched psyops culture against customers.
On top of that, salespeople are often poorly trained on the products and dealerships seem to have an adversarial relationship with corporate, especially around the corporate website differing from the local story.
And then the dealerships steer you for whatever benefits them. In 2017 I tried to test drive a Chevy Bolt: Motortrend's car of that year. One dealer hadn't even heard of it. Another said he couldn't get one. Another tried to dissuade me from looking at it by dissing the product. Finally I found a knowledgeable dealer that knew the product, had some, and revealed they were easy for all dealers to get in our area: the others were just being obstreperous. Suck.
I don't like being ripped off but I'm not particularly price sensitive, so I'd love to just show up and pay a reasonable sticker price that was just The Price Everyone Paid and not have this vague haggling system expected, and not have to worry whether I was being ripped because I wanted a particular option. I hate having to think about decisions I don't particularly care about, so I would prefer to constrain my choices to "maker Y's automobiles", be it GM, Ford or Toyota for the rest of my life, so that once I have the capabilities I need out of an automobile there's only one or two choices. I would love the simplicity of the dealer being the default maintainer of my automobile and I just show up once a year or so and they take care of it in a pleasant experience.
It's irksome instead that it feels like they just want to trick you out of as much money as possible up front and that they don't want to have a maintenance department, but are legally required to.
I would walk away if I saw that.
And that is probably why they go out of their way to not show you that.
They really don’t know how to develop a compelling product, even though they have an enormous amount of experience and engineering talent.
It’s very similar to the failure of Boeing but GM doesn’t make anything that can kill 200 people at once so we don’t notice.
Global-wise, there are also a few Chinese companies in the robotaxi market. Pony.ai and WeRide both recently went in public market.
Not yet with customers, but in the testing phase.
BlueCruise vehicles were updated with the most sophisticated driver attention system ive ever seen. It will complain in less time than it takes me to change a radio station.
Before that update, I was enjoying the automated driving a little bit too much, and using it to take liberties with my attention in the car. That was awesome as a user experience.
After the update, I am paying more attention to the road than I did when I was driving a manual. There is no benefit to being able to take my hands off the steering wheel. That’s a good thing for safety, but it means that “level 2.5” driving (or whatever) adds zero marginal utility.
> Starting with 2024 model-year vehicles, we will deduct points if an [active driving assistance] system doesn’t have adequate [direct driver monitoring systems]. Right now, only Ford and GM’s systems meet our criteria for earning additional points, but others could be available soon. ...
> Ford’s BlueCruise sets a high standard among [active driving assistance] systems, aided by an infrared camera that monitors the driver’s eyes to determine whether they are looking at the road. If the driver glances away from the road for more than about 5 second... the system will give the driver a visual warning and an audible chime.
It depends how you look at it, but in the (super limited) cases where you can use Mercedes Drive Pilot you can legally and safely read a book, watch a movie, or work while in the driver's seat. [1] That's not the case with any Tesla product.
Sure you might not be able to legally claim that a Tesla is totally "autonomous," but the fact of the matter (to me, at least) is that they are putting in way more effort to solve the problem than legacy auto manufacturers are.
I can step into a Tesla today, press a destination, and go there without touching the wheel or pedals. Sure it won't be flawless but the fact is, I can. I can't do the same in any other consumer car, and the closest thing is a Waymo. The effort is there, I think its just a matter of time before we start seeing the legal stuff play out.
Tesla autopilot 1.0 has been around on tesla since 2015. It's actually pretty good on the freeway as traffic aware cruise control + autosteer. I think it's a pretty good balance of driver + assist.
https://www.youtube.com/watch?v=Hv9HtWUf27s
Arguably this is Waymo in its "home turf"
We know how big the taxi market is and it's growth rate. There is clearly room for a few businesses here alone. Then consider driverless will go beyond taxi to general transportation like trucking which is massive market. Also likely play a significant variable in what cars consumers choose.
I think the risk here is software tends to a winner (or small number of winners) gets all market.
That has to be a major risk/reward concern on the companies investing in this tech.
Before the iPhone came out, some pretty decent/popular smartphones existed. I had clients in the industry. They sold a lot of n95s and such.
But... data plans sucked. Apple did an exclusive deal and forced them to include 2gb of data. They could not sell phones with data-less plans. Customers could not buy iphones on data-less plans.
At some point you have to nudge the paradigm.
I've had one Tesla for 6 years, another for 2: I've gotten to a point where I find Autopilot boring and turn it off just for something to do.
My older Tesla has enhanced Autopilot. As impressive as it is, it's very glitchy. I primarily use it so I don't speed.
Because enhanced autopilot was so glitchy, when we added a 2nd Tesla to the household, we didn't pay for FSD. (Enhanced autopilot was too expensive given how glitchy it was in my older car.) The free trials of FSD in the 2nd Tesla are impressive... And glitchy. I constantly need to take over on surface roads; but it is very nice on freeways.
The thing is, "self driving" as a feature just isn't worth the sticker price. If you're sitting in the driver's seat, and there's nothing to do, you can just save yourself a boatload of cash and drive the car yourself. Especially if you have to remain alert at all times, the best way to cut the boredom is to drive the car.
But if I have to maintain a hand on the steering wheel and monitor the car I may as well be driving.
Today, I think they should do this. It is rare I need to touch the wheel or pedal these days and people should experience it.
While SuperCruise is a highway only system, riding in a friend's car with it was pretty much a non-event.
Self driving probably works great on Interstates with numbers like 5, 10, and the feeders. Based on my experiences with subaru self driving and watching videos and watching tesla drivers around here, a self driving car would give up after maybe 3 minutes. There were more teslas and other expensive "self-driving" style cars a couple years ago, but now i rarely see them. I wonder why?
I can consistently get my wife's subaru self driving to swerve into another lane without warning, without jerking the wheel, with very minor control inputs. Subaru keeps updating the firmware, but they haven't fixed the suicide merge!
> Chrysler was the first manufacturer to implement the device in 1958. They called it “Auto-Pilot” and it appeared in their luxury model as an upgraded option. Soon after, General Motors installed it in their Cadillac vehicles, naming it “Cruise Control” which has stuck to this day.
That was a your-dealership issue, not a GM issue. GM doesn't run the dealerships (though they sure have plenty of influence).
I test drove the Silverado EV a month ago and got to try out super cruise.
The top-selling EV maker in the US doesn't have any dealerships, so clearly it's not an insurmountable problem. GM just hasn't prioritized surmounting it.
We are seeing this play out with VW’s attempt to direct-sell Scout vehicle's and immediately getting challenged/sued by dealerships over it. [1]
https://money.usnews.com/investing/news/articles/2024-10-25/...
This is just a car dealership tying its hands behind its back and not competing well. If it isn't widespread it isn't a GM problem.
vehicles regularly catching on fire for no reason
vehicles driving into the back of emergency vehicles, slamming on the brakes in the middle of highways - both because the CEO decreed that RADAR/LIDAR isn't necessary
Issues with its "self driving" randomly swerving at objects, pedesrtians, cyclists, other vehicles...
Enormous service/parts/bodywork backlogs
vehicles blacklisted by insurance companies becaue of high crash rates and repair costs
windows that randomly shatter
drivetrains that fail because you drive them in too heavy a rainstorm
a truck that cannot be driven in the snow (snow reflects light off the DRL bar, blinding the driver, the headlights getting blocked by snow, and the vehicle is stymied by barely a few inches of snow):
https://www.tiktok.com/@molesrcool/video/7446853436198358303...
GM doesn't have any of these problems (except for a small number of Bolts which had defective cells made by LG - all replaced now, and given extended ten year warranties.)
https://www.cnn.com/2024/07/18/cars/china-baidu-apollo-go-ro...
That doesn't sound like it's nearly at par.
I'd love to see some direct safety comparison stats.
It does seem strange but when one considers that the CEO of one of these competitors spent ~$260M to help elect the next President and he will have his ear on matters of autonomous car regulation, maybe not so strange.
If future autonomous car regulations are influenced by Musk and end up being lax and favour the technological approach that Tesla has taken, then GM/Cruise may have made significant irrelevant investments to solve issues that are no longer relevant, and this puts them on poor footing in the new competitive landscape.
It's possible they're bowing out early on the assumption that this will no longer be a competition that they can win.
To be clear, the "no longer relevant issue" is safety regulations designed to protect citizens from getting killed.
Yes. I think GM leadership has finally learned enough to realize how far behind they are.
I can safely and legally take a Waymo home after getting drunk at the bar, this is not true of a Tesla.
Btw that kinda thing is one of the few use cases that really does require 5g’s latency and bandwidth… which Tesla currently doesn’t ship. More hardware updates are going to come, despite musks promises)
On top of that, one offered free oil changes for 12 months, IIRC. They'd regularly try to sell unneeded services on top of it (alignment after <12 months of ownership without even checking it first, for example).
My experiences with Tesla have been imperfect, but vastly better by comparison. There are issues, but I'd take them over the dealer experiences 100% of the time.
A Subaru dealership tried to charge $200 to reset a passenger window like this (literally just pushing the button to roll down the window and back up):
I don't know how I will switch to another car brand after this.
I can’t imagine how someone would compare it to a multi-sensor setup - a single camera can’t match the level of prescience a Waymo or Cruise have. But also as the overall experience is just absurdly different. Waymo was the first tech thing I felt was truly magical in _decades_.
Meanwhile the best I can get out of Tesla FSD is straight highways. And even there, there’s random glitches (the forever annoying ghost braking being the one that makes me feel particularly unsafe with it)
Tesla's equivalent system appears to be 100% vaporware. I don't understand the case for them not being far behind.
My gut tells me Waymo is ahead, probably by a good amount. But calling Telsa FSD "vaporware" is absurd, and leads me to the same conclusion that GP already called out.
Seems like a) the salesperson might not know his vehicles and perhaps it just wasn't a Super Cruise enabled vehicle (meaning, it wasn't added-on to that specific vehicle). b) the actual Super-Cruise module was broken in that vehicle, and that was his excuse c) he didn't know how or wasn't comfortable with telling you how to enable it. (some people are still quite afraid of giving full control over to computers)
I have tested numerous vehicles with Super Cruise; I quite enjoyed it.
I could be wrong, but unless it's a new GM rule, afaik, and with my understanding of how dealerships work (as far as the ones I've visited), dealerships try to avoid messing with the vehicle programming as much as they can. It's rather sad, as aside from the manufactures, dealerships have the most tools and access to the vehicles programming, but rarely-if-ever stray from the service manual.
Custom programming kind of is a market for that reason.
Living in a state with a very large automotive industry, there are a lot of people who have the tools + knowledge that offer custom programming services like: Programming the key fob that if one was to hold the unlock-door button for 5 seconds, the driver side window opens. And if the lock button is pressed once, then held for 5 seconds, all the windows are rolled up. Enabling a bunch of different video codecs/formats and enabling video-playback while the car is in motion if the passenger seat detects something siting there, etc, etc...
I wonder if this is a liability thing (driver would be responsible).
Or it could suck and the chances of buying the car go down.
I remember walking into an apple store and seeing the apple vision pro headset. I asked to try it, and they said "you can't. You have to make an appointment and blah blah". If someone adds friction to the buying process, there must be some reason - they must be hiding something (poor performance, complexity, price, etc)
GM later had a PR release for UltraCruise which was supposedly developed in house and maybe that uses Cruise technology. However, AFAIK it was never released and is rumored to be shelved.
The facts don't matter anymore.
Just perception.
I think you mean interest in level 3 autonomous driving appears to be cooling across the board
>Google/Waymo ... Amazon/Zoox ... Uber/Yandex ... GM Super Cruise ... Ford BluesClues
That's the one the press release is referring to, I guarantee it.
Other companies have neat AV demos (some that they even make people pay for), but fundamentally they can't and won't scale.
GM was majority stakeholder in Cruise, which was gearing up to a ride hailing service. Cruise's cars have instruments like LiDAR on the roof, unlike Super Cruise-equipped cars.
Ford wrote about an "initial launch of a commercial self-driving service in 2021" but I don't think it materialized. A photo of their robotaxi bears the logo of Argo AI, a startup that shut down in 2022.
I just spit up my drink. Thank you for that.
Waymo isn't scalable.
"It seems like drivers did not find the feature compelling and were not renewing. Interest in autonomous driving appears to be cooling across the board."
No, because BlueCruise sounds like an asinine feature. To make me pay $2450 more for a car for some arbitrary selection roads for that feature to be available doesn't make sense to me.
Cruise really was on to something, well w/e GM's loss!
My pet theory is that it really took off because a bunch of investors were hoping to capture and capitalize on the baby boomers aging out of being able to easily drive themselves. They tried to speed up technological advancements in order to meet that window of a large more-affluent cohort of potential customers.
So it might be interesting to somehow plot total "self driving" investment against that kind of customer base projection over time.
That's still a huge market. And the baby boomers are the most affluent demographic in the western world.
Many others on forums report the same, and here's an example of a component being borked.
AEB, Automatic Emergency Braking, is on by default and always turns back on at vehicle start.
It's dangerous. An example? Even with a camera and a separate radar, it is borked. I've been behind another car, slowing for a light, and hit a downward slope of a large speedbump and SLAM the brakes come on.
Why?
The radar suddenly sees pavement, as the nose of the car is pointed down, but the camera sees car + brake light ahead.
This is at 30km/hr, and everyone is braking fine, the car ahead is not even close. And proof it's not me is the 5 times it happened in 2 weeks, it was always something like this.
I was almost hit 3 times.
It literally panic brakes. Horribly designed.
I finally had it and removed the radar sensor with forscan.
With behaviour like this in something as simple as AEB, by no means would I trust anytiong else. Forums show people complaining about:
* adaptive cruise slowing down, or speeding up because it sees speed limit signs on offramps, or even adjacent roads
* the lane assist pulling strongly because the car thinks an offramp is the main road.
* reports of swerving into other lanes when passing some vehicles, the car confused
It's just junk.
This beta testing on our roads needs to stop.
What cracks me up is this.
If you asked a skilled software dev, if they'd prefer to have software designed by a team, any team, responsible for tgeir life? They'd likely say no.
Because we've all seen the crappy code that gets produced. We know of bugs. We know of failures in testing, automated or not.
These cars ae all extremely buggy, just as any software project is. But beyond that, auto manufacturers are notorious for cutting corners.
I just can't imagine why anyone, this decade, would want to trust auto drive anything.
https://news.ycombinator.com/item?id=33554679
Lidar obstacle detection algorithm from a Git repo leaked onto Tor
This is a drivable region mapping (obstacle detection) algorithm found in what appears to be a git repo leaked from an autonomous vehicle company in 2017. The repo was available through one or more Tor hidden services for several years.
The lidar code appears to be written for the Velodyne HDL-32E. It operates in a series of stages, each stage refining the output of the previous stage. This algorithm is in the second stage. It is the primary obstacle detection method, with the other methods making only small improvements.
The leaked code uses a column-major matrix of points and it explicitly handles NaNs (the no-return points). We've rewritten it to use a much more cache-efficient row-major matrix layout and a conditional that will ignore the NaN points without explicit testing.
This is an amazingly effective method of obstacle detection, considering its simplicity.
Fun fact - Cruise did actually have a source code leak a few years back, some engineer decided to embrace the 4 hour work week and outsource their role to a contractor, who uploaded their work to a public github repo (I assume this was accidental and they intended to make a private repo). But it was not core AV code.
It may not be a perfect driving solution (probably 1:1 teleoperator to car in the beginning), but their go to market speed is undeniably better than Waymo et al. Waymo, Cruise, etc have to map out an entire region and have long rollout timelines. In theory Tesla could say they will deploy in X city, talk with the local government to get necessary approvals, and get it going after a few test drives. Thats the tail, anyway. With that kind of go to market speed you can start acquiring customers pretty easily and start scaling up the business. Its basically the typical SaaS setup - cut some corners for market share and become the defacto player in it after some time.
I wouldn't write them out purely out of spite. Their solution is the most conducive to scaling laws, on the ML, operations, and business side.
Meanwhile Waymo bets on jamming the car full of sensors to ensure it has superhuman 360 perception in all conditions with multiple sensor redundancy. Waymo takes safety insanely seriously. Tesla doesn’t care if 100s die in their car while on FSD. Waymo doesn’t want to make cars or manage fleets. They want to purely focus on automating the driver end to end.
Unless Tesla has a magical perception algorithm breakthrough, Waymo may expand all US states that allow it, while Tesla won’t even get an operational robotaxi with 100+ rides per year.
Elon has tremendous success in SpaceX with falcon & starlink, that’s a fact. But Tesla FSD is so much hype compared to substance.
My understanding was that in many states, it's mostly just paper requirements and not any strict testing, etc.
That's a good thing, we shouldn't be lowering the bar for AVs.
The report comes days after Trump named the automaker's CEO, Elon Musk, as a co-head of the incoming administration's government efficiency department. Trump's team is looking for policy leaders for the transport department to develop a federal regulatory framework, the report said, citing people familiar with the matter.
Last month, Musk criticized the state-by-state approval process, required for self-driving vehicles, as "incredibly painful", weeks after unveiling a two-seat "Cybercab" robotaxi without a steering wheel and foot pedals, set to go into production in 2026.
a) Early technology adopters are more forgiving of minor problems and are your free word-of-mouth marketers. The company benefits from lowered GTM costs.
b) Fast, reliable service depends on density; there needs to be enough cars on the road that there's always one available within a few minutes when you open the app; users will develop a habit of choosing whichever provider is reliably faster and that habit will become ingrained, making market entry for later movers more costly. First movers also have the advantage of their early adopters helping pay for their fleet as it's built up; later movers have to flood new markets with subsidized capacity to get to equivalent density. That works (Uber did it for rideshare) but it's expensive. So again: lower GTM costs.
Regarding whether this will be a viable market - Lyft/Uber have already demonstrated there's tons of demand for on-demand transportation services, and Waymo in SF is already very popular. One can back-of-the-envelope whether the economics work but generally speaking any time you replace labor with automation in a large market, it does.
In the long run though you're right that there isn't a moat that can't be solved with massive capital investment. But you could say the same about many markets; turns out one can have a successful company within a competitive market the old fashioned way.. by just being really good. It's harder, but it works.
Baidu/Apollo, maybe. There are deployed robo-taxis in China. Reviews of driving quality indicate "meh", they are sometimes directly teleoperated to get out of problems, and difficult intersections have fixed cameras the cars can see through.
Of course, if they're actually a more general kind of sensor but called a camera then it's fine.
I've been very puzzled at people treating FSD as non-existent when it's very reliably good.
It's very similar to legacy space and internet, the existing players saw SpaceX coming years out, did nothing and got crushed.
And Waymo has that whole time to bring down their costs of construction, learn how to solve problems of scale, and so on.
I agree that a CyberCab would have structural advantages if the service were at parity w/ Waymo, but that is a very big hypothetical!
If they do pull that off, or are able pivot to a fused sensor architecture then I do think their hardware potential will make them a good competitor.
Beyond, idk, saving my relatives the task of driving me to the airport, I don't see my family using an autonomous taxi at all. I think I'm like many Americans in that I like my personal vehicle because it's mine. I'm comfy in it, it's filled with only my stank, and I have tailored it to my liking.
Wonder if they'd sell Cruise back to the founders for say a buck? Sure Cruise could then cut a deal with another auto maker. Maybe even get financing from one?
https://www.cnbc.com/2024/12/10/gm-halts-funding-of-robotaxi...
Dec 4th, 2024
GM braces for a $5 billion hit as it fights to keep up in China’s intensifying EV price war
SAIC-GM revealed in a regulatory filing on Wednesday (via The New York Times) that it expects to write down between $2.6 billion and $2.9 billion in the fourth quarter. The automaker is also expecting another $2.7 billion in restructuring expenses.
https://electrek.co/2024/12/04/gm-faces-5-billion-hit-ev-bat...
Now they've gone back to what I think GM's original plan was for acquiring Cruise — using their tech for passenger cars. Selling cars is GM's bread and butter. This wasn't going to end any other way.
Then they didn't deliver.
Waymo hasn't left blood on the pavement. Cruise and Tesla have.
As long as we're clutching our pearls and being overdramatic, how much blood have the humans left on the pavement in the meantime?
I can't speak to their technology, but generally speaking Cruise seemed too cautious, both in the aggressiveness of the driving, and in how quickly the company capitulated. I've taken Waymo a few times and what stands out to me is how aggressive Waymo cars are. As a passenger this aggressiveness makes them feel more sophisticated, not to mention faster, but the cars also spook pedestrians to a greater extent as they nose their way into traffic or through turns. Nonetheless, Waymo has been more adept at moving past bad publicity than Cruise, including in the face of accidents.
Above and beyond whatever technological shortcomings there may have been[1], I think there was a management and business development mismatch between GM and Cruise. Cruise manifestly worked well enough. With better management--more honest, but also more aggressive--it could have been a contender; heck, it was (is?) the only proven competitor to Waymo for autonomous public taxi service, at least in the U.S.
[1] In 3 Waymo trips remote intervention was required once, which is a comparable ratio to my Cruise trips.
I’m serious about this point too. Eg. Intel pays less than half of what tsmc in Taiwan pays for similar jobs on a ppp basis after 20years of wage freezes and cuts at intel. They can’t poach any talent and investors want more layoffs and wage freezes next quarter to fix the decline they see. GM is similar. It’s not going to turn around and start competing in the high end labor market reasonably. It can’t. The big shareholders would sack the ceo in seconds. Think GM could ever compete with Google (Waymo) in a market where ai engineers can earn ) 1m/yr? Ha! The MBAs wouldn’t allow it for a second.
For example: https://news.ycombinator.com/threads?id=reTensor
It was well-known and well-leaked that something was wrong with Cruise.
As soon as Cruise closed up shop, the complaints went way down.
https://sfstandard.com/2023/09/12/muni-bus-killed-san-franci...
My thoughts today are the same as they were back then. Self driving will only be solved city by city and street by street. Local governments will need to create dedicated closed-off autonomous lanes with sensors. Auto manufacturers will need to abide by a common standard to talk to local infrastructure and to each other.
Instead we chose to solve the problem in the most complex and inefficient way possible, and are now seeing the results.
The answer is that people want cars. They want to drive. They want to be taken directly from point A to point B without needing to walk to transit stops and switch buses/trains and face unpredictable schedules and delays.
Autonomous road infrastructure is a sensible, incremental improvement. And in fact city buses will tend to benefit from it the most.
> The answer is that people want cars.
Actually in many cities in the world, only a minor % of people own a car. And even those that do own a car don't use if for that many trips.
> They want to drive.
Driving in cities is almost universally hated and is proven to cause lots of stress.
> and face unpredictable schedules and delays.
Because in cars its impossible to have unpredictable delays, roads are always completely open and there is never any concention?
In most cities in the world, public transport has less unpredictable delays compared to driving.
> Autonomous road infrastructure is a sensible, incremental improvement. And in fact city buses will tend to benefit from it the most.
A sensible incredmental improvment are regular bus lanes with regular trolley buses.
EDIT: and yes, it's a completely reasonable question to ask why invest in special new kinds of roads when more people would be better served by public transit.
Even if the claim (you need Europe density) wasn't trivially false, several parts of the US DO meet that level of density and could easily support good public infrastructure.
Or do you not know that in the early 1900s nearly every city had copious public transportation that was relied on by literally everyone in the city?
Seems pretty reasonable for difficult new tech.
So I must be missing a lot of cars and a lot of extra cities with service. Or you've confused ambitions (Cruise wanted to have thousands of vehicles and wanted to be in dozens of US cities by 2027) with reality (Neither of those things happened)
https://deepmind.google/discover/blog/genie-2-a-large-scale-...
That being said it’s usual in US for CEOs to be rewarded in short time with stock raises when they cut costs.
Innovation is not a game for the light hearted.
Google is playing that game with Waymo and ploughing billions every year. GM doesn’t have the same risk bearing culture.
Where is the money? I don't mean to be as snarky as Cruise founder, but how can you burn that amount of money and not show results?
They operated a robo taxi fleet for a time (until they got involved in an accident and mislead the following investigation), what other results would you like to see?
QUOTE: Cruise and GM’s technical teams will be combined into a single effort focused on developing autonomous technology to offer in future models sold by GM, according to a statement Tuesday. GM said it will no longer fund robotaxi development work “given the considerable time and resources that would be needed to scale the business, along with an increasingly competitive robotaxi market.”
It’s a big retrench for GM and Cruise, which survived a shakeout among autonomous-driving companies and restarted operations after one of its cars dragged a pedestrian last year. ...
The move has significant implications for GM. Chief Executive Officer Mary Barra wanted to transform the automaker into a transportation technology company and double GM’s revenue by 2030 in part by generating $50 billion from Cruise. Without a robotaxi business to bring in fares, that goal looks remote. /QUOTE
The business of robotaxis has always been a bit suspect, given the unit economics of running 1000s of taxis, each with ~$100K of equipment (even as the COGS were coming down each generation).
My 2c: the only reason Waymo is succeeding (so far) is that its parent Alphabet has deeep pockets, it has a a 6th gen (?) technology of its AV driver, and they already exited the other AV trucking segment... implying they need to succeed in robotaxis, so it's all hands on deck. It also helps that they have, so far, NOT blundered their way (=> a significant pedestrian collision) the way Uber ATG did or Cruise did.
This is a loong game to win, far longer than anything the tech bros have promised their their investors... We are still in the initial innings
I wouldn't underestimate the advantage of not having to bootstrap a planet scale backend/ML platform. Look how much R&D Uber spends on software and they're not even doing the same level of ML. Waymo can just start using excess capacity across the google fleet whenever it wants for training and simulation. That's a huge advantage.
It felt like there was a lot of buzz around it a while back but then it only happened in a couple of cities which were just US cities with favorable climate. Did it ever move beyond that? You'd think that if the market is really competitive, then operators would bring it to cities where it doesn't exist, rather than just all competing in the same small handful of cities?
Their last great innovation was Saturn, and they killed that too.
Waymo has been a little less expensive than Uber/Lyft for me. I just saw a Zoox autonomous vehicle today. That weird looking box with no drivers seat.
If Tesla can do level 5 in 2-3 years as you say (and that might be a pretty big “if”), that places them 5+ years behind Waymo.
What leads to the win here, then? Waymo constrained by the cost of LIDAR? Is it truly such a massive % of build cost that they can’t succeed? Is it that Tesla is vertically integrated?
https://blog.comma.ai/a-100x-investment-part-1/ https://blog.comma.ai/a-100x-investment-part-2/
Cruise cars were way more numerous around SF, although the service was worse than Waymo's. That stupid October 2023 accident really snowballed, and it wasn't even their (primary) fault.
where is YC's top tech success now?
Shows that most yc companies are clownshoes who pale in comparison to the tech chops of actual tech companies
Why not just keep the meter running? Then it ceases to be a problem.
Lots of people are ignoring this because of politics.
The Elon hate is so ingrained here, people are ignoring what’s happening right in front of their eyes.
It would, inevitably, lead to more cars in cities. Probably faster cars. Both are the opposite of what we need; it isn't just about safety.
People consuming less energy to get to their destination [2] is good for society
[1] https://engineering.nyu.edu/news/subway-air-pollution-dispro...
I say "we" do - ie, if people want to buy autonomous cars, then that's what's going to happen.
Incidentally if there's a large number of autonomous cars available in a fluid and efficient market, that would probably mean fewer people need cars in total.
Baked in to this is the assumption that the operation of markets are democratic, but they are not. A very small number of people decide where investments go, what tech to design, and what to buy. The alternative - public transit focused infrastructure, is not available to buy in small quantities. Huge investments have to be made before anyone can “vote with their wallet” on public transit. And similarly, billions have to go in to AV development before anyone can buy in to it. And those with more money always have more “vote” in a “vote with your wallet” scenario. Fifty percent of the people in the US have so little money as to make their votes nearly meaningless in such a scenario. That’s not “us” making a decision that is very much certain people making that decision.