Cruise was supposed to generate $50 billion in revenues by 2030, according to GM’s former hype, which Wall Street ate up.
By Wolf Richter for WOLF STREET.
General Motors announced Tuesday afternoon, after blowing $10 billion on its misbegotten robotaxi-startup money-pit Cruise LLC which it had acquired for an undisclosed amount in 2016 amid enormous hype, that it would:
- No longer fund Cruise;
- Bring its ownership from the current 90% to 97% via “agreements with other shareholders”;
- Acquire the remaining shares;
- Shut down the robotaxi operation;
- Fold Cruise’s technical team into its own autonomous and assisted driving effort for personal vehicles which produces the “hands-off, eyes-on” driving feature now available on nearly two dozen GM models.
So adios, Cruise.
Until Cruise started to implode in October 2023, GM CEO Mary Barra had hyped Cruise’s financial potential at every twist and turn, expecting it to produce $1 billion in revenues by 2025 and $50 billion by 2030, which caused Wall Street to lick its chops.
Cruise costs GM currently about $2 billion a year. Today GM said in the statement that it “expects the restructuring to lower spending by more than $1 billion annually” when the restructuring is complete in the first half of 2025.
Following the announcement today, shares jumped initially by 3.6% afterhours but then gave up part of it.
Alphabet’s Waymo, among the adults in the robotaxi room, has been operating driverless robotaxis in San Francisco, Phoenix, and Los Angeles. Here in San Francisco, riding a Waymo – they’re retrofitted Jaguar SUVs – has become a must-do tourist attraction, along with riding a cable car. The Waymos are everywhere, especially in the touristy part of the City.
Amazingly for a lot of us erstwhile cynics, Waymos are working pretty well compared to the mess that human drivers are making out there. This year through October in San Francisco, human drivers have killed 20 pedestrians, 2 bicyclists, and 8 people in vehicles. Nationwide, humans cause about 40,000 traffic fatalities a year. Not to speak of the serious injuries they cause. Human drivers are terrible drivers.
Developing this autonomous driving technology is a very risky, very costly business. Even minor accidents become global clickbait. The reputational risks are huge, as GM found out, and as Uber found out when it scuttled its robotaxi development after one of its test vehicles killed someone pushing a bicycle across the street.
But the top people at Cruise, a San Francisco startup company founded in 2013, were a bunch of shoot-from-the-hip startup cowboys, living by the Silicon Valley motto, “fake it till you make it.”
And they didn’t make it. They collided with self-inflicted trouble in October 2023, just two months after Cruise had received the approval for 24/7 robotaxi operation in San Francisco. A human driver in San Francisco hit a pedestrian and knocked the woman into the path of a driverless Cruise, which ran over the woman and dragged her, trapped under the car, for more than 20 feet before pulling over. That was bad enough.
But what was worse were the shoot-from-the-hip fake-it-till-you-make-it cowboys running the place. They tried to cover up the incident by failing to report it properly, by then misreporting it, and by lying about it, which infuriated local regulators and got them into deep legal trouble with the US Department of Justice.
Following the incident, Cruise’s permit to operate in San Francisco was suspended in October 2023. Cruise then halted operations in the other cities it was operating in – Phoenix, Austin, Dallas, Houston, and Miami. In November 2023, it issued a recall for its 950 robotaxis and laid off a portion of its 4,000 employees. Among other housecleaning GM did at Cruise at the time, the founder and CEO Kyle Vogt plus nine executives were forced out.
Last month, under the fist of GM by then, Cruise admitted to the allegations by the DOJ, accepted the criminal charges brought against it by the DOJ, and settled with the DOJ.
During the call with analysts following the announcement today, Barra still tried to put lipstick on this pig and said, “Cruise was well on its way to a robotaxi business, but when you look at the fact that you’re deploying a fleet, there’s a whole operations piece of doing that.”
Enjoy reading WOLF STREET and want to support it? You can donate. I appreciate it immensely. Click on the beer and iced-tea mug to find out how:
Would you like to be notified via email when WOLF STREET publishes a new article? Sign up here.
That $10 billion could have been used for so much better purposes at GM…
Like even more share buybacks to pump stock options for executives.
Beat me to it.
This is an excellent counter-example for the no-buybacks Jihadists (although I agree that the scale of aggregate buybacks has become more and more troubling over 40 years).
There is a long history of insiders making crap “investment” decisions (self-serving too…just in different ways than buybacks) when retained earnings get too large.
So returning money to shareholders ain’t per se awful.
In the end, it all comes down to the quality of the CEO and the integrity of the board.
If the Board is compromised or brain dead…
And the CEO is a dummy making vanity investments…shareholders are screwed,
Or if the CEO is a weasel using doomed buybacks to prop up an illusion of organizational health…the shareholders are screwed.
It is hard work to avoid both outcomes.
There’s a substantial chance that 2banana was being sarcastic.
Sarcastic or not, a share buyback would have been a better use of money…
kramartini
A share buyback would have been a total waste, they did share buybacks, many billions of dollars worth, and they’re still doing them, and they have nothing to show for them, nada, money gone, total waste. Despite the hughest stock market rally ever over the past 14 years, GM’s new shares since the IPO in 2010 have gone up only 30%. The S&P 500 has gone up over 10x as much, 350%.
With Cruise, they built and advanced the autonomous-driving technology and went through the learning process and they now have the IP and the engineers and they’re using the autonomous-driving tech in their personal cars. GM isn’t giving up on autonomous driving tech for its personal vehicles. It’s giving up on robotaxis.
As I said, the scale of buybacks has long become troubling – especially since the idea of a scumbag CEO literally loading up with corporate debt to finance doomed buybacks to fraudulently game his options (and then fly to non-extradition country)…is no longer inconceivable.
(See also, politicians and facially ruinous fiscal policies, multiple decades.)
But that said, there is more than one way for a scumbag CEO to rip-off/ruin a company.
Before buybacks were oceanic, facially idiotic “investments” were not uncommon (with indirect CEO financial interests in said idiotic investments later popping up. That’s how fiduciary requirements, no affiliated interests, and conglomerate disfavor got their start in earlier eras).
Like I said, it is hard work policing the perfidy of insiders (corporate and political) and not usually amenable to simple blanket solutions.
Because the insider scumbags, like the demon girl in The Ring…”never sleep”.
At least buybacks have the relative advantage of being somewhat transparent and splitting the loot(ing) with shareholders.
Not wonderful and ruinous to long-term corporate survival…but so are self-serving corporate “investment” delusions egged on by equally self-serving political and media interests.
But a share buyback would have avoided the waste of $10 billion. Investors who sold GM would have the money to invest in the higher yielding Investments you cite.
And those who didn’t sell would not have to share future profits with those who would.have cashed out. Investors as a whole would have been much better off if $10 billion had not been wasted on the Cruise fiasco…
“share buyback would have been a better use of money” … only in hindsight, after the capital investment failed to produce needed results.
But that’s the nature of enterprise: you invest capital in a risky venture hoping that it produces profit, but knowing it may result in loss.
This is such a narrow take.
Never take a risk, never make a bet. I rather a business try to do something and fail at it. Hype is part of it. Buybacks are the worst.
Case in point:
https://wolfstreet.com/2024/10/28/boeing-launches-22-billion-share-offering-to-get-some-breathing-room-dodge-junk-credit-rating-after-having-wasted-64-billion-on-share-buybacks/
Companies buyback because we’ve allowed share value to be the sole indicator of a company success. Hence the CEOs have to trumpet and build hype so they can manage the breathing room to burn billions.
Come up different incentives and people will behave accordingly.
The idea of autonomous driving is always going to succeed with a 99% failure rate.
If this was WeWork, I’m all for not funding crap ideas like that. There’s no net value by _democratizing_ commercial renting space. It did the opposite, and created an artificial shortage. It wasn’t like AirBnb or Uber before it, where the supply actually expanded.
I would be very disappointed if Cas127 turned out to be in IT.
Shareholders indeed view share value as the sole indicator of company success. Same as it ever was.
Rather than spend the $10 billion on buybacks, they should have just given it all to me. I would have kept $100 million or so, just for modest living expenses, and then invested the other $9.9 billion.
Or perhaps give $1 billion collectively to Wolf and subscribers at the “beer mug” level (again for modest living expenses), and invested the other $9 billion. Away from GM’s bureaucracy, we could have made a go of it.
“I would be very disappointed if Cas127 turned out to be in IT.”
I have worked in IT in the past.
Look, I’m all in favor of well-reasoned *investments*.
But decades of ZIRP led to the $200 million “science experiment” – VCs and startups so blindly/absurdly flush with cash that actual pre-investment due diligence on things like feasibility, product moat, and actual consumer demand were airily waved away (remember, a shit-ton of this money is actually coming from public sector workers unions’ pensions…who if they and their “advisors” get lazy/f-up stick the pension shortfalls on…the taxpayers. Again).
These people *aggressively* learned *nothing* from the 1998-2002 internet bust. Because it would have stood in the way of them getting their “investment management” cut.
So what should have been a doomed $20 million investment gets turned into a doomed $200 million investment.
And this aggressive/predatory investment sloppiness isn’t just on the front end.
However you feel about Musk, it is pretty clear that Twitter had a *ton* of people (well over 50%) who really weren’t necessary for the actual running of existing products.
But what of *future* products?
Well, I’m pretty sure Twitter really wasn’t exploding in new product innovations in the 5-7 years *before* the layoff bloodbaths.
General insider knife-fighting, endless political bitchery, and sub rosa censorship – yes.
Actual effective product advancement…not so much.
As between insider constituencies of all colors appropriating wealth that isn’t their’s and returning more money to the actual owners…isn’t a tough call.
In the end, everything requires balance.
“So returning money to shareholders ain’t per se awful.”
They should do that via dividends, not buybacks.
Why are dividends superior to buybacks, in terms of diverting money away from “worthwhile, corporate growth investments”?
In both cases, the money is exiting the corporation and being returned to the shareholders…you know…the actual owners of the corporation?
Like I said, the sums involved in buybacks have become troublingly titanic (and have been for probably 20 years) – primarily because they suggest that the market indexes could collapse without the buyback form of stock manipulation.
(That is, at the very same time that an earnings decline damages a company’s fundamentals reducing its share appeal, the company’s ability to phoney-baloney prop share price via buybacks would evaporate. So a 45 PE ratio – ahem – overnight becomes a 15 PE ratio – ahem)
But the same general stupidity could/would be operationalized via dividends.
I’m not sure why people think of the latter as superior to the former.
It isn’t like the corporation is just buying back the CEO’s shares…everybody’s shares are being bought back.
A buyback is superior to a dividend as it allows each shareholders to decide how much cash to take rather being forced to take an arbitrary sum.
A buyback is INFERIOR to a dividend because you have to sell your (appreciated) shares to get the cash.
With a dividend, you just keep collecting the income without selling the shares.
Buybacks and divs are not comparable. Buyback doesn’t guarantee any additional return of anything, as the stock can collapse later. Dividends are money, really money, today.
Yup. For most of the 20th century stock buybacks were illegal, considered a form of stock market manipulation. Which they are. It wasn’t until Ronny Reagan in 1982 that they became legal again, and we’ve been stuck this way since then, with apparently neither political party willing to go against that fat wall street cash.
Mary Bara is, in fact, a weasel. Barra recently took advantage of the price runup from a recent buy back to unload a big chunk of her GM shares, allowing her to pocket $84 million.
At the same time she’s attempting to fleece Michigan taxpayers by asking for a $250,000,000 payment from the State to redevelop private real estate owned by GM and laying off thousands of local employees.
Her history at the company is shameful.
“returning money to shareholders”
That would usually mean paying dividends in a normal world.
The funny-world version – buying back my shares with my (i.e. treasury) money, leaving me with no shares despite the fact that if I wanted money and no shares I could have sold them in the market instantly.
Wall street loves that B(lue)S(ky) rationale for buybacks; the one-day blip they can make some bucks on by selling to the patsies.
Oh, and mopping up cheap options by management. But Michael Brush went into that in detail decades ago … to no avail.
And for the umteenth time: it used to be illegal – for a reason.
I think some of you are confusing buybacks with tender offers. You, as a shareholder, don’t do anything, you do not sell, when a company buys back shares. You simply have a larger % of ownership of the company.
I am sure they will use the savings and put it towards more AI. And Wall Street will lap it up.
Wolf, please pardon the ignorance but if they already have 90% ownership, why bother with the 7%? Can’t they just walk away without bothering?
GM could try to sell Cruise. But they wouldn’t get much for it. Conversely, GM is probably not paying a whole lot to buy the remaining shares. Cruise probably has some tech that GM wants to integrate into its own efforts and some engineers that it wants to onboard. Those two could be valuable for GM.
I think it’s the IP and core engineering talent as well that already has the experience they need.
As per the article (and all other news reports), GM is absorbing the staff & technology into its own autonomous vehicle efforts.
Autonomous driving will shortly be a required feature for any automaker. Despite the business failure of the robotaxi effort, that advanced tech & organized staff is still a valuable asset.
I have work for and around these large automotive companies for over two decades. The way they throw money away on these “going nowhere projects” is sickening. I can only think of two reasons why. One, they are ignorant to the point of criminality. Two, they are making thier friends rich. You can choose which one is the truth…
Self-driving isn’t “going nowhere”. It will become the dominant vehicle control in the not-to-distant future (within, my guess, a decade).
In area with difficult climate (like lots of snow), it may take a bit longer but will still happen.
Projects at big companies are like start-up venture: of 10, some will fail outright, a few will break-even, a couple will be profitable, and one will be outrageously successful and will make huge profits overall.
The problem is that you can’t tell the 1 from the other 9 at the start.
Both.
Full autonomous driving now has proof-of-concept operating in 3 cities. This is not a “going nowhere” project. This is now an “inevitably offer it with our vehicles or die” project.
They could of used the 10 Billion to pay down on the 200 + Billion Debt they owe. Another thought would have been to use the money to Fix the Transmission and Electrical Problem on the Profitable Pickup Trucks. No Mary instead through it all away ; she still remains as one of General Motors largest shareholders.
Why not focus on automated interstate trucking. Automate travel on interstate highways, then switch to humans in urban areas. Perfect this then work on the next level. At least they would generate some return while they continued research. A truck on a limited access highway has got to be easier then urban driving. They could even fuel them up such that they have no stops between the start and end of their journey.
I suspect that GM was relying on this revenue given that EV profits were always going to be minimal.
The politics of it.
There are almost 2 million truck drivers in America-it’s one of the five most common jobs. They are the hardest people to place elsewhere and ornery as hell.
Nobody is ready to open the AI can of worms by putting these guys out of work and back on the Kentucky sofa.
GM does not sell heavy trucks.
Even if it did, there are over 10x as many passenger vehicles sold in the USA every year vs. heavy trucks. i.e. 10x as many potential customers.
Cuz there would be teams of fast driving tuners in black honda civics stealing the loads of VCRs and DVD players out on the open road with no shotgun wielding driver to fend them off.
Excellent reference of…. Speed.
/s
There are companies doing that. But it doesn’t get as many clicks. And you still need someone to load it and anload your cargo.
Railroads?
Self driving vehicles are licensed/regulated at the state/local levels, which I suspect makes driverless interstate operations difficult.
What GM needs is a new CEO. Preferably a chick with cool sounding name. Preferably wearing a leather jacket (a la Nvidia). Preferably pushing AI tech; maybe self-driving cars or something.
Great piece. Today, I was walking East on Pacific Ave in SF, not far from the Wolf Den. A driverless Waymo car was stopped at a stop sign, at the bottom of a steep hill. A large truck then attempted to make a right turn, heading up
the same street that the Waymo was heading down. The driverless car elegantly backed up just enough to permit the truck to negotiate its wide turn, without hitting the car stopped just a few feet behind the Waymo! The passenger in the truck and me (the pedestrian) couldn’t help but smile at the Waymo’s brilliant maneuver and gave one another a thumbs up. Good on ya, Brilliant Team Waymo! 🙏🏼
So here is my Waymo story. I was walking by the Marriott here. A Waymo was stopped near the curb maybe 20 feet past the hotel driveway, and a woman’s voice said, “Please close the rear door, please close the rear door,” which was slightly ajar. So I go on to do my thing, and five minutes later when I come back by, the Waymo is still there and the indefatigable lady is still saying just as patiently as before, “please close the rear door…” I look inside, and there’s no one. It seems the passengers left without closing the door. So I shut the door, and a few seconds later, the Waymo (happily, I imagine) drives off.
The Waymos should have power doors that close automatically, like the rear doors of taxis in Tokyo had for decades. But these are retrofitted Jags, not taxis. I assume that purpose-built robotaxis will doors that close automatically so that they don’t have to rely on good Samaritans walking by.
Next time, the some Waymo will identify you on a street and will take you to ones needed their doors to get closed. This would be the best machine learning example lol
It definitely requires the public to understand the needs of a robot. I was leaving the park (GG) after a show and saw a Waymo stuck from the exiting crowd surrounding it. No one was trying to interfere with the Waymo as far as I could tell, there was just a lot of people being funneled to the park street. I feel that a human driver would have slowly inched out moving through the crowd. The Waymo could not get the context in which is was surrounded by humans.
Waymos as a matter of principle don’t threaten pedestrians. They will not try to inch toward people to force people to move out of the way or to hurry up, as human drivers do. And that’s good. Human drivers often roll closer to a pedestrian, instead of stopping, as he crosses the street; I guess they’re trying to pressure the pedestrian to hurry up. My wife does that too occasionally. That’s really bad, and in many places, it’s illegal.
Wolf, IME, that happens a lot when a car wants to make a right turn, and peds have left the corner well after the green hand or whatever “it’s ok to go” signal switches to a flashing “no go”. That’s jaywalking and hugely inconsiderate. The worst is when the flashing red is just about to end and some jackass leaves the curb. That costs at least 5 cars from being able to make that right turn, and multiplies congestion.
Incredibly, plenty of ignorant pedestrians think that’s a legal move. “If I can get across before the traffic light goes red, it’s legal.” Had that argument many times, usually ending in a google search and their defeat.
” Open the pod bay doors, HAL.”
It looks like Waymo is in the lead with only Tesla having a chance to catch up in the next few years. Hardcore engineering is what Google has always done the best.
So why are they trying to break up Google?
Tesla, lol. Their effort is principally a stock promotion scheme that Elon has been successfully for a decade plus. So many highlights over the years, but my favorite is when he said Teslas are appreciating assets because robotaxis were thisclose
Only a somewhat educated opinion… Tesla has not been using LiDAR (helps with object distance determination), were as everyone else has. My guess is because it is expensive. But after many years they might be coming around as it seems they purchased some.
“Waymos are working pretty well compared to the mess that human drivers are making out there. This year through October in San Francisco, human drivers have killed 20 pedestrians, 2 bicyclists, and 8 people in vehicles. Nationwide, humans cause about 40,000 traffic fatalities a year. Not to speak of the serious injuries they cause.”
I love how Wolf states totally unsubstantiated points as facts.
If you want to do a comparison between two things, please provide data for both things (and ideally in an objective manner). How many human driver miles in San Fran over what period, how many Waymo driver miles for the same, what are the results of each. If you don’t have this info, then don’t make statements of fact.
Just because you refuse to look at the data to stay within your dark hole doesn’t mean the data doesn’t exist. It’s published for all to see. You just have to want to look.
As of October 2024, Waymo performed 150,000 paid rides per week totaling over 1 million miles weekly, according to the most recent update by the CEO.
Through July, Waymo robotaxis have driven 25 million rider-only miles (no safety driver):
Waymo driverless rides compared to human benchmarks, per million miles driven:
AND: unlike human drivers, the autonomous tech is improving all the time.
“Waymo’s data is derived from crashes reported under NHTSA’s Standing General Order (SGO) and uses the same criteria as described in Kusano et al. (2024).
“We are intentionally using publicly available data to allow other researchers to replicate the results. To link the data shown on this dashboard to NHTSA’s published SGO data, researchers can download a list of SGO report IDs and boolean membership in each outcome group in the download section below.”
More details here: https://waymo.com/safety/impact/
Thanks for that link. The results are much more impressive than I realized or expected.
When the country was first started by the founding Slave Masters and their northern enablers, it had the purest form of capitalism; i.e., Slavery. The AI movement is simply the interest urge to find slaves, this time electronic ones. The urge for Capitalists is so overwhelmingly strong to have slaves that it blinds the wannabe slave master of the true cost and practicality.
@Gary do you have a washing machine (an electronic slave) or bring your laundry to someone that you pay to wash it? We will eventually have almost self-driving long haul trucks, but the roll out will have to be slow so we don’t put all the people that drive trucks out of work at once just like the roll out to washing machines was slow so we did not put everyone out of work at once. I knew a guy who grew up in Pacific Heights who had a lady that came and washed stuff my hand in his basement in the 1950’s. My immigrant grandmother never owned a washing machine, and we still have the wood and glass “wash board” that she probably used the day before my parents had her come to live with us on the Peninsula in the 1960’s
Robot rights activist?
America just does not do cars very well. The U.S. does 3 things superbly: computers, entertainment and military hardware. It should stick to what it knows best.
🤣❤️
But they’re too expensive.
The Japanese design the best cars hands-down.
Sorry not sorry, SoCalBeachDude.
I was not happy with my Infiniti G-35 which we bought new in 2006 (designed and manufactured in Japan). Gas guzzler, and lots of things wrong with it over time. But our Ford Fusion hybrids, both bought used, have been gas-sippers and no problems at all. The first was rear-ended and totaled, and so now we’re on our second. I’d never buy another Infiniti.
I honestly forgot Nissan is Japanese. I lump them in with Hundai/Kia in terms of reliability in my mind.
Wolf, have you ever owned a Honda? Just curious.
addicted to my zero problems 2005 lexus 430 sc – 121000 of comfort – just the right amount of performance- and no issues
The Japanese and Koreans make nothing except low end junk compared to BMW and other fine European manufacturers and there is no argument about that whatsoever among automobile affectionados. Have you ever actually driven or owned a BMW? They are indeed the ultimate driving machines from Mini up through Rolls-Royce with nothing else even coming close.
I bought two new 5-series in my life, the first one with a V-8 in the US (1999), and the second one with a 2.8L inline 6 as expat in Europe (2002). They were very nice cars, no mechanical problems while I had them, which was just a few years each. The V-8 had a calibration problem with its traction control system when you stepped on the accelerator to get across an intersection: if a wheel slipped just a little, it would essentially slow to a crawl in the middle of the intersection that I wanted to get through quickly; I guess it figured I’d hit ice. I liked the 2.8L better for that reason, less power, but enough power, better weight distribution and no problem getting across intersections quickly. Inline six engines are a marvel.
Dont forget moneyprinting…
Why not spend 10 billion on making a better car, instead of share buybacks and incinerating it on hype and speculation? Make a great product instead of pieces of junk!!!
Elliot Estes, CEO of GM at the time was asked by automotive journalist David E. Davis of Road & Track, back in 1969 whether GM could make a car as great as Mercedes-Benz. Elliot’s response to David was ‘yes we can, but why should we?’
Used cars pushing up overall CPI just as Wolf predicted.
and wall street is celebrating such a great cpi read with tsla and bitcoin up 5%, nasdaq up 2%, and a 95% chance of a rate cut next week.
what a cancer upon society.
I don’t imagine I will ever own a driverless car at my age but assuming the significant cost of Uber or Lyft is labor I could see it being less expensive like in the early days when losing money to gain customer base was the strategy.
Has GM made any significant revenue off the purchase or has been a complete money pit?
My guess is that they don’t give that level of detail?