Kyle Vogt resigns from Cruise
techcrunch.com
techcrunch.com
Kyle is the smartest, most technically capable, driven person I have ever had the opportunity to work for. He had a deep technical understanding of the system, a great intuition for unnecessary complexity, and a constant desire to find and resolve engineering bottlenecks. And he was deeply motivated by a goal of safety and saving lives.
After the GM acquisition, he could have enjoyed an early retirement. GM were eager to give us all Windows laptops and turn us into a San Francisco outpost, just as they had done with their Sidecar acquisition. But instead, Kyle worked tirelessly to ensure we had sufficient freedom and independence to build.
I know this wasn't everyone's experience with Kyle. I know plenty of people he pushed too hard and who left much earlier than I did. But for every person who burned out, there were others motivated by his energy and loved who working for him.
Is Cruise over? Probably not. There are still a ton of extremely smart people working there, and they are still in possession of a world-leading AV technology (probably second only to Waymo, who had a 5 year head start). But I hope they can maintain their Silicon Valley culture and not succumb to becoming a complete GM outpost.
As a person that had to deal with Cruise vehicles on the streets, they were by far the worst.
Also I am happy to be your next choice after Kyle. I'll join Stripe, Patrick!
pc gave me his mobile number and answered at 3am when I was in a panic during our launch of e-commerce.
Much respect.
Whose decision was that? I'm really curious if there were internal disagreements.
1) There were ways around it via a secure proxy for outbound
2) If proxy was blocked because of your zone, you could escalate and get deviation. Just need the right security person to approve. I've definitely launched stuff in prod and it's like whoops didn't know that would be blocked or forgot to submit to be opened.
3) Maybe you encountered a power tripping "security architect"...gotta find ways to avoid those.
4) Unfortunately in non-IT corporate, you need to not just be able to be good at engineering but good at cutting red tape to keep your team moving. It's time consuming & frustrating.
5) The in-house alternative is PCF, which is phasing out for Azure. I thought PCF was a dumb idea to start.
6) Same red tape exists for Azure :(
I'd say it's worse for security. Maybe this has changed...left GM 6 months ago. I actively refused to migrate to Azure till they worked out the kinks. I'd have to have a frankenstein solution to satisfy their policy. I.e, app is in Azure but dependent web services are internal... Azure is considered external even though it's walled off for us...but it's external and we can't connect internal stuff to it.
We had actual products to deliver and didn't have time to be guinea pigs.
This is not technology to ever be created and or operated with those who are tech bro minded or following tech bro mantras(i.e. fake it til u make it)! Innocent people lives are put in danger if so as we've seen with Uber and now Cruise!
She was hit and lodged underneath one of Cruises's robot cars on training wheels who unlike what a human would do (stop the car & help the other human being) ..it's algorithm (there's no way these robot cars can account for all the million human scenarios that a human can) executed dragging the woman to the side of the road further mangling her body.
As an aside, I wonder if all Waymo and Cruise vehicles will include undercar cameras in the future.
Further you are saying what I detest about this dumb technology ..that being ..oh once we hurt someone or killed them we will have learned something from their death and push out a software update with each person we harm or kill. Thanks for sacrificing your life for technology progress and expanding tech bros wallets.
As for Waymo they need to be strung up too if it's found out they learned something from cruise's incident and made changes. Which scarily shows they are learning as they go and more people are going to be harmed or killed.
Love technology but love human beings more!
Two pedestrians were killed today in my home town, for example. Tens of thousands die because of cars every year, either inside them or outside them.
Automobiles are dangerous technology. Before they existed you could be in a carriage and tell the horse to go home while you passed out. Thanks to the automobile, you’ll be dead instead.
And there is no excuse of learning going on either. Just hop in cars and tens of thousands will die. Again and again and again.
More importantly Any one profiting entity Putting robot cars which they surely know aren't ready for prime time but say or act they are is beyond egregious! Such startups or companies who do such things to win this race at all cost is putting profits/market share over the safety they bang their drum about and thus such behavior willfully harms & has harmed innocent lives. Both Uber and Cruise acted as such and both now are no longer because if it!
Human beings over technology that kills everyday for me unless they've proven like Waymo's track record /behavior shows they truly are about safety first .. human beings over profits/market share. Though I'd still love to know after cruise's incident did Waymo push out a software update ..uggh if so.
The limited framing rhetoric you're using here is exactly why Kyle had to quit. You - like Cruise - are choosing not to communicate the full story so that Cruise looks good.
Silicon valley has not shipped a successful product in the space that Cruise is in, and its quite possible no company ever will, so maybe "succumbing" to a different culture could be a good thing.
The future is here, just unevenly distributed.
This in particular
> But I hope they can maintain their Silicon Valley culture and not succumb to becoming a complete GM outpost.
is the last thing Cruise or the AV industry needs. SV culture is suitable for cranking out CRUD apps. SV has no business trying to produce safety critical software.
https://www.atlantamagazine.com/great-reads/no-accident-insi...
> it meant GM had produced a part that failed to meet not only its own internal minimums for torque but also basic federal safety standards. For all safety-related defects, NHTSA requires auto manufacturers and part suppliers to notify the agency within five business days after learning of a flaw that could lead to an accident. GM had not.
> The documents showed that engineers first noticed the engine stalling problem way back in 2001, four years before Brooke purchased her Cobalt. What’s more, the Cobalt wasn’t the first GM car with a defective switch; that would be the 2003 Saturn Ion. Despite internal probes over the following decade that recommended changes to the key system, no changes were made. Indeed, GM appeared to double down on the bad ignition switch, integrating it into other compact cars, including the Chevy Cobalt, Chevy HHR, Pontiac Solstice, and Pontiac Pursuit.
> As it turned out, DeGiorgio’s 2006 order came four years after an email he wrote to the ignition switch manufacturer, in which he ordered the company, rather than fix the part and incur delays, to “do nothing.” The email was signed “Ray (tired of the switch from hell) DeGiorgio.”
> The report also described the disturbing pattern of disregard for the defect. This could be witnessed in gestures such as the “GM Nod”—in which GM employees affirmed the need for a plan of action, but understood no one actually intended on following through—and the “GM Salute,” described by one employee as the “crossing of the arms and pointing outward towards others, indicating the responsibility belongs to someone else, not me.”
From the leaks it is clear that safety was never the top priority at Cruise.
Few people were surprised that Cruise car seriously injured a pedestrian by unnecessarily dragging her screaming for 20' in a completely unnecessary maneuver.
I'll stand by my opinion that Cruise is not long for this world.
CEO is a different role, and without any checks in pace he’s pushed the organization so hard it is where it is today. Even the incident aside, just look at the handling and drip drip of news since going from bad to worse.
IMO Cruise's fatal mistake was choosing San Francisco. It was hubris to think that they could change the minds of a populace and government clearly set on hating everything to do with tech. There are probably a bunch of cities that would be proud to host self-driving cars, rather than antagonistic. Things could have played out differently.
This is exactly the core problem. It wasn't the crash. Regulators expect there to be problems with new technology, but they want to do deep investigations when they happen so everyone can learn from them. This could had been a great learning moment for Cruise, the regulators, the rest of the AV industry, and the general public.
But then Cruise lied to the regulators, which is a HUGE no no. They don't take kindly to industry trying to hide information like that.
As opposed to one hit-and-run followed up by a hit-and-drag-on-tarmac. If I were that poor woman, I know which I'd rather endure: I think a human driver would have the decency to back-up first.
I think it was covering up for the accident that caused regulators to pull the plug.
If that were traced directly to the CEO, then he's out for good reason.
Otherwise, the CEO likely has to take the fall anyway to satisfy regulators who want consequences for executives who don't ensure regulatory submissions are credible.
But I suspect Cruise was already on CA DMV’s bad books after having so many publicized issues in SF and this was just the last straw.
It's a shame they tried to hide it in this situation. They could have had a relatively smooth rise if they hadn't; a few years of awkward coexistence with robot cars is a small price to pay for the benefits of (almost totally) freeing humans from the task of driving.
Correct me if I'm wrong, but a Cruise has still never avoidably struck a pedestrian. Even when there have been V2V crashes, the overwhelming majority were caused by poor driving from humans.
Trust is very important and if you can't trust your CEO to do the right thing and comply with regulations you haven't got anything.
"Whether Kyle Vogt keeps his job or not is I think a secondary question. The big question is whether Cruise, trapped as it is between the inexorable demands of the capital it has taken on and the impenetrable difficulty of the remaining technical challenges of deploying their fleet commercially, has a path forward at all. "
Worth noting Cruise service paused (i.e. not even supervised rides) in Austin on Nov. 15: https://youtu.be/1o1CTfsZy_k
Fine, if you want to judge Vogt on a CEO-friendliness scale, how to scam a company into paying billions of dollars for a no return, or how to occasionally posts on Hacker News / Reddit like a "Real Human" TM then feel free to consider him/Cruise a huge success.
In the real world Cruise currently has 0 cars giving 0 people driverless rides. The game is over. The mask has been lifted. It's over. Cruise tried to expand too quickly without the technology, profitability or safety to match. And has now crashed and burned. Now, this may be controversial, but maybe the CEO of such a company bares some responsibility for that?
I know what the replies will be: shut up random poster! Vogt built a $1 Billion dollar company from his garage, that is success by anyone's standard. But by that same standard Theranos was a success, for a time... Now both are virtually worthless.
At some point we have to start judging the success of self-driving companies by their profitability not by their PR.
What’s really unfortunate is I worked with many people who worked with Kyle and no one had a single good thing to say about him. And I’ve often heard that before with people who knew other CEOs but they kinda grudgingly admitted they did what was needed to get it done.
However over and over I just heard phrases like “lost” “aimless” and “confused mess” with Kyle and it never was the “he’s an asshole but it makes the company work. It was just. He’s lost, distracted and constantly sidelined.
I say all this because I actually feel for the guy. Because I know if I was CEO of a company going up against trillionare competitors I’d probably be the same way. Probably even worse. And that’s the state most CEOs are up against — completely unable no matter their skill to outmatch the largest monopolies the world has ever seen — who can simply outspend, outlast, and out do you with capital alone.
not exactly a stellar rating
I worked with Kyle at Justin.tv. He's great. A good guy and a tremendous engineer.
Now you know someone who does, so you can stop saying that.
Some of these are qualities that I could see leading to someone thinking he was "aimless" or a "confused mess", but these are also things that engender a lot of respect from employees. For better or worse his presence was very much felt by engineers, and I think everyone thought that he was very technical.
I think Kyle's departure is going to hit hard for a lot of the engineers working on onboard software. It's been reported that they are spending something like $250M/quarter so in order to keep the ship afloat they are going to need to raise funding soon. It's hard to think of a worse situation they could be fundraising in, which I assume is going to mean a massive down round, which is going to further dishearten the people working there.
Respect from employees is a fickle thing. If someone can do all that you describe and still succeed as a CEO then power to them. When it comes time to choose between one or the other, I think most employees would rather an effective boss than one that seemed to understand them. Their opinion on the CEO might change rather quick were they to be laid off because the CEO just can't cut it.
It's sort of like the cool parent that's also a friend. If they can do that and effectively parent, they're better at it than me. If it comes down to being a friend to your child or parenting them, one is nice and the other is a responsibility and has major repercussions when done wrong.
I think aCEO should have people that work under them that can actually perform the duties they were hired for. Spot checking department and employees to make sure what's being reported to the CEO is accurate is probably a very good idea, but if the CEO is spending a bunch of time in meetings with people that they've already hired others to manage, that either shows there's a severe problem in that department that requires the CEO's constant attention, or a lack of trust in the people they hired to run that department to actually do what they were hired for. Neither are signs of what you'd want to see, IMO.
Is this a positive thing, tho? I am not sure. I am just curious.
[1] https://www.cnbc.com/2023/11/06/cruise-confirms-robotaxis-re...
After giving me the offer ("We would like to offer an internship position to you. I'll call you today with more details.") the recruiter emailed me that Kyle wanted to talk. Since I'd already been offered the internship, I took it as a sell call to learn more about the company.
But shortly thereafter when I emailed to accept the position, the recruiter responded
> After your conversation with Kyle, we felt that we will need a more experienced intern for the projects we have. So, we are not going to pursue your application any further.
At no point had I been told—until I tried to accept the offer—that it had been rescinded.
Thankfully, I hadn't yet reached out to the other companies who had given me offers and ended up at what turned out to be a much more successful startup. But if I had done it in the other order I would have been screwed.
This was more than 12 years ago, and I still don't know what happened internally. Having worked in many startups since then, I can understand the level of dysfunction that can occur even with well-intentioned leadership.
But in any case, Kyle clearly didn't care what happened to some random college student, and it's left a bad taste in my mouth for all of these years.
I've heard it said that Musk has had an even better week, for the reasons you mentioned plus the Starship launch.
https://blog.comma.ai/a-100x-investment-part-1/
https://comma-ai.medium.com/a-100x-investment-part-2-569b714...
>Last month, we had 1,209 cars drive a little over 1,000,000 miles
Let's say they've had zero growth since then, so 48,000,000 conservatively?
Actually, from their website [1]:
>100+ million miles driven and 10k users.
[1]: https://comma.ai
He also claimed that he can fix the search in Twitter, but he left after 2 months and now he is working on something else.
To put another way, there is a better chance that Musk is the next OpenAI CEO.
Personally, after CA DMV pulled their license to operate. Cruise ceased all operations in other states (particularly in TX).
Their vehicles were by far the worst to deal with as both a pedestrian and driver. As a pedestrian, I have had more close calls with their vehicles compared to human drivers. They have caused unnecessary traffic jams (in Austin, TX there was an infamous incident in West Campus that took forever to clear). When their vehicles pulled over, they often did not pull over completely and thus becoming obstructions on the road. Cruise flooded the streets of DT austin at night to pump their mileage numbers as well. It was quite a common sight to see at least 2-3 cruise vehicles circling the same neighborhoods and even hogging all of the EV charging infrastructure.
Good riddance. I say.
At least in this case, Vogt is doing a clear resignation on his own terms.
Cruise is an extremely tough spot right now after its handling of the San Francisco accident, and it's nationwide pause of testing.
OpenAI has been extremely successful but is having strategic disagreements within its non-profit board.
> At approximately 9:30 p.m. on October 2, a human-driven vehicle struck a pedestrian while traveling in the lane immediately to the left of a Cruise AV. The initial impact was severe and launched the pedestrian directly in front of the AV. The AV then braked aggressively to minimize the impact. The driver of the other vehicle fled the scene, and at the request of the police the AV was kept in place.
Why was cruise under fire for this?
Oh well. Maybe next decade.
And no, most human drivers would easily detect a human being stuck under their vehicle within seconds.
There's a good chance the woman would have been _killed outright_ if it hadn't been a Cruise. Are we willing to accept strange and inhuman failure modes if they overall lead to better safety than human drivers?
https://www.sfchronicle.com/bayarea/article/woman-run-autono...
It should be "was the eventual outcome better than what a human driver would have achieved in the same situation?"
It's not actually clear that a human driver would have done better, and given that the human driver in this incident did a hit-and-run and still hasn't been caught, assuming that is... questionable.
The safety record of these cars is actually quite good when you consider the environment and times of day they operate in.
Furthermore, Cruise attempted to cover up the fact that the car dragged the woman. Now that I refresh myself on the details of the situation, it seems you have the "cleaned" PR statement that was pushed by Cruise and had omitted the above details
https://www.latimes.com/business/story/2023-10-26/cruise-rob...
Anyway Altman gives me the ick, serious lizard person vibes.
I don't really see the parallel.
It seems impossible to have success in self-driving without safety. Cars that are unsafe will literally kill people, which will then kill the program as seen with Uber.
The AI safety/success tradeoff seems far more nuanced. The threats are abstract, largely hypothetical, and safety seems to be directly in opposition with success. So it's easy to see why product-oriented people would feel justified in making a call to ignore or down-prioritize safety, in a way that nobody running a self-driving car program can afford to.
Push forward at all cost, expand as fast as possible, see regulators as stifling forces not to be trusted or taken seriously. Can almost guarantee the Cruise CEO was in that circle of thought and drank the Koolaid too hard.
And then we had all the other leaks.
That they couldn't reliably distinguish children from adults, making them (even by their overly optimistic models) more dangerous to children than humans - which is why they started operating mostly at night, when there are few children.
And then it was reported they had 1.5 employees per car in operations, requiring support every 2-5 miles. Very surprising given that they've reported a disengagement every 95,000 miles last year.
Cruise was a company that could not withstand transparency.
If true that alone should have disqualified them from being street legal.
That's a utterly ridiculous statistic. If you need a human to monitor to prevent a bad result, it doesn't matter how often the human has to intervene.
For example: With Tesla Autopilot, the driver doesn't have to do anything the vast majority of the time. Still, your life expectancy will be measured in days if you don't monitor it.
Even if the vehicle makes completely random decisions when faced with a binary code, you could still say that 50% of the support events don't result in the remote operator doing anything.
A human driver can easily get away with the latter, and often do. A self-driving car can't, because people hold them to higher standards.
And they need all those people because their software cannot reliably handle it itself.
Otherwise they would not used so many people.
> A human driver can easily get away with the latter, and often do. A self-driving car can't, because people hold them to higher standards.
LOL. They've been blocking streets all over the place.
https://techcrunch.com/2022/06/30/cruise-robotaxis-blocked-t...
They're also overstaffed to deal with surges - I recommend you go read the (ex) CEO's response to these claims. As the number of cars on the road goes up, the number of employees required to deal with surges increases more slowly. At some point the ratio increases from 1:20, which is where it's already at.
Looks like they haven't increased the number of operators enough.
> They're also overstaffed to deal with surges
Are they? How are they going to handle the next bay area earthquake? Will they have enough operators for that, or are they going to block first responders all over the city?
Fortunately, this is a moot point, because it is doubtful whether they are going to survive.
> I recommend you go read the (ex) CEO's response to these claims. As the number of cars on the road goes up, the number of employees required to deal with surges increases more slowly. At some point the ratio increases from 1:20, which is where it's already at.
Nice theory. Note that the number of remote operators shocked everyone that was following Cruise. They were constantly asked about number of cars per operator, what exactly the duties of a remote operators were, etc. They always avoided these quetions, and for a good reason. It turns out this was far worse than anyone outside Cruise could have imagined.
Who will be willing to spend a billion or two every year with no end in sight?
Only Waymo and Cruise will make it, they said.
Will be very interesting to see who is left standing.
Maybe Tesla will never crack self-driving, but an insolvent company definitely won't either.
Meanwhile, as an occasional tesla passenger, I get to enjoy watching the car's representation of the surrounding traffic have artifacts, like:
1. Cars disappearing, re-appearing, and disappearing again.
2. Trucks and buses slowly drifting until they phase 'into' other cars when traffic is stopped at a streetlight.
It's a fantastic advertisement against self-driving.