I suppose on some level it isn’t exactly surprising that John Krafcik, the ex-CEO of Waymo (2015-2021) doesn’t think Tesla’s Cybercab will have what it takes to be an actual, functional self-driving taxi. Waymo currently operates robotaxi services in Phoenix, San Francisco, Los Angeles, and Austin while Elon Musk, the little-known CEO of Tesla, has been promising that everyone who has a Tesla will be able to make money from their Teslas getting side jobs as robotaxis for about a decade now, has yet to make that actually happen. So it’s not too surprising to find that the former head of one of the companies actually operating robotaxis may have things to say about Tesla’s very public and hyped plans to get into the same business.
Krafcik gave an interview to Germany’s Manager Magazine earlier today, where he did a pretty comprehensive job of explaining why he felt that Tesla is nowhere near ready to start a robotaxi service, based on their current level of automated driving hardware and software.
Perhaps most obviously, Krafcik notes that Tesla’s cars, including the new Cybercab concept, which was designed specifically to be a self-driving robotic taxi, lack the type of hardware Krafcik believes is needed for an effective and safe robotaxi, including redundant sensors, high-resolution cameras, and equipment to clean cameras and sensors from grime, ice, and anything else that may impair their use.
In fact, Krafcik seems to have nothing but issues with Tesla’s design of their robotaxi, the Cybercab, saying of the prototypes:
If a company were serious about building a safe robotaxi business, the robotaxi wouldn’t look anything like this prototype. A serious robotaxi would demonstrate the primacy of safety; the manufacturer would place sensors in optimal positions—on the roof, as well as on the sides and corners of the vehicle. These sensors would also have cleaning and drying functions—windshield wipers, compressed air nozzles, and so on. A serious robotaxi also wouldn’t have a low-slung coupe body design. This design makes it difficult for people to easily get in and out; not everyone will be able to use these robotaxi vehicles comfortably.
We’ve wondered about the design of the Cybercab as well, though more about the logic of a two-seater doing taxi duty, logic that, frankly, eludes me.
Krafcik’s criticism and skepticism of Tesla’s robotaxi service, scheduled to be launched in June, is pretty dismissive, with him noting that
“There are many ways to fake a robotaxi service,”
… and implying that this is, in fact, what Tesla may end up doing. Krafcik mentioned teleoperation, cars that lead or follow, and extremely limited operating domains as ways Tesla could “fake” a robotaxi service, but he also suggested that maybe it just wouldn’t happen at all, and be another missed milestone from Musk.
I think the most surprising thing revealed in the interview had less to do with Tesla and their chances at deploying a robotaxi service by June, and more about the challenges of automated driving itself. Counter to most people’s estimations, Krafcik believes that highway driving will actually be the bigger challenge for automated vehicles.
While driving in urban, high-density areas seems like the greater challenge for a self-driving vehicle, with much more complex terrain, higher traffic density, the presence of pedestrians and animals and all manner of other obstacles and distractions, Krafcik states that the increased speeds and presence of many large vehicles like big rig trucks actually makes for a more challenging self-driving environment.
As far as the lack of “edge cases” – a term used in the self-driving world to mean pretty much any normal thing you may encounter when driving that isn’t strictly part of the expected set of elements – on highways, which most people assume would less frequent, Krafcik has thoughts on that, too, in the context of the possibility of automated trucks on highways:
Almost all of the challenging circumstances and vulnerable road users found in cities also exist on highways—only less frequently. We’ve seen cyclists, scooter riders, and pedestrians on American highways. The rarity doesn’t make things easier—it makes them more difficult. You can’t ignore these extremely rare events; you have to solve them robustly, even if the speeds are much higher and the stopping distances are much longer. This means that the sensing, perception, behavior prediction, and path planning aspects are much more demanding for autonomous trucks than for slower-moving robotaxis in the city.
The entire interview is interesting, certainly because Krafcik should understand the challenges of operating a fleet of self-driving cabs better than anyone.
Will Musk and Tesla manage to start everything in June and make Krafcik look like a fool? I suppose it’s possible, though I wouldn’t counsel anyone to do any breath-holding.
Picture a Johnny Cab.
Only he’s wearing a little Nazi SS uniform.
“You can’t ignore these extremely rare events; you have to solve them robustly.”
Serious question – why? Why must you solve extremely rare events? In today’s America, why would any remaining authority figure have any interest in holding today’s taxi vehicle vendors accountable for preventing these events?
I took my first Waymo ride in SF earlier this year. I was really impressed with how it drove – it felt very human its in responses. The way it accelerated, braked, etc. all were familiar and not unnerving. When it ran over a bumpy section that was pitching the car side to side, it slacked off the throttle a little to calm things down. When it was at a four-way stop, it started to creep forward after stopping, but a car coming at 90 degrees didn’t fully stop and went on into the intersection. Instead of the Waymo jumping on the brakes to avoid a collision, it realized that its creep speed (maybe 1-2mph) it wasn’t going to hit the car, so it just kept creeping forward until the car passed and then the Waymo sped up.
Leon Skum has never: Repeat never brought a product to market at the time he said it would. He uses hyperbole to inflate his ego and the stock price of his vaporware, along with deposits that the company seems to have a difficult time refunding. Is Tesla a good car? Ef if I know, since I would never purchase anything that had Musks name attached to it, but some people seem to like them. Then again, some people like Andouillette. (look it up, as I had the experience of ordering it thinking it was andouille sausage)
Wish I didn’t look it up
BIG SAME. Don’t do it, folks. Tell your curiosity to shut up.
I just got back from Toulouse, France (home of Airbus). Toulouse has the most wonderful autonomous Metro. Trains are small but come every five minutes. They are building a third line that should be complete in 2028.
All I kept thinking was…. “Why can’t we have nice things too?….”
We COULD do that, but it would mean that a few billionaires have slightly less money, so you can see the predicament we’re in.
You’re basically describing Vancouver’s Skytrain system, as well. Trains every 2-3 minutes during rush hour, 5-7 minutes outside of rush hour.
“ … it isn’t exactly surprising that John Krafcik, the ex-CEO of Waymo (2015-2021) doesn’t think Tesla’s Cybercab will have what it takes to be an actual, functional self-driving taxi…”
I don’t have his credentials and even I know that.
His credentials are certainly hard to match! He’s the guy who coined the term “lean manufacturing” and wrote the paper on it in 1988. It’s an interesting read and is publicly available online
He was also behind Hyundai’s turnaround in North America in the 2000-2010s and helped shed the perception that Hyundais were simply crappy low-quality copies of Mitsubishis
Hopefully we’re getting close to a realization that’s been true the whole time: Human beings are incredibly good at recognizing patterns, observing deviations from those patterns, assessing telemetry and responding to observations. And when driving a car almost all of that information comes in through one set of “sensors”: human vision.
and those sensors come with a state of the art self-cleaning feature.
Digital olfactory clearing function is among the most elegant solutions for sensor cleaning and maintenance ever devised.
With bonus gustation feedback loop (optional).
Vision, sound, and the many different kinds of physical feedback that comes from driving a car.
Then consider that Tesla’s “camera-only” approach uses two dimensional images from cameras that can be occluded or completely disabled by various conditions, and the human eye is an inexorable part of a larger neural system that’s evolved over millions of years to detect even the most subtle event in order to keep us being eaten by whatever lurks in the dark tall grass.
As you say, humans are incredibly good at this. A couple cameras and a bit of circuitry, not so much.
I’ve been working on autonomous vehicles, with industrial sized robots on private property, for almost a decade now. Even with the extremely helpful constraints in a distribution center yard with no public vehicle access, autonomy is incredibly difficult. Krafcik points out something that you would likely only discover from hours and hours of testing – the keeping sensors clean problem, that truly is a challenge to handle, plus ensuring you have the correct sensor suite to handle whatever real world challenges you encounter.
the shorter version: Musk is a moron.
Musk is a snake oil salesman. His customers and fans are the morons.
The model 3 and Y, and the Falcon 9 are not snake oil. NASA, the NSA and the people who buy the best selling car in the world are not morons. Take a smaller bite next time.
Correct, but if we were to valuate Tesla for only model 3 and Y sales performance it would not be valuated half the current value.
I agree, Tesla stock is over-valued for that reason.
Every product of his you’ve just named has exploded multiple times, and two of them have caused fatal accidents because of their alleged full self-driving capabilities.
Never really thought about the process loop before. Yeah, he’s right. The faster you go, the less time the process loop has time to run, and the feet of earth travelled during said process loop is much higher, thus leading to more things to process. Interesting.
Want to know what would improve autonomous taxis? Putting them on rails so the automated part is a lot easier. Then we could chain a bunch of them together and maybe make most of them just dummy follower cars, while the lead one could be powered. This would reduce the cost of most of them. Then we could make all of them bigger so we could improve economies of scale. Heck, what if we scheduled them to take off from fixed locations at fixed time, and arrive at other locations at fixed time so it was incredibly easy to figure out how to get places.
And so that they could avoid traffic, maybe we could move these ‘rails’ you speak of into underground tunnels, or maybe elevated above roads/buildings etc.
Nah, on second thoughts, it’ll never catch on.
To build on what you’re thinking, I bet if they were in tunnels or elevated rails, we could probably make them travel at much higher speeds than cars can drive safely, saving people taking these “automated taxis” a lot of time.
The funny part is Elon Musk has enough money to deploy an actual rail system across the major cities of America, make money from it and get even richer without even trying. Of course this only works in a world where he is sane and normal.
He doesn’t have THAT much money.
Phase 1 of California High-Speed Rail is like $28-35B. A fraction of Musk’s net worth. Total cost is expected to be like $130B last I heard, which is still less than half of Musk’s net-worth of a couple days ago (~$301B).
Yes, net worth is not the same as money. But he managed to get enough money to buy Twitter for $44B.
Lol I’m sure if he fronted 10B for an ACTUAL rail system, not the outright scam that was Hyperloop, institutional investors would be clamoring to put down to rest. Of course, all this makes sense in a different timeline, not our “Dark MAGA” timeline.
See the comment from Doughnaut below. he definitely does.
“across the major cities” was the quote. He has enough money to do CA and maybe Boston-NYC-DC and that is it.
I’m curious if anyone really believes that this gold plated 2-door hatchback is anything more than a stock pumping device. They can’t build a self driving car, they don’t know how and they’re not really trying.
This is the Fuel Shark of self driving cars.
But remember that people actually buy the Fuel Shark. Because they’re stupid. Lots and lots of stupid people make these scams possible.
Oh STOCK pumping… I read that very differently the first time thru, although I guess the meaning wasn’t too far off.
Anybody can start a robotaxi service if they know they will receive the Congressional Medal of Freedom and Bronze Star rather than a warrant when they run over a kindergarten class in a crosswalk.
The only self-guided motor vehicles I trust are at (how appropriate…) Autopia at Disnelyand.
> Will Musk and Tesla manage to start everything in June and make Krafcik look like a fool?
I’m sure he’s devoting 100% of his attention to this industry-changing innovation in order to hit the deadline as well as safety and quality standards.
Tesla can’t manage to self-drive in a *very* limited access tunnel under Las Vegas doing a one-way loop with two defined stops. What on Dog’s Green Earth makes them think they can make successful robotaxis?
Ketamine – that’s what makes them think that.
“There are many ways to fake a robotaxi service,”
In a parallel situation, Elizabeth Holmes had cut a deal where Theranos actually had blood testing machines in a few drug stores.
Except they were fakes which didn’t do any testing. Theranos would send someone over to pick up the samples pulled by the machines and then take them to a regular lab for standard testing. Then they’d send results to customers pretending that they were from the supposedly automated blood testing machines.
What’s more, the blood draws the machines managed to perform had unacceptably high rates of flaws.
What’s crazy is that every reporter and editor in the tech world should know this based on the Wall Street Journal exposes, the book Bad Blood, and the related movies that came out. Almost none of them will connect the dots between Theranos and Holmes and Tesla and Musk, but they’ll all be offended if anyone points out how they’ve dropped the ball.
Came here for this.
I may have mentioned this before, if so ignore this.
Single track road, really one vehicle at a time single track road, the passing places are farm gates, field gates and one every five miles for 40ft wagons. No road markings, the width of the road is seasonal. In a snowy winter there is no road, in springtime there is a road, and the trees and bushes and grass are doing what they do, in the summer there is a road but it narrows, cow parsley and bilberries, second brood partridge and pheasant chicks.
I have not mentioned the sheep, or the Belted Galloway, or the pigs. Or the steam engines and rally cars.
It takes a week to negotiate terms with the only local taxi firm.
Just make the bus service align with the train!
As soon as they showed off their crap “taxi” design, we all knew they weren’t serious about actually doing it for real.
I like that this guy mentions having redundant sensors and sensor-cleaning equipment. You don’t hear many people talk about such details. I’d like a compressed air nozzle for my backup camera, incidentally.
My Bolt EV sprays the backup camera with washer fluid at the same time it sprays the rear window. Works surprisingly well, I’m not sure that adding compressed air would make it any better.
I figure compressed air is a good way to blast water off the lens without much hardware needed; I’m not so worried about other types of debris. But I’m jealous that you have anything!
My GTI did it even better. The camera is stowed away when not in use, and in the process of going in and out gets wiped off. Always clean and ready to go.
Those are pretty fun to watch when someone shifts into and then out of reverse. It will be interesting to see how well they hold up over time.
It’s an electrical device on a VW what do you really expect.
Right?
I was just in Japan, and your self-cleaning camera does exactly the same thing as the automated self-cleaning bidets over there.
My partner has a Honda Pilot and when it’s been raining, a large drop of water pools on the backup camera lens. I’m going to try to 3D print a cover or something to address it.
That’s just what I mean. My Civic had that problem pretty bad. Maybe it’s a Honda thing, actually. It’s inevitable, since it’s a spherical lens, and pointed slightly down, so you get that big hanging drop right in your line of sight. A flat cover would fix it, or at least make it much less of an issue.
From their job descriptions it sounded like they were trying to do “engineer” VR take overs for when the AI freaks out. That also seems like it’s not that safe. It feels like we have been waiting for self driving for 15 years at this point there are some systems that get close but don’t know how to get out of trouble often will go right past road closures. If you wanted to trap a bunch just put a road closure sign and all the self driving taxi things seem to bunch up and make a big mess.
I think we’ve been waiting a lot longer than 15 years. Remote control cars were demonstrated in the 1920s, full-scale autonomous highway prototypes and accompanying cars were built and tested in the 1950s, the first car capable of driving autonomously without assistance from any external infrastructure was demonstrated in the 1970s, most of basic technologies being used as the basis of current experimentation were in place by the end of the 1980s, at least in preliminary or conceptual form, and Congress passed a law in 1991 specifically mandating the US Department of Transportation to develop and demonstrate an autonomous driving system by 1997, which lead to the creation of the National Automated Highway Consortium of industry, government, and academia, in order to develop a single universal standard for such technology.
We’ve heard it could be practical and widespread in one form or another by 1960, by 1981, by 2000, by 2020, by 2025, by 2026, by 2030, etc, etc.
Self driving is five years away. And it shall remain five years away well into the foreseeable future.
I really think self driving technology is a solution in search of a problem. Do people really hate driving that much?
If so, there are plenty of options that don’t involve questionable tech that could easily suffer an ABEND/BSOD/whatever-hang-condition-exists-on-the-platform while carting you down the road at at over 100 feet per second.
I don’t want to be in such a machine, nor do I want them on the roadways with me.
Yes, there are plenty of bad drivers I’d really not want on the road with me either, but at least that’s somebody to sue for whatever damages they cause.
I think advanced cruise control, the kind that automatically adjusts speed to maintain safe following distance on Interstates, is realistically the most that people really want
Although, I could see actually autonomous driving as being useful for the elderly, who can’t drive themselves anymore, but are still mentally sound, haven’t designated their son power of attorney, and absolutely refuse to move out of their existing house to a place with decent public transportation. But, they are also probably the least likely consumers to trust a computer to do such things for them, and, also, in my state, the DOT provides free shuttle service statewide to those sorts of folks, albeit with advance planning and scheduling
I think the ideal use case are the elderly or others that aren’t capable of driving safely. Having seen this play out with my grandparents and my wife’s, safe and reliable self-driving cars would be a huge improvement in their quality of life and the rest of the family trying to help them out.
Obviously what Tesla (or anyone else) has at the moment is no where near good enough for that use case, but I hope that we’ll be able to get there by the time I’m in that position where I can’t drive anymore.
Yeah, my grandmother is 94 and somehow recently passed a test and renewed her license, I say somehow, because her Camry has been through some serious shit in relatively few miles. A couple of years ago, she got in an accident on her way to the airport, took a taxi the rest of the way, and my uncle had to fly back with her because she wasn’t sure where she left the car and he needed to track all that down
Still not as bad as my grandfather, he died of Alzheimer’s several years ago, but, before that happened, had been getting in all sorts of weird accidents and then trying to hide the results in creative ways, like ripping the bumper off his truck, then placing the bumper behind some boxes in the garage, because it wouldn’t look at all suspicious if the truck was sitting in the driveway missing a bumper
Thing is, I really don’t think she’d ever use a robotaxi even if it was an option. She’s been using computers continuously since sometime in the 1960s, but seems to draw the line at desktops, every cell phone we’ve tried to get her over the years has gone unused and shoved in the back of drawers, never even being charged once.
I know an old couple in their 90s both still driving and have 5 vehicles. Each one has seen some things. Both of them are basically blind and still somehow pass the yearly exam they make them do. I often get calls about cars making strange noises and when I check I find road signs or cones under they she ran over and somehow dragged back home. They hit a guy on a bike at least one time but just bought him a new bike because his was destroyed. He didn’t want to get the police involved. They go though tires alot too hitting things. I can’t tell you how many side mirrors I’ve replaced for them from them hitting signs and who knows what else. I’ve told their kids about it and they just seem to think of them as an inconvenience and seem to be waiting for them to die for the inheritance.
I’ve talked to them about the ride sharing apps. I got them both smart phone the old guy doesn’t like it and went back to a kinda smart flip. But I know they wouldn’t get in a car that drives it self. They have enough fights with a printer. When I’ve talked to them about it they said they think it would crash into the water.
I knew a guy like this; I knew this driving medical was coming up and I wrote a stern letter to his physician saying in no uncertain terms that he was unfit to drive (I’d been a passenger). He lost his license.
I can relate. My mother is 88 and just crashed her car into the back end of another in the stop and go at a 4-way stop last week. Fortunately, no injuries to either party. But her 13-year-old Accord with maybe 65K miles on it is almost certainly totaled.
She lives by herself five miles out of town. I gave her a PC back in the 90s and she got reasonably competent with it. Years ago, I flew down and gave her a used cell phone to use primarily as a wi-fi hotspot. She was on dial-up until then. I also explained she could use it a phone wherever she went and could call my brother and me without long distance charges. A few months later, I received it in the mail, carefully wrapped in the original packaging. So, I have replaced it with various cellular-based solutions over the years.
Now, she doesn’t remember how to use the computer and claims that icons on the desktop are constantly changing location or disappearing. They aren’t.
My brother and I have some tough decisions ahead. Physically, she is in good health but I’m afraid her body is going to outlive her mind. 🙁
Oh, that’s familiar, I have an uncle who has refused to set one foot into her house for the past 5 years, because she had a habit of blaming everything that was slightly amiss on him messing with her things when she wasn’t looking – TV turns on on a different channel, his fault, no new trash bag in bottom of can, his fault, remote control somehow placed inside the crisper drawer of refrigerator, his fault. She also moved into an assisted living facility a few years ago, but refuses to sell her house and still sleeps there once a week (refuses to use the laundry room at the facility, because it might require interacting with people, so she drives back to her house to do laundry there and typically spends the night, then drives back). Seems like an expensive way to do the wash, but, what do I know?
My sympathies. I think my brother and I would be okay with taking her back home to do her laundry, but when brains go squirrely, there are so many landmines. Best wishes to you going forward, when/if you have to deal with this.
Oh, I don’t have to deal with it, my uncle who’s an employment lawyer in California has power of attorney, the two of them apparently wanted it that way, so anything that happens is between them from now on
I saw a conference presentation from the head professor of a computer vision department at a major research university last week. A true scientist not trying to sell anything.
He made the 5 year joke, but said we’re realistically decades away, if it ever happens at all. There’s a reason all these Waymo towns are in the desert. Plain rain throws off cameras and lidar.
I’ve been joking that it’ll be ready in 7 years for the next 10 years, just like fusion!
“Do people really hate driving that much?”
When its the exact same boring AF 1-2 HR crawl twice a day every day?
YES!!!
Dare I say your problem here isn’t that the drive is boring AF, your real problem is that a commute like that is even necessary. Even if you weren’t bored behind the wheel, that’s a lot of time given up each workday.
I guess if you could do something productive while the car drove you to and fro, there would be some improvement, but that would still come at an increased risk of death or serious injury due to software failure. (I write code for a living – trust me, software ain’t perfect and edge cases pop up and bite you all the time!)
If you haven’t seen it, go find Sam Kinison’s “cure for world hunger” bit on YouTube. You might find some inspiration at the tail end of it.
Preaching to the choir. I find the return-to-office movement despicable, at least for employees whom benefit from it less than they suffer from it. Which I figure is a lot of them.
“If you haven’t seen it, go find Sam Kinison’s “cure for world hunger” bit on YouTube. You might find some inspiration at the tail end of it.”
The difference being food can’t be sent electronically whereas telework can. There’s also limited living space in commercial areas and IMHO living in tiny, cramped apartments in high density housing units sucks ass.
The truth is return to office isn’t about increasing productivity and improving company culture is only tangential. The real motivations are to prop up sagging corporate real estate prices, to increase control over workers, to increase tax revenue, to prop up businesses that serve corporate real estate, to force workers to relocate to depressed downtown areas and to try to shed workforce without severance compensation.
The stealth layoff thing is very real – and it’s pretty disingenuous to claim “oh, we need it for our culture” while offshoring departments overseas.
I hadn’t considered the “need” to prop up the value of their real estate holdings. That’s probably a fool’s errand at this point. But I guess they have to try.
That “need” is very real to the bottom line and has been a main motivation since the very beginning of the lockdown. Companies and even government have no problem doing whatever it takes to prop up those values including theft of employee non work time.
We’ll have “universal” (anywhere, any weather) self-driving cars right about the same time as flying cars. Those have been 5 years away since the 1950s too.
I fully expect we will have commercial nuclear fusion-based power generation first.
Silly Valley is really good at coming up with solutions in search of problems to fleece the punters with.
According to Popular Mechanics in 1991, I’ll have a flying car by 2000!
They said the same thing in 1951, 1961, 1971, and 1981 too. And probably 2001, 2011, and 2021. And they continue to be wrong.
Yes, it’s hard for us as car enthusiasts to understand, but many people really do hate driving that much. And the solution is not self driving cars, but walkable cities.
A cool side effect of this solution is that you can still make the unit economics for interesting cars work, and they don’t get out-competed by blobby CUVs, because the people that would otherwise buy those vehicles just… don’t. They walk or bike or take the subway.
Remember that Ford didn’t get rid of the fiesta and the focused because they weren’t selling. They got rid of them because selling compromised products to uninformed buyers who aren’t interested in cars offers higher margins.
That Fiesta/Focus take is pretty cynical. And likely dead on.
I always thought it was a bit dumb of Ford to drop the Ranger for the Transit Connect back in 2011, but you’re probably right – forcing certain buyers out of the truck option likely benefited the team running the van product.
While I know it wasn’t a 100% conversation (our local O’Reilly stores went to Frontiers instead of Transit Connects), it probably made sense to the bean counters.
I have absolutely ZERO interest in living in such close proximity to enough other humans for it to make sense to have “walkability” – BTDT, never again. But if you want to live like bees in a hive, you do you. If I had my druthers, it would be a 20 minute walk to get down the driveway to my mailbox.
The driver is the single most expensive component of commercial driving – trucks or taxies.
The goal of self driving tech is to reduce costs for commercial fleet operators.
If a self driving vehicle hits you – you sue to the company that made it. Waymo has much deeper pockets than the average Joe.
I think over-the-road stuff is going to be viable long before in-city automated driving, especially with larger vehicles. That’s actually relatively realistic in the mid term, as long as the transport happens on controlled access highways.
I don’t see taking humans out of the last mile any time soon, whether it’s delivery or taxi service.
I do see drones working well for the last thousand feet. I can imagine a truck pulling over and sending out drones to specific addresses in order to drop packages (size permitting). That will at least enhance productivity.
There is no plans to automatic the last mile in the near term.
The plan for autonomous trucking is hub to hub. A human driver picks up the load, drives it to a hub right off an interstate, and then the trailer is transferred to an autonomous tractor. It gets a DOT inspection and is sent on its way. The autonomous tractor drives 300, 400, 500 miles to the next hub and then gets another inspection and refill. It either gets sent on again or the trailer switched back to a human driven tractor that take the load to the delivery dock.
The automated driver does the big miles and can run outside of DOT hours as there is no human to get tired. Trucking becomes a hourly job for the human driver with them home every night in their bed.
Of course this only works for clear roads. Nobody is expecting autonomous trucks to drive through snowy blizzards any time soon which is why development is concentrated today in Texas and along Interstate 10 East of the mountains.
I appreciate that Musk was smart enough to get into the wireless ISP game when his self-driving project would likely require remote control of the vehicles. The internet is still not as reliable and ubiquitous as water or electricity or natural gas (which have a 100+-year head start), and so relying on public internet latencies or even availability would be super dangerous. Running their own ISP reduces some of that risk.
The modern version happened in the early 2000s. During the Iraq war DARPA set up funding for autonomous cars to handle convoys.
https://en.wikipedia.org/wiki/DARPA_Grand_Challenge
Almost all autonomous driving systems today come from that work.
I think it’ll happen quickly. The rest of AI is moving *fast* – AI won two Nobel prizes last year: https://www.nature.com/articles/s41746-024-01345-9. Tesla’s biggest advantage is xai/grok which is also moving fast, and years and years of data.
Apparently people in SF simply put a traffic cone on the hood of the Waymo taxis, and they stop dead in their tracks.
https://www.npr.org/2023/08/26/1195695051/driverless-cars-san-francisco-waymo-cruise
Waymo can do remote takeovers when needed, but it’s only used when absolutely necessary
That’s cool. Today I watched 2 waymos go around a road closure due to a fallen tree and get stuck until someone came to rescue them.
“Former Waymo Head Doesn’t Think Tesla Is Close To Ready To Deploy Robotaxis”
Ya think?
Me either, and I’m just a guy on his couch.
I have my metaphorical bowl of popcorn
So I have to ask: how is John Krafcik‘s couch?
Interesting the guy from Whammo home of the frisbee and other fun toys that while operating legally is definitely not ready for prime time player status is dissing the guy who started the robotaxi concept. Now both of these vehicles are killer, of people who have no choice when they were killed by these vehicles. But the Tesla owners are so close to making money they just need to rent them out for the renter to drive instead of a robot. But I bet if we had an economics person at Autopian we would do the math and discover renting a car out to an individual who drove it by the day as opposed to renting a car out by the Mile with a robot driver was more profitable. The rental requires the renter to have insurance the self driving taxi must provide it’s own insurance which I bet isn’t cheap. Proper maintenance gets the owner off with rental self driving it is all the companies fault. Taxi is 24 hours of running for x a rental is still parked most of the time. But alas no proper economics.
So is there a /S at the end of this rant or did someone slip you a few hits of ketamine?
Not so much /s as a tongue in cheek approach to say that the guy is really no expert as his approach is failing too. And to suggest that after companies realize the whole self driving is not possible or practical they would be better off just renting their cars out on the app rather than allow the car to just be ordered by a stranger. IMHO
I did think my comment on self driving cars to the moon pretty much said he was not trustworthy when he makes promises maybe It wasn’t radical enough that people needed an /s?
“It wasn’t radical enough that people needed an /s?”
Its how your audience knows you didn’t forget to take your meds that morning. Or that you didn’t take someone else’s meds.
Did you just claim that Musk started the robotaxi concept because that is actually hilarious.
Nope I believe that self driving taxi concept has been around longer than any of us has been alive. Musk MAY have been the first to PROMISE it, but hasn’t really come close yet.
Ah, I misread your comment then, sorry.
Waymo had actual autonomous taxies on the roads giving rides bake in 2015. That was before Musk made his first Robotaxi promise back in 2019 when he needed to pump TSLA shares.
Setting aside the last year or so of politics, hadn’t we already gotten to the point where we should know better than to believe anything Musk says? He’s been telling lies people want to hear so they stop asking questions for at least as long as he’s been running Tesla.
IIRC wasn’t Elon’s original line about Tesla cars that they were essentially just a proof of concept to license tech to other manufacturers?
Okay maybe if Elon was right we should have self driving cars we could fly to the moon. Of course we don’t. But let’s be honest without him we wouldn’t have one decent EV in the world.
I fail to see your point. This unintended consequence doesn’t make him trustworthy or honest.
Yeah, I still think there would be decent EVs sans Musky. Both BMW and Nissan make good efforts that I firmly believe were decent. What he did was push range standards, albeit with some real optimistic math and marketing. However, that would have happened eventually with or without him. I personally would prefer without.
And he didn’t even originate the tech nor found the company. The mythology he came up with about himself helped sell it and get funding to build it up and keep Tesla going through long years of losses. I think the tech would have sold itself at some point, but I believe he pushed the schedule ahead at least a few years, if not a decade.
Yeah, Musk’s contribution here has been his relentless drive to get enough money (investors + customers) to make mass market electric cars mainstream. He’s always been an exploitative egomaniacal asshole, but he does deserve some credit for dragging the whole industry from EVs being a cute extracurricular side project to pacify environmentalists to a legit market segment with at least one profitable EV-only major manufacturer in the US.
My theory is that he took too many stimulants and other drugs to keep up with his self-imposed workload at Tesla and SpaceX that he either fried some of his brain’s regulating mechanisms or got too addicted to quit, and now he’s turned into a full-time mini-Hitler who thinks 14-word tweets are funny.
He was always an absolute turd of a human being, but something changed substantially in the past 5-6 years.
Sorry. We lizard people forgot that there is side effects to our “suggestions.”
The EU wrote their CO2 mandates in 2007 – a year before the Tesla roadster when into production.
2015 – 150 g/km
2020 – 95 g/km
2025 – 81 g/km
2030 – 43 g/km
2035 – 0 g/km
(NEDC numbers)
A 2nd gen Toyota Prius that was on sale when the EU adopted those rules was certified at 104 g/km so it was very clear at the time that the rules required automakers to transition to EVs.
There’s a big difference between gradually achieving these numbers in Europe on the one hand, and on the other making America want EVs and getting the investment money and execution to make it happen. Look how many EV companies have failed.
Again, Elon is an evil mfer I might push into traffic if I ever saw him on the street, and he didn’t found the company, but Tesla being a profitable major car maker is an accomplishment.
Yes, Tesla being a profitable car maker is an accomplishment. Starting up a car company from scratch is very hard. That doesn’t mean that Tesla is responsible for EV sales.
Above I have an example for Europe. Here in the USA CARB has similar numbers with hard percentages. China has more aggressive mandates then the EU.
Regulators around the world have mandated EVs – they will get built with or without Musk. Without Musk they likely wouldn’t have gullwing doors, fart noises, or be made of stainless steel.
Musk founded Tesla the same way Ray Kroc founded McDonalds
This connection dawned on me recently, and it needs to be mentioned more.
Mars. Haven’t you seen his hats and T-shirts?
Personally I think Tesla and Musk get too much credit for EVs. What Musk proved that people were willing to pay high prices for a performance EV. However, with or without Musk the reality of the EU’s fleet CO2 mandates and other regulations around the world still exist. It is simply becoming uneconomical to attempt to hit fuel economy and emission standards without EVs in the mix.
Then there is China – the largest automotive and EV market in the world where Tesla is a rounding error.
Con man gets called out by someone in the know. Good.
As if that matters one whit to lord lies a lot and his new bestest friend capt grift.
It’s super-hard to figure out which one is which, when you put it like that.
Ones older and doesn’t speak like a drunk 14 year old.
I wrote more and then erased it. I promised DT to try to be less political, less divisive, and more civil. It’s not always easy, but today I did better than yesterday. So, progress…
I promised myself to dial it back but the just keep pulling me back in.
SO MUCH of what Elon shows as groundbreaking tends to be utter bs or smoke and mirrors. Show an actual robotics expert Optimus and they fall over laughing, for example.
The sole exception is SpaceX, which he has the least involvement in. They truly are groundbreaking and FAR ahead their competitors.
Gywnne Shotwell, that’s all I’m gonna say.