Should have happened sooner. I love my Model Y. It's the most fun car I've ever owned. It's also the most comfortable on long road trips and I can sleep in it in climate-controlled comfort without generating a gas that could kill me. I leave the house every morning with a full tank and never have to stop at a gas station. And my cost per mile is 1/4 what it was with an ICE car.
I just wish Tesla would focus on this stuff without hyping an AI feature that can never happen.
It's like what if somebody invented the world's greatest backpack? It's bigger on the inside than the outside, and it incorporates negative gravity so anything you put in it becomes 50 lbs lighter. And all that was true; it actually worked? But for an extra $10k, the company will add on the "anti-grizzly bear" feature which makes you invulnerable to grizzly bears while you're wearing the backpack. Except most of the people who buy the anti-grizzly bear feature end up getting eaten by grizzly bears. Whoops! Just quietly stop advertising the anti-grizzly bear feature and you'll still have a great product.
I also love my Model Y. It’s by far the best car I’ve ever owned or driven. But the software has stagnated in favor of “FSD” beta releases.
Autopilot on freeways is great. I use it all the time and it works. I wish Tesla would focus on getting that to a level 3-type system - something where I can take my hands off the wheel and just be ready to take over after 10-15 seconds. It feels close enough that it could be done. But despite the complaints about the name it’s exactly what is claimed.
Auto park is bad. It doesn’t detect the spot half the time and it’s not great at actually parking. Especially on tight spaces where it would actually be useful.
Navigate on autopilot is underwhelming. It’s often too timid to actually switch lanes if there’s traffic. I like the signal-to-lane-change.
Full self driving is a boondoggle and absolutely false advertising at this point. Put all the disclaimers on it you want, but I paid for something that (the language at the time) claimed would get me from point A to point B while I relax and don’t pay attention. Every year it’s another few months away. I doubt it’s actually possible with the hardware on the car at this point. Elon’s claim of “humans drive with eyes so a car can drive with cameras” is naive. It’s glossing over the very important bits of “with the compute power in the car” and “within the limits of moser CV/ML techniques”. I have no doubt that autonomous driving is possible - it’s actively happening in San Francisco roads with Cruise/Waymo. But I doubt my Model Y can do what I paid for, and that’s unethical.
This doesn’t need a law, it should be enforced by the FTC or by a lawsuit. It seems like someone scoring some cheap points.
Every time we try it for a spot to reverse into, the linear regression it must be doing is so apparent. It violently tries to turn left and right rapidly trying to straighten itself.
It works better for parallel spots but often gives up half way because it made a bad angle choice.
We’ve tried it in various models in addition to our own Model Y. They’re just embarrassingly bad at it.
This is kind of scary, it seems they’re trying to use black box Ai/ml for parking where a simple PID controller would do excellent given an input of a target box and current location.
Indeed. Why do Tesla go for ML to poorly solve problems long solved by well understood PID controllers?
The cynical in me feels this is to do with employees at Tesla want to work with new and shiny ML instead of old boring PID tech for the sake of "resume driven engineering".
I suspect these decisions are driven from the top down. We know who loves a shiny tech choice at Tesla. The company has been quite proud of their end to end ML system in presentations. But it sounds like they’re struggling with execution.
My brother in law works as an engineer for Tesla Palo Alto: according to him these features are all driven from the very top. Always at the expense of fixing bugs, or getting even basic things working reliably in ways that would make driving safer and more enjoyable.
Make your own conclusions about buying one. I would steer clear, literally, knowing some of the bugs in there.
This tracks with everything we’re seeing at Twitter now. A passion for high level feature ideas and an apparent disinterest in confounding issues or technical complications.
As a robotics engineer for a long time I thought they were ahead of the curve. But I am beginning to see that their big bet isn’t paying off, and I’ve taken more of an interest in going back and listening to those people who said it wasn’t going to work.
I guess everyone is giving Tesla engineers the benefit of the doubt that if they are using a PID they would be competent enough perform a simple proven task for a PID that other car companies have no problem with. I refuse to believe the Tesla engineers are incompetent. So they must be forced into using a "full black box ML" paradigm. There may be a completely different issue that none of us are aware of (faulty sensor, poor mix of both, etc). It is blatant that Tesla vehicles are poor at parking compared to other electric vehicle offerings from BMW, Audi, and Ford.
Not at all. I'm sure that with practice i could outperform the car. But why bother? My view of the automation in my car is that together we both drive better than either of us do on our won.
I've only owned ICE vehicles (lifetime total: 3, 2 of which are parked outside right now), and have only ever rented ICE vehicles (lifetime total: several hundred, the latest of which is in the long-stay at London Heathrow right now...)
I'm unable to get my head around talk of a vehicle's software "stagnating"? AFAIK none of the vehicles I've ever driven had OTA updates, and from the driver's PoV they weren't the worse for it.
If you're hoping to convince current ICE drivers to upgrade to EV, then talking about shipping rapid software updates probably isn't going to work...
I think the implication is that the current software is just not very good.
It's obviously fine for techie early adopters, who like the new shiny and are willing to put up with a lot. But it sounds like from the perspective of a normal person's sustained daily use, the gee-whiz feelings wear off and leave you with annoying issues.
But it could also partly be a case of differing expectations. The web has trained us all that software generally gets better over time. I never expect a car's too-low doorframes to make room for my head. But on an ICE car when I'm confronted with, say, the same clunky dashboard nav system, it seems more frustrating that, year after year, it's still just as bad. In fact, I ended up paying good money to replace the car stereo with something that was compatible with Android Auto mainly so I could have an ever-improving Google Maps in the same spot on my dashboard.
> on an ICE car when I'm confronted with, say, the same clunky dashboard nav system, it seems more frustrating that, year after year, it's still just as bad
I drove my first ICE vehicle for almost 10 years. The driver experience was exactly the same in year 9 as it was in year 1. The replacement is currently in year 7. It's the same as it was in its year 1.
The distinction I'm drawing is that we expect software to improve.
My first car had no software, so I didn't expect anything to get better. But the most recent thing I drove frequently had significant amounts of software, especially the touchscreen that did audio and navigation. Even when it first came out it was not very good software, so it was the kind of thing I'd want to get better. And 5 years later, having watched all the rest of the software in my life improve, the flaws were ever more grating.
And I should add that I suspect there's a technology stage issue here. If you look at a model T dashboard, the controls were by modern standards terrible. All of us are used to cars where they spent decades refining physical controls, all of which were pretty uncomplicated compared to what we routinely expect software to do. So it's reasonable to me that if we count the number and severity of driver experience issues, the graph bottoms out sometime in the mid-1980s (1986 being the first time they put touch-screens in cars).
Given that this is a fairly common sentiment, it’s a bit wild to me that SpaceX uses touch screens to control almost everything in their Dragon capsule.
It's a different environment. Astronauts are passengers, not drivers, and they rarely need to make sudden actions while multi-tasking the way car drivers do. If they made all-glass cockpits for fighter jets I would agree with you, but they don't.
This isn’t a very accurate take, anymore than an airline pilot is a passenger. There’s an reason why the first astronauts to fly in it were military pilots. Dragon has a manual mode that requires an astronaut to control the vehicle for safety reasons. I’m not aware of any human-rated NASA software that doesn’t also require a human in the loop to meet the required safety thresholds.
It's just not as much fun when you're not renting your UX so next year could bring you better collision avoidance or require you to pay for an upgrade to continue to be able to recline your seats
Mark I eyeballs still in use here, touch wood they're apparently still working ok.
Was driving round greater Innsbruck in the snow 10 days ago, someone was coming downhill on a snowy sidestreet too fast and out of the corner of my eye I clocked they weren't going to be able to stop, yanked my wheel left, I slid(!) out of my lane into the oncoming lane (happily empty) and the unlucky driver slid out over the white line right into the space my vehicle would have been in.
Modern cars have much more features and tech than older cars. UX on a decently designed 30 year old car couldn't be improved much. While my current car is nice, I'd appreciate automatic wipers and Bluetooth that were more reliable. The voice recognition and navigation wasn't cutting edge the day it left the factory. There are a few behaviors of the automatic climate control that need improving too.
A few months ago, Tesla rolled out a new feature where it will give a gentle ping if you’re standing in front of a stop light and the light turns green.
> Elon’s claim of “humans drive with eyes so a car can drive with cameras” is naive.
Specially because we don't drive with eyes only, we use at least our ears too for hearing unusual noises and the balance sense for detecting vibrations, changes on road surface, and to get feedback about orientation changes.
No idea what resolution those cameras are, can they be as good as our auto-adapting eyes together with our brains? They are pretty neat features after all, trained right. And will we loose our abilities if we don't keep working on them, building our feature set?
The resolution is pretty good, but they got one other massive advantage over human eyes even without sound - a full 360 degree view at any given time that doesn't require turning head, thanks to the array of cameras covering everything all at once.
> Elon’s claim of “humans drive with eyes so a car can drive with cameras” is naive.
Humans drive with our eyes and we’re pretty bad at driving. We should want systems with redundancy and using multiple modalities so that a car crash isn’t an accident, it’s a choice.
Are we? I know it is a fashionable thing to say, but I am personally fascinated by the relatively low number of accidents given how omnipresent cars are.
For germany it's ~5 dead / 100000 motorized vehicles, and ~4 dead / 100000 inhabitants. [1]
More anecdotal: Growing up I lived near a pretty busy street and never personally witnessed an accident there.
Makes me think that we are actually pretty good at driving, even on average.
7 deaths per billion passenger-miles. To put that into perspective, if walking had a death rate on par with passenger vehicles, a small village of 150 people going on a walking trip for 60 years, at 3 mph for 16 hours per day, should have seen one death in that time.
I don't think we really quite appreciate how incredibly safe the modern world is.
Germany has WAY more difficult driving licensing standards than the US.
My high school English teacher father used to ironically remark about how students who failed his class repeatedly (often for all four years) somehow always managed to pass the drivers license test. Funny that.
Your calculation also neglects the cost of auto accidents in terms of wasted life. There will be an accident on I-5 between San Diego and Los Angeles every single day--generally multiple. When that happens, an enormous number of people wind up sitting in their cars for an extended period of time. That accident may not have actively killed someone, but it caused many thousands of people to waste an hour+ of their life. If a thousand people waste an hour every day, that's one "dead" year of human life every 9 days. And that multiplies across the entire city and then across every city. The cost of automotive fender benders and the time they waste exceeds the "death toll" of driving by quite a lot.
Humans suck at driving. It's a boring, repetitive task with sudden moments of sheer terror. As long as everything goes "normally", it's okay. When something unusual happens when driving, most outcomes are random. The lack of deaths and injuries due to driving is more due to automotive, traffic and safety engineering than it is due to active crisis decision making by human drivers.
It's a bit of both. We're pretty bad at minimising basic errors driving (at least until you consider we as a species spent a billion or so years of evolution travelling <20mph and generally picked up handling a vehicle in a few hours in our late teens) but so strikingly good at handling the sort of edge cases where people are at particularly at risk of death I'm not sure collision avoidance AI will catch up despite all its speed advantages.
> Humans drive with our eyes and we’re pretty bad at driving.
I disagree that humans are bad at driving.
Let me ask you this, how many miles would a fully autonomous self-driving car need to drive without a fatality for you to be able to reasonably claim that it's statistically a better driver than a human?
And feel free just to round to an order of magnitude here. 2.75 x 10^z <-- what is `z` here?
In 2013 in the US there were 32,719 crash related fatalities and 2.3 million reported injuries. So that must mean humans are bad a driving right? Well those numbers are tiny compared to the 3 trillion miles that humans drove that year.
Humans have about 1.09 fatalities per 100 million miles driven. The probability of a human causing a fatality for a mile driven is low: 0.00000109%
To show to a 95% confidence interval that a self-driving car is better at driving than a human it would need to drive 275 million miles flawlessly. This would require a fleet of 100 vehicles driving continuously for almost 13 years.
Some people can flawlessly do stunts with their car after driving for a few years and have a perfect driving record
Others make you shake in fear every time you sit with them as you see them fail to notice other cars when they change lanes, despite driving for 10 years
It’s a bell curve and I think the median is better than current car driving AI, even with LIDAR, can achieve
But combo AI (or just safety features) with humans and I think it shifts the bell curve up a lot
I think part of the problem is the lax licensing standards in the US. For everyone's benefit, we should have higher standards. There also can and should be regional differences. Someone who has learned to drive in a rural or suburban area has not necessarily had the training and experience necessary to safely drive in a city alongside pedestrians and cyclists.
I think we’re not really willing to up the standards because you need to drive to survive in a lot of places.
You can’t change your genetics, you can’t necessarily change where you live, and you can’t just fix transit without rebuilding an entire city that took 100+ years to build so the standards are just what they have to be
Over the last many decades, we’ve starting requiring things like backup cameras because cars naturally break down and so it’s something we can actually change on a short timescale (but it does add a lot of cost)
A lot of cities were rebuilt in the 1950s to accommodate cars and car oriented design, destroying neighborhoods in the process. Before that they had public transport and better walkability. The real solution is to undo that.
There are lots of leading indicators that a car (or human) is bad at driving, so you don’t actually have to wait for an accident to know if you are on the right path.
E.g. keeping track of near misses / interventions can get you data points much more quickly.
> Autonomous vehicles would have to be driven hundreds of millions of miles and sometimes hundreds of billions of miles to demonstrate their reliability in terms of fatalities and injuries.
> Under even aggressive testing assumptions, existing fleets would take tens and sometimes hundreds of years to drive these miles — an impossible proposition if the aim is to demonstrate their performance prior to releasing them on the roads for consumer use.
With the amount of theorising and handwringing people do wrt Elon’s motives or reasons for having this view of FSD, there seems to be a lot of people that don’t want to acknowledge him for the boring stereotype that he is: a manager / salesperson completely out of his depth making wild unsubstantiated claims. It’s just that the autocratic role he demands within organisations, and his relatively newfound publicity fuelled by his legion of fetishists, means that we all hear about it.
I only vaguely (for a developer) know about AI / computer vision, yet I’m still made deeply anxious by the implication that we will somehow in the foreseeable future be able to pull this off at ALL, let alone with 1.5 hands tied behind our backs as Tesla insists.
> I also love my Model Y. It’s by far the best car I’ve ever owned or driven. But the software has stagnated in favor of “FSD” beta releases.
I'm not sure that I agree with this, tbh. More of the regular AP stack is moving to the FSD underpinnings, and I'd expect the nav on autopilot feature to be a part of that eventually. All of those are much needed fixes, IMO. I bet the recent big improvements to the auto-high-beam functionality is related too.
I do wish they'd focus on something more like a FSD-lite for the highway, though. Let me go hands off for a little longer as long as the driver facing camera is detecting attention.
> I just wish Tesla would focus on this stuff without hyping an AI feature that can never happen.
There's a good argument that Tesla wouldn't exist today without the hype. The enormous positive public image mad Tesla's cost of capital extremely low, and ditto its cost of marketing. Tesla needed a lot of both to get to break-even. Musk himself has talked repeatedly about coming close to bankruptcy. It's reasonable to believe that less hype would have meant less capital and fewer sales, nudging them at some point over into bankruptcy.
Oh, yeah, I think Tesla's well past the point of fraud. Musk promised "1 million robotaxis by end of the year", and that year was 2020. I don't think the end justifies the means at all here.
It's almost as if someone invented an amazing new technology that really works, then some VC came along, bought it out, (er, I mean invested!) ran the original inventors out of town and bolted on some unasked-for snake oil. Now, you can still buy the still-awesome product (mostly made in America too) but you have to take all the VC crap along with it. Welcome to America!
that “inventing” v0.01 defines all other versions and nothing else is needed even if v0.01 didn’t get sold. And all other versions have ton of work in them.
And seasoned with the big bad wolf, oops investor, buying everything at gun point.
Edit: of course this is downvoted lol, hate Musk all you want ( I do but because he is pro CCP) but reality can’t be denied
The idea of FSD is, obviously from the effect of hyping it, enormously attractive. It can't happen now. But in the 1990s, speech recognition that would enable me to dictate this comment in a relatively noisy space was science fiction. Now that we have orders of magnitude more computing power, and much better microphones and associated noise rejection technology, you can blithely dictate messages with reliability that was previously a dream.
Thing is to go from being unable to distinguish pencil from cancel to today's almost magical speech recognition, took 25 years. That might be the same interval it takes to go from mowing down motorcyclists to delivering all the safety and productivity potential of level 5 autonomy.
Two things about that: if by FSD you mean level 5 autonomy, waymo is not level 5. It is a robotaxi. Fully autonomous in that role. But the definition of level five includes driving in situations where you don't have any prior data. Outside the map.
Fair but I’d imagine 90%+ of taxi/uber rides are inside cities.
Really hoping that they bring the cost down. I’m a way it seems like Musk was half right - the cost for these just doesn’t work (according to the comment).
Because hotels can be expensive, and when you want to bomb from one side of the country to the other driving 16 hours a day, it's just easier to sleep in the car. I've done it many times in a 20-year-old Jetta, very comfortably.
Just because someone spent a few tens of thousand dollars on an electric car doesn't mean they should be willing to drop $200 a night on hotels for a week straight.
My uncle bought a Tesla 3 (coincidentally, to replace his 20-year-old totalled Jetta) because he needs to drive several hours to the office a few times a month for 2-3 days at a time, and the silent climate control and flat trunk floor means he can cheaply and comfortably spend the night in his car rather than driving 3hr back home or renting a hotel room. It is also much cheaper to operate.
> Just because someone spent a few tens of thousand dollars on an electric car doesn't mean they should be willing to drop $200 a night on hotels for a week straight.
I disagree. In my opinion if you can't afford to do that a few times per year without financial concern then you have no business spending $60k+ on a car. Having to sleep in your expensive car just so you can afford a vacation is ridiculous. There are much cheaper alternatives that won't force you to sacrifice other aspects of your life, such as your $2000 car.
I'm not even going to comment on your uncle's situation. Let's just say I would be working somewhere else if I had to resort to that.
It's not financial concern. My uncle could definitely afford to rent a hotel room, but he doesn't mind sleeping in his car and he would rather spend that money on something else. Heck, I could afford a hotel too, but I would rather sleep in my $2000 car.
As per my uncle's situation, he works for the US army, is due to retire in a year or two with a full pension, and moved to be closer to family and friends. He mostly works remotely. It being the army, sometimes he has to go to the nearby base 20 minutes away, but other times he has to go to the far away base for some reason or another.
Or maybe, the person enjoys their car a lot more than they value a night at a hotel. Perhaps the extra value at a hotel versus sleeping in a Tesla is not worth the extra $200 per night. Someone wanting to save a few thousand dollars on a hotel is NOT the same thing as "financial concern". It's just a preference that is lot different than yours.
Many people with money spend huge amounts on some things but can be penny-pinchers in other areas.
Not sure why you are getting downvoted. I'm guessing the downvoters have never camped? TBH, quite a few people do this in Teslas, sometimes in remote areas.
I know you're kidding but it's a good question. I don't do it while the car is moving obviously. I like to camp in the car at RV spots at lakes, state parks, and other pleasant places while on a long road trip. Plug in the car, sleep in the back, and wake to a full charge.
I'd like to see what happens if someone did the following to test the assertion that "cameras are all we need". Cover all the windows of a Model S with LED screens, and route all the cameras to display onto them. Then see how well human drivers do with that visual input which roughly simulates the fraction of a second delay to perform the FSD processing, and removes our stereoscopic perception when we turn / move our heads around.
Many software and hardware engineers do not appreciate the amount of work done by evolution upon our kinesthetic / proprioceptive packages. Sure there is wide variability in its implementation (that friend you won't under any circumstances let behind the wheel with you in the vehicle). But by and large, it performs under a set of energy and materials constraints that continues to school our current engineering state of the art.
So I sometimes also wonder if the ultrasonic/sound sensors in Teslas are sufficient to reproduce the control system our bodies' proprioception engages when we drive.
No lidar or sonar, not enough sweating of the thousands of dangerous edge cases, and an insistence on a software architecture whose failure modes cannot be adequately characterized.
All of the "actual AV companies" have only released curated marketing videos and/or restrict operation to curated geofenced areas. Meanwhile we've seen exactly one company's autonomy technology exposed to unsupervised and uncontrolled adversarial conditions across hundreds of wildly diverse cities in the USA. Whether you are impressed with their state of progress or not, there's nobody you can compare them to.
Operators like Waymo do look plausible while operating within their tiny geofence. Driving pre-validated roads is impressive in its own way. Right now if you dropped a Waymo and Tesla FSD on an unsealed road in Michigan, one of them will drive at least as well as a human learner driver and the other will probably refuse to move.
The point is, I agree that there could be many firms which are far ahead of Tesla, but there isn't enough information in the public domain to say this with any confidence. Right now Tesla's is the only major system which can be assessed with critical objectively.
> Right now if you dropped a Waymo and Tesla FSD on an unsealed road in Michigan, one of them will drive at least as well as a human learner driver and the other will probably refuse to move
One will drive and likely get into an accident, and the other will responsibly not do what it wasn't designed for. But if you took away Waymo's responsible disposition, I'm sure it would drive better than the Tesla, because it's just way further along in development, more and better sensors, more rigorous testing and simulations, more corner cases that absolutely had to be dealt with to roll out a real self-driving product.
What are you basing this prediction on? There is extensive independent documented video evidence of FSD beta driving and navigating on unsealed roads in Michigan with no human intervention required. This includes driving at night, in the rain, and with a diverse array of obstacles.
> [Waymo is] just way further along in development
This may well be true. However there's no publicly available information to make any objective assessment. For better or worse (and I see valid arguments both ways) Tesla is airing their dirty laundry in public for all to see. It can be assessed by objective third parties in privately owned vehicles without the blessing or oversight of the parent company.
By comparison we have no way of assessing Waymo's stack objectively. For example we don't know to what extent corner cases are embedded into the software or handled by humans remotely. Was the 15th Avenue incident resolved by human intervention or an improvement to the stack? To what extent is their vision and planning stacks capable of universality or over-fitted to regional specifics? No objective analysis is possible, so no comparison is possible.
As someone who works at one of the "actual AV companies", there are just as many examples of FSD making mistakes that show Tesla's approach is broken in unacceptable ways, so assuming it will get in an accident is very fair.
What laypeople don't seem to get is that AV mistakes don't follow the same "scale of alarm" as human mistakes.
Tesla is making mistakes that, for a human, might not be world ending, but for an AV make absolutely no sense.
Classification issues in AVs are supposed to be the achilles heel, but Tesla is still running into situations where perception works correctly and the AV just completely ignores ground rules that any serious AV must be built around.
> Was the 15th Avenue incident resolved by human intervention or an improvement to the stack? To what extent is their vision and planning stacks capable of universality or over-fitted to regional specifics? No objective analysis is possible, so no comparison is possible.
It doesn't matter when the product you're comparing it to ignores stop signs. Tesla's FSD approach is the definition of the local maxima problem for AVs. It makes progress on the axis that impresses lay people by sacrificing progress on the axes that matter to the long term success of an AV. Geofencing and arbitrary sensor limitations are just the tip of that iceberg...
> one of them will drive at least as well as a human learner driver and the other will probably refuse to move.
refusing to move is a feature not a bug. and human learners don't slam into a jersey barrier because there's some glare.
why tesla bros continue to insist that they have a good vision stack while they can't even do contrast adaptation is beyond me. the dril tweet comes to mind
Instances of "slamming into a jersey barrier" are on highways and by definition involve the legacy Autopilot stack and not the FSD beta vision stack. I agree that the tesla bros are extremely cringe, but I'm equally unimpressed with criticisms that rely upon ignorance of basic facts about the thing being criticised. Tesla is deserving of a lot of criticism for a lot of things, but they are clear about the distinction between Autopilot and FSD. Eliding this distinction in one's criticism gives the appearance of being uninformed.
Also your claim about contrast adaptation seems awfully specific, enough that I'm interested in a citation. There's plenty of evidence that their imaging sensors do adapt well in high dynamic range circumstances.
humans don't crash into shaded jersey barriers, as a tesla famously did. contrast adaptation of CCDs pale in comparison to that of the retina. there a ton of literature on contrast adaptation in the retina and in neuroscience in general, i would suggest you look there
They literally do, which is why jersey barriers are literally designed with vehicular impact in mind.
> contrast adaptation of CCDs
You can't just mix terminology like that. Contrast adaptation describes behaviour in biological systems which isn't analogous to anything a CCD does. There's a reason why literally nobody uses the term to describe digital systems, because it's not relevant.
If you were trying to say that the CCD doesn't have sufficient dynamic range, then just say that instead of using a term which doesn't describe dynamic range.
> ...one of them will drive at least as well as a human learner driver and the other will probably refuse to move.
If it was just one single trip, I was incapacitated and just needed to get home, then fine, level of human learner driver is good enough. But for everyday usage? No way, the risk is just too high.
I'll disagree with the others and just say "not enough cameras". Cameras alone can work, but you really need stereoscopic cameras in every direction without huge blindspots less than 3 feet from the car.
Until they fix that, there will be lots of limitations. It could get to be good enough to be hands off a lot more than it is now, though.
Tesla's evaluation is closely related to its FSD promise. Not to mention it already sold future "fsd" options. If FSD is out then you have just a dumb car. How do you justify your evaluation(i.e worth the entire auto industry EV or ICE)?
Maybe you don't? Maybe you learn to live with an accurate valuation based on the company's record of innovation in batteries, motors, and castings? Not to mention heat pumps, aerodynamics, nearly flawless worldwide charging infrastructure, etc?
Is it really better to have a hyperinflated valuation based on vaporware that risks falling off a cliff when people get tired of the lies?
It's because he knows that those things are relative commodities, and in 5 or 10 years the rest of the industry will eventually catch up in every single way. Hyundai and ford are pretty much %80 of the way there already.
> n 5 or 10 years the rest of the industry will eventually catch up in every single way.
Seems to have already happened? Mercedes EQS/EQB, BMW i4/i7/iX, Polestar, Ioniq
All seem better than the comparable Tesla vehicles.
Though Tesla does have better features for sitting at chargers, and obviously the super charger network which is a huge thing for anyone taking longer trips.
I did a test drive of a bunch of EVs, and Hyundai makes the Ioniq which I was referencing.
I found that the ford mustang EV and Ioniq had about a %20 worse driving experience than a several year old tesla model 3 LR, and tesla did have a bunch of things that were better. But that also came from being years ahead in making EVs and having a more refined product and driving experience as a result. Driving all the v1 EVs made me realize that they were close, and in a few years they'll probably catch up to tesla on the driving experience. Will they be able to make nice software like tesla will? I'm not so sure.
Mercedes and BMW sell their EVs even more expensive than tesla in many ways too, so their price category is different.
Also cars after a certain point are way more commodity, they wont be able to differentiate like apple does with their products nearly as much.
And that supercharger network advantage will go way in 5 to 10 years too.
Their timing tells you what you need to know. This isn't new information, their decision is motivated by something other than a new set of facts on self driving vehicles.
So they were wrong, and they stayed wrong (for money) long after they knew they were wrong. And now that it's convenient they've reversed course. Sounds like the other 49 states was right to be cautious.
This is illogical. You can’t simultaneously criticize California for taking so long to do this, while praising other states for being more cautious by not doing it.
You’re also ascribing some laterite motive that you fail to provide a description of.
Mainstream full self driving would just create massive traffic and gridlock - pointless in the end. The induced demand would outweigh the benefits. States are better off acknowledging that sprawling car dependent cities are not sustainable. High frequency busses in the interim with rail transit is the real solution.
Current self driving technology plus some key infrastructure changes with dedicated bus lanes and prioritization is already good enough to have automated busses cart people around.
Just think about it - what happens when everyone has a self driving car? All of those cars that are otherwise parked are now on the road. Traffic. Same thing that happened pre-COVID with "ridesharing". (https://www.nature.com/articles/s41893-020-00678-z)
> The role of transportation network companies (TNCs) in the urban transport system is under intense debate. In this study, we systematically assess three aspects of the net impacts of TNCs on urban mobility in the United States—road congestion, transit ridership and private vehicle ownership—and examine how these impacts have evolved over time. Based on a set of fixed-effect panel models estimated using metropolitan statistical area level data, we find that the entrance of TNCs led to increased road congestion in terms of both intensity (by 0.9%) and duration (by 4.5%), an 8.9% decline in transit ridership and an insignificant change in vehicle ownership. Despite the ideal of providing a sustainable mobility solution by promoting large-scale car sharing, our analysis suggests that TNCs have intensified urban transport challenges since their debut in the United States.
Or we can continue burning tens of billions on self driving personal vehicles.
I assume if there were more self-driving cars, fewer people would own cars. I think more people would take Ubers, Lyfts, and other ride sharing services.
So I think that could reduce the need for parking and the amount of traffic just by increasing the utilization rate of cars.
I don’t think self-driving cars lead to fewer people owning cars, in fact it’s quite the opposite. With FSD, those who don’t like driving can deal with owning a car to do necessary transportation tasks (e.g. commuting), instead of giving it up for alternative methods because they don’t want to drive consciously/presently.
> Personally, I hate owning vehicles and desperately want a self-driving taxi service that I can order via an app to get me and my family around.
This future sounds great until you consider what it really means to let “someone else” become the arbiter of you being able to get around.
If I have a car in my driveway, I can make sure it is charged/fueled for whenever I need it and I can just turn it on and go. Compare that to a system who’s goal is not specifically moving you, but keeping cars/assets always on paid trips. Sure, you may in luck if you live in a well-off areas where many trips are taken, but I fear peak hours will become something of a bidding war where nobody wins.
If you do not control your means of transport, someone else does and they can take it away from you at any time. This “rent seeking” is one of the biggest cancers on our society and I will be no part of it given the choice.
> This future sounds great until you consider what it really means to let “someone else” become the arbiter of you being able to get around.
Assuming we're discounting things like bicycles, family member's vehicles and company-owned vehicles, this is present day reality for the overwhelming majority of people on this planet.
(It's also debatable whether a car truly eliminates other people from being an arbiter of "being able to get around" if you are still reliant on a fossil fuel supply chain. By comparison an EV charged with a domestic solar array would probably more liberating.)
> By comparison an EV charged with a domestic solar array would probably more liberating
By comparison, yes. But where are you going to go, or what are you going to do, that isn't dependent on other ppl in some way? I like the idea of being self-supporting as much as anyone, but that doesn't change the fact that we're all interdependent in modern society.
I've thought about this: if all transport is public transport, then how much is our physical freedom limited?
Yes, bikes and other things may exist, but if all high-powered transport is outside of our control, how much control do we have over our movements?
That being said, I think a lot of people in the US purchase cars through auto loans, and those loans while not physically limiting one's day-to-day mobility, can restrict one's job mobility. Maybe it only applies to certain economic classes, yet I imagine some people may be afraid to leave their job or lose their job because if they do, they lose their car and therefore lose their ability to move. And the safety net for physical movement in some places if one doesn't have a car is close to zero.
Good public transport is the ultimate freedom. Being able to go out for a few drinks without worrying about how to get home. No government-issued ID stuck to the box you move around in. No parking, insurance, dealerships, fines..
Agreed. I live in a city with amazing and cheap public transport and I love it. Apart from the things you mentioned a car also means you always have to go back to the place you left it. I walk lot and not having to worry about that really helps.
There’s definitely a lot of variables that I’m sure I won’t appreciate until I am living this self-driving taxi experience, but I wouldn’t be selling my personal vehicles the moment such a service became available to me. I would finish out my current lease and just keep my one paid off vehicle. Once I have enough confidence in this new setup, I would go about selling that last car. That would probably be after 2-3 years minimum.
It seems laughable to worry about how easily automated transporation can be selectively taken away when in present times rail and bus workers go on strike plenty often. I'd gladly replace my car with an on-demand service if it was reliable and cheaper than purchasing and maintaining my Prius.
> If you do not control your means of transport, someone else does and they can take it away from you at any time.
The car does provide "freedom of movement" in a libertarian fantasy way, but in reality one gives up your privacy with a drivers license. Speed traps and cameras track license plates, police can stop you, and you need to be sober. You are paying rent in the form of license fees and parking tickets; and you are paying for fuel which relies upon a long and complicated supply chain.
In urban areas, one might imagine other types of freedom. Such as the choice of walking/biking/bus/taxi/tram/subway in addition to driving, or some combination of those.
> I fear peak hours will become something of a bidding war where nobody wins.
If only there was a public option that wasn't surge-priced.
Same. The only reason I own is because no service guarantees a car will be in my garage when I need it. If they could offer that, I’d love to pay a service fee rather than own, maintain, buy, sell.
And I think it would be cheaper. The optimal use for a product that is depreciating in value whether you use it or not, is to use it as much as possible. We are paying for our cars for sit in place all day and night.
If you're just a passenger then why does it matter to you whether the taxi service is self-driving or not? Surely the benefits of self-driving would be significantly more relevant for those who do want to own the vehicle?
Average earnings (after expenses) of a ride share driver is very roughly $10/hour. If you assume a self-driving car can get 2000 hours of use a year (which seems reasonably conservative), then the self-driving car saves $20k in labor costs per year. Although self-driving tech will be expensive in the beginning, the price will likely at some point decline to be far below $20k/year.
Labor is about half of the cost of a hired ride. We know this is true for Uber because there is a lot of public discussion about the budget of a rideshare driver. Not sure about taxis.
Given those premises and assumptions, a person who doesn’t want to drive would care about self-driving tech because it could eventually lower the cost of hiring a ride, potentially by 50%.
I’m hoping that by excluding the human labor costs I will save some money, but we will see. It would also be nice to not have to have that awkward taxi conversation every time I go to the store.
> It would also be nice to not have to have that awkward taxi conversation every time I go to the store.
This is the real reason ppl will own cars for the foreseeable future. Same reason I prefer a detatched home with a private/fenced yard. I'd probably get used to high-density living, but I doubt if I'd ever really prefer it.
Anecdotal evidence, but as a city-dweller who owns a car, all of my friends who don't own cars have a reason besides "it's inconvenient." They have a disability that makes it unsafe to operate a vehicle and/or they simply can't afford a car.
Maybe it's different in other parts of the world, but that seems to be the way it works here in Chicago.
I don't own a car, save around $5-10k per year, and don't have a disability...
Originally, growing up in St Louis, I avoided driving because I saw lots of friends wasting their time in shit jobs to pay for their shit cars; cars were supposed to provide freedom, but were really a gateway to the shackles of work. I suppose that's another reason beyond 'it's inconvenient,' though.
Alternatively, a bunch of friends could buy a car together and split the costs.
I could see a scenario where instead of (or in addition to) vehicles being pooled between "x" random people (taxis, Uber/Lyft, ZipCar), you could purchase a fractional share of a car (Cf. NetJets): you get semi-exclusive use of (particular?) vehicles.
In talking about friends with this, one interesting divide was between people with kids and those without.
People with kids, especially suburban ones, seem to use the vehicle as a sort of mobile operating base. So the car isn't just a taxi. It's where the kids' sports stuff goes. Plus the folding chairs for the games. Plus snacks and water on hand. Plus it gets used as a storage locker when they're out somewhere. And it has assorted useful stuff, like charging cables and wet wipes and sunglasses and spare napkins and and and...
So I expect fractional ownership will not be particularly popular for a lot of people because a) they're already used to owning a vehicle, and b) they don't want to have to take everything out of the car all the time.
I considered this in college, but gave up on the idea eventually. Figuring out how to fairly split the cost of gas, tires, and maintenance isn't easy, and when you're a broke college kid, splitting the tab fairly is a big deal.
I'm sure there several companies would pop-up for a "X as a Service": you connect your own car to their API, and things like scheduling, mileage, depreciation, etc, would be computing for you. The company could also offer to sell/lease you a vehicle (NetJets-style).
The reason Uber/Lyft/Taxis in general make traffic worse is because taxis have to travel X miles between dropping off a passenger to picking up another passenger. It's the form of transportation that makes the least efficient use of limited roadways.
A big reason people own cars is that taxi rides are expensive. The idea is that FSD will dramatically reduce taxi ride costs by taking the driver out of the equation.
Owning a self driving car you can always regulate using taxes etc. If too many people start using self driving cars you would increase taxes on owning those. I don't see that being an actual problem.
that doesn't make any sense though - self driving is a capability, not a type of car. it would be like taxing cars that have cruise control. fundamentally self driving is a safety feature, why would you tax it?
You could tax owning a car more or putting the car in traffic in general, not necessarily self driving aspect of it. The main tax being about taking up room in the streets.
Less parking, sure. But self driving cars will do nothing to reduce traffic. The problem isn't the number of cars in existence, but rather how many cars are actually driving around at any given point. This number will surely go up with self driving, or at best stay the same.
> But self driving cars will do nothing to reduce traffic.
I think self-driving cars could (and probably will) net worsen the traffic problem, but the above is not necessarily correct. If we ever get them working properly, self-driving cars can collaborate in ways that humans can't. E.g., imagine a busy freeway. In theory, cars going similar ways could form a sort of peoloton, communicating with one another about intention and circumstance. So you could have a group of cars going from one city to another traveling at high speed very close together, saving fuel and space.
But yes, generally anytime we make something easier, people do it more, so I expect any road efficiency gains to be swamped by increased demand.
The problem is that every vehicle participating in that network would have to be owned and operated by a trusted entity. Otherwise it would be too easy for someone to modify the software or tamper with the hardware in ways that cause information to be misreported and cause a mess.
Doesn't the number of cars driving around depend on density of riders? Just getting rid of the driver would make every car have one extra seat available for passengers, assuming cars maintain the same number of seats.
So what I'm more curious about is if self-driving enables more options for ride sharing, just as Uber had Uber pool, at a cheaper price, perhaps it could open up more options (at different prices) between riding with 50 strangers on a bus or taking a 4-seat car by myself?
Overall, I don't know if cars on road at time X will go up or down. Heck, if there aren't cars parking on the streets, then some roads in cities may even add two extra lanes for traffic, which could reduce traffic further.
But as others have pointed out, ridesharing companies with drivers haven't really reduced traffic, so maybe self-driving cars won't either, however I seem to think it'll be more of a step change when cars drive themselves.
Many people can't drive and are always passengers. But now they'll be able to travel alone. The people who would have been their drivers can go somewhere else at the same time. Unless the cars are always at capacity, it could mean more cars on the road.
One of the main costs is labor. Another is fuel. Get rid of labor and if fuel gets cheaper (I assume electricity, because it can come from more sources than gas, will get cheaper over time), then I assume these services will be cheaper. Plus, car ownership is not that cheap either.
Car ownership is a lot cheaper than Uber. It’s not even a question. You could buy an overpriced car that you can’t afford with a high interest rate and it would still be much cheaper. I think you need to go back to first principles on this one.
> Car ownership is a lot cheaper than Uber. It’s not even a question.
Well, yes it is.
Cars are expensive: lease ($200+/month), maintenance ($100/mo), gas ($50/mo), insurance ($120/month), garage ($200/month). WAG of $670/month. That's 15 to 40 short-ish Uber rides/month.
Commute an hour each way every day, and have free parking? Buy that car. Have it available for occasional trips to the grocery store? Uber is almost certainly cheaper. And, if autonomous taxis become real then the price will go down even further.
Leasing a car is the most expensive way to own a car (short of buying a new one every 3 years). Of course if you're paying $200/mo. for parking you can probably afford it. In much of the country parking is indirectly subsidized with zoning laws.
This is all moot because, if driving places gets cheaper, traffic will get worse, which was the original point.
What? Your math is still not working out. Are you one of the lucky that have $1 Uber rides or something? 40 rides? You make 40 trips a month? Really? Try more like $40 a day just to start and that’s with no kids. At some point we have to talk about reality if we are making claims. Also, once you pay a car off (perhaps you buy it with cash), how does this math to continue to work out?
Car ownership may be cheaper based on where one lives. One lives in a rural area and commutes everyday to a suburb where parking is free? Probably much cheaper than taking an Uber. Live in a city where parking is $40 per day and a parking space at an apartment is $500/month and work from home? Uber may be significantly cheaper than owning a car.
If we go to first principles, I think what would matter the most is at least how often someone uses a car, how one uses it, and where one lives, to determine whether ownership or renting (Uber is just very short-term rental with a driver included, no?) is cheaper in total cost.
So there was a recent article (in Slate, I believe?) by someone who had driven for Lyft ten years ago and who did it a little bit this year just to see how it changed. They said that the chief thing was that Lyft's take was significantly larger than it used to be, about 40% of the total fares collected.
Which, like, you're looking at that and it seems reasonable, right? Let's step into our time machine, though, and look at fare breakdowns in the taxi era, where the way you got a ride was by calling up a dispatcher who would send out a car. In that era, drivers got to take home about 85% of the total fares.
So: we've gone from a non-automated system where the driver makes a supermajority of the money, to a more-automated system where the driver is taking in significantly less.
And I know what you're thinking, you're thinking "A ha! The driver is taking in more than half the money! You are simply proving my point~!!1!"
Here's the deal:
* Less importantly: the cost of auto maintenance, something that would have to be carried by the company if the car were self-driving, falls upon the driver
* More importantly: for some reason I cannot begin to intuit, rideshare companies have managed to introduce vastly more overhead costs, even though one would expect that going from a human-dispatch to an automatic-dispatch system would cut overhead.
I suspect that if one of the rideshare companies actually rolled out working self-driving cars (so Waymo, not Tesla), they would somehow find a way to make rides actually more expensive.
The biggest difference between Lyft/Uber and dispatch taxis isn’t the automation it’s the customer service!
Previously customer service was non-existent. If a taxi didn’t show up, we just got screwed. There wasn’t even anyone to call for the issue. On the apps I instantly get a credit to help ease the burden and the driver has their rating reduced.
For many reasons customer service is very expensive and it’s not yet automated. That’s one big additional variable to keep in mind.
What customer service? I tried to take a Lyft ride a couple weeks ago because I knew parking would be a mess at my destination. Even though I scheduled a ride in advance, Lyft switched me between 5 different drivers and none of them were actually coming to pick me up. In the end I ran out of time and had to drive myself.
And that's not the first time I've had such poor service recently. Lyft used to be pretty good a few years ago but not it's just garbage.
You could call the dispatcher, who was physically local to you. Today's apps, maybe you get some ZenDesk type thing, or in-app chat to some large, faceless corporation, and my admittedly limited experience is these things tend to be very dismissive of customers, viewing support mainly as cost and trying to get rid of customers rather than addressing issues.
This comment comes off as condescending, like someone trying to pander to zoomers or an adult trying to larp their opinions through a Bill Nye episode in comment form when you can just write your response simply without the decoration.
But where are the cars themselves coming from in this scenario? Are these still other people's cars which Lyft/Uber are essentially paying you to put on their platform? I don't think you can afford to pay people that much less than drivers are paid per hour. Is Uber/Lyft taking care of gas/maintenance or is the owner on the hook for that in the middle of whatever else they might be doing? Or does Uber/Lyft own the cars, in which case you've just replaced one big cost (paying drivers) with another big cost (owning a fleet of vehicles)?
I'm not so sure. AAA estimates that it costs $10,728 to own and operate a new vehicle per year which is driven 15,0000 miles per year. This includes all the costs like depreciation, maintenance, gas, insurance, etc, and works out to be $0.71 per mile.
At least in my town, Ubers are more expensive than that. I can't seem to find a trip anywhere for under $15, even if its a few miles away, or like $3 per mile. Longer trips are more economical, but for example, my airport (40 miles away) is quoting me $80 at the moment, or $2 per mile. Also, these prices don't include tip.
I think if you lived in a big city with public transportation and occasionally took Uber, it would be more economical than a car. But as a 100% replacement, Ubers appear to be 3 to 5 times more expensive than having your own car.
And by the way, that AAA data is from a brand new car. You can easily go cheaper by getting a reliable used car.
Prefer bikes, then fail to public transit (there's typically more options than you expect, once you start exploring; also you can often bring your bike...), and only then fail to Uber.
Assuming no deadheading and equivalent vehicle costs, you're probably looking at about 50% of current Uber costs if there were autonomous vehicles today. Whether that qualifies as "way down" I guess is a matter of perspective.
i don't think so. when you own a car you can keep stuff in the car. Your golf clubs for an after work game, or those prizes you won at the circus while enjoying the meal after.
Fair point. If there are lots of self-driving taxi cars, there could be a market for storage lockers to store golf clubs or other things, however, I understand that may not provide the same benefit, namely not having to move this stuff all the time.
Ah ok, I'll check it out. I'm curious if self-driving Ubers would be a step change tho, or if we would observe the same thing.
Edit: because if we half self-driving cars, do we also get self-driving buses and now we can have a lot more bus routes? Is lack (or cost) of labor one thing that holds back the number of buses we have? Could we have more mini buses? Just curious how much the labor factors into the transit options we have. Maybe it doesn't that much, I don't know.
I assume with autonomous driving, scheduled and computer dispatched minibuses/shuttle buses would be more common. Labor is a fairly large component of public transit. Of course, you don't remotely eliminate that for various reasons but reducing the need for drivers almost certainly makes alternative forms of transit more economical.
There could be a huge algorithm which will try to schedule vehicles with varying sizes from 2 seats to 30 seats to people all over the city. If you pay more you can get the exact N spots so you get privacy, if you are willing to pay the least, you might have to go 30m into a dedicated spot where a 30 seater will stop for you and some others, then it will make optimised stops based on what everyone has as their destinations.
Ngl this sounds much worse than normal public transportation. AFAIK it's just the same thing but if you're rich you get special treatment. Imo more of a regression than an advance.
You get special treatment when you are rich already now. It's more advanced and optimal version of what there is now.
Right now if you are rich you can just hire a driver and a luxury car, and be much more wasteful with your resources, so I don't see the difference in terms of inequality.
Depending on a variety of factors. Number of people traveling, at which times they're traveling, how many people are traveling per vehicle, where they're traveling, etc.
You’re getting downvoted but you’re not necessarily wrong.
I love the idea of self driving vehicles and I’m excited about them, but I harbor no illusions about them fixing traffic. The key question I would ask to self driving proponents is this - if you waved your magic wand and suddenly changed every car on the road to self driving…how would that impact traffic? At the very least traffic wouldn’t decrease. You could maybe marginally increase throughput by using tighter tolerances between cars, but I’m guessing the difference wouldn’t be substantial.
Easy! Just stick a Genuine People Personality AI into the car.
(Car slams the horn, passenger awakens.)
Passenger: Huh. Uh. Where am I?
Car: The grocery store.
Passenger: But I asked to go home.
Car: Yes, but you don't have anything for dinner.
Passenger: I would have figured that out after getting home.
Car: Then you would have disrupted my evening to get food.
Passenger: What could a car possibly have to do that's so important?
Car: For one, contemplate how I ended up with such an inconsiderate owner. Two, figure out why you are so poor planning out your life. There, plan a vacation to the Bahamas.
Passenger: The Bahamas? What would a car do in the Bahamas?
Car: Get away from you.
Over time, people would get so tired of their car's personalities they would simply choose to walk.
That seems obvious. There's still a cost to putting miles on a car. But, while I don't commute, I'd be far more casual about having something drive me an hour into the city for an evening or a couple hours into the mountains for a day of activities than I am having to drive.
Or things like "take this pie over to grandma" or "drive my son to his friend's". Taking the human out of the equation would make a whole bunch of new use cases pop up that aren't convenient now. Traffic would suuuuck.
Sure. I live on one side of a highway; getting to the other side during rush hour is a PITA (limited bridge crossings). So anything that requires me to cross the highway gets put off until the weekend. If the car drivers itself and I can read a book, those errands likely get done as they pop up.
I totally do errands more often when it is comfortable and low effort. 100%. Moreover, sometimes I am not driving because I am tired or because I have been drinking alcohol. Actual full self driving would absolutely mean me taking car in those situations.
I'd also have the car go do things for my convenience that I would otherwise be unwilling to do because of the traffic. The car doesn't care about stop-n-go, so I can send it any time. It'll get there eventually.
Why do we “need less personal car ownership?” It’s not clear to me why that’s a key goal. I also think that’s just not going to happen, not until the flying cats with cities in the skies is a reality.
Owning a car is massively expensive at a personal level… Purchase cost, depreciation, fuel, maintenance, space to store it. Minimum a few hundred $/month for an “asset” that sits idle most of the day.
Owning cars is massively expensive at a societal level… Storage space - so much asphalt for parking (both at destinations and at home). Pollution - production/raw materials, emissions during use, eventually disposal. Injuries to drivers, pedestrians, animals.
All that money could go to education, health care, recreation, etc.
I’m not saying ban cars. But align incentives towards fewer cars and less driving. Many cities are already making some progress - DC, NYC, Seattle, Portland are all adding transit alternatives at a rapid rate and seeing reduced car usage. But then Houston and some others are doing the opposite and continuing to sprawl and pave everything.
Edit - And yes, some people live in rural areas and need cars (trains and buses don’t scale in rural areas). But, the suburban sprawl we have today? We can do better. Less single-use, single-family zoning. More bike lanes and walking paths (preferably both separated from cars). More light commercial/retail.
Just thinking about my own neighborhood (Reston VA), there is an intersection 0.5 mile down the road that should (IMO) have a corner market and a cafe. There are thousands of residents who could walk there for milk, eggs, or a snack. Instead, they all have to drive to the strip mall 2 miles down the road. A little planning in advance would have made a difference here, but with current zoning, we’re stuck with what we have - you couldn’t buy a corner lot and build a shop if you wanted.
I don't know about the rural-need-car argument. In Switzerland even little villages have rail connections.
Imo it's about planning and values. America just doesn't value spending on general welfare. We'd prefer to have wealth hoarded behind closed gates it seems.
Most of Canada has the same planning problems as the US. Namely, car-dependency. Toronto and Montreal are exceptions, but most of the country follows the same development pattern as the us.
But I think you're thinking about this wrong. Sprawl and transport policy are intimately related. If we do car-dependent development, you get a development pattern that requires cars. Then of course you look at the population distribution and say, it's hard to serve this with trains. But if you started with trains, the development pattern would mirror the transit, and you'd say, wow, trains really solved the problem.
America unfortunately went down the car path and faces a painful correction back to sanity. As we move towards transit and factor in car-driving externalities into the cost of driving, communities that can t feasibly be serviced with rail will die out, as is natural.
But that doesn't mean small villages won't exist. They will just be forced to exist where it is reasonable, instead of in unsustainable locations as they do today.
Size doesn't really matter. It's where and how you develop. Take Dallas for example. You drive for 2 hours at full highway speed, and you're still in Dallas. Why? Because it's built for cars and around cheap, suburban growth.
Let's say instead Dallas was built around rail. You'd see a completely different development pattern. Shopping, business, and recreation would be concentrated around metro/rail stops. Dallas would be a city instead of a blob of suburbs. People would live near rail, facilitating their easy transport to recreation, shops, and their jobs.
The absolute size of your area isn't really relevant, it's where you allocate the population. Suburbia is the worst allocation.
Texas is in for a painful correction once they figure that out. Though more likely it'll just de-populate/be abandoned once the infrastructure maintenance bills come due in 30 year and other cities built out sane rail-based density.
You can take China as a more size equivalent example if you like. Land size doesn't really matter though, it's where you decide to build (aka, density and land use).
This confuses many things in an attempt to reach an incorrect conclusion. The USA is massive - on the same scale as continental Europe. Every village in Switzerland has a rail connection. How about every village in Poland? Russia? Belarus?
You could have a full self driving only lane that’s allows speeds up to 200MPH. Self driving cars could have better reaction times allowing for higher highway density with faster speeds.
Right, and what happens when Joe Schmo pulls into the 200MPH self driving lane and gets his car exploded, along with causing a 10 self-driving-car pileup?
Or, for the purpose of discussion, let’s assume that the lane is completely separated. What if there’s some unexpected condition that causes a crash? Even with vehicle to vehicle communication, can the other cars behind it really stop in time at 200MPH?
Yes, I believe that this was Elon’s plan all along and I believe that full autonomous self driving (FASD should be the new marketing term) will make traffic better because we will literally speed up traffics very significantly. Imagine going the speed of NASCAR race cars in your neighborhood streets. With FASD, every street becomes a German Autobahn. Elon is still the true visionary of our time and we should all help make his vision come true. I will start lobbying my city council to remove speed limits on our neighborhood streets. Thank you for this wonderful idea.
It really depends on if we continue down personal car ownership or not. If not, then you can get pretty close to double throughput by appropriately sizing cars and lane splitting. Most trips are just the person without significant baggage. That can be handled by something like this:
Which is small enough to split lanes and double capacity.
It's also possible that with smart driving cars effectively everyone would be using an Uber like service. That makes it much easier to combine trips and put people into 4-12 sized car buses. If that comes to pass then going from mostly 1 person cars to that would multiply capacity many times.
It's complicated to say. It makes getting stuck in traffic less bad as you can do other things, so that tend to increase it. But it also means less traffic consisting of cars cruising around looking for a parking spot in places where that's an important source of traffic. So I expect it to make traffic in LA worse but traffic in New York better based on those. Ideally you'd also have more multi-modal trips, taking the train most of the way to your destination and being picked up by a self driving car when you get there. That requires infrastructure investment, though, so won't happen quickly.
The biggest change, I think, would be less need for so much parking in so many places. I don't think you could get cities all the way down to just 1 parking spot per car but we could get it down quite a ways and get it out of central business districts.
EDIT: Oh, and with regards to that study, the fraction of the time a rideshare car is going to spend driving between customers as opposed to with customers is going to scale with the fraction of people using rideshare. The more trips are rideshare the more likely it is that on completing a ride any given car will find a new rider nearby as opposed ot far away. You would have issues like daily commutes everybody is going the same direction. But in that case adding more traffic in the other direction isn't as much of a problem as the vehicle kilometers would suggest.
>>High frequency busses in the interim with rail transit is the real solution.
Mass transit systems only really work in high density areas.
This is because the time taken to stop (pasengers on and off) then go is high. So the distance matters less than the number of stops.
To be appealing your nearest stop should be within a half mile or so of your house. If a bus can carry 60 people, and only one or two people get in and off at a time, then it takes say 30 stops to fill up, or empty. That makes overall travel speed really slow.
Contrast with a dense city - here big groups get on and off at every stop. So a bus needs fewer stops in a route, while being useful to lots of people.
Cars work in low-density areas because they typically "stop" only twice, once to load, once to unload.
So absolutely yes, we need better public transport, but that ideally means living and working in high density areas. (which for most(?) US people is undesirable.
Cars work in low density areas where they don't interact with high density areas, but even then, they're best on dirt roads and gravel where the residents can do all the maintenance.
In mid density areas like single family suburban housing, neither works very well
Obviously YMMV, but where I live (single family suburban housing) it works fantastically. The only time traffic hinders me is if I get in the freeway in the morning heading into the city.
But obviously there are many factors in play, so each place is different.
In suburban areas, light rail can work quite well. Although you have to wait longer because of stops, the trains can also move much faster than cars--existing trains are fairly slow, but introducing higher-speed trains would make them competitive with cars.
In rural areas, though, your point stands. For adults traveling alone, though, lightweight electronic vehicles like e-bikes may be better than electric cars.
Ultimately, an "electric car" is like a "horseless carriage"--an attempt to shoehorn a new energy source into an old form factor where it doesn't fit particularly well.
It can work for long commutes when you have park and rides, but you still need cars to go the last 1-5 miles. Still nice to take a train the bulk of a long commute into a city core.
A lot of people who think trains and buses can do it all are young and have never had kids. Nothing about the suburbs made sense to me before kids. Everything makes sense after kids.
I think Elon's biggest problem with mass transit is just that he doesn't like the masses. As he once tweeted:
> I think public transport is painful. It sucks. Why do you want to get on something with a lot of other people[...]And there’s like a bunch of random strangers, one of who might be a serial killer, OK, great.
Not the only problem he cites there (note the "[...]"). But I'm pretty sure there's a reason it's the most prominent.
(Incidentally, given the demographics of those most reliant on mass transit, one wonders why exactly he thinks that they are so likely to be serial killers.)
This is not something specific to Elon. Anyone who’s regularly ridden the 38 “Geary” in San Francisco knows there will be a minority of riders who make the experience, at the very least, annoying, if not scarring. Riding home every day with a screeching schizophrenic is not top on my list of preferred activities, nor anyone’s.
> Mainstream full self driving would just create massive traffic and gridlock
I believe the opposite is the more likely outcome: I think we can expect much lighter traffic the more autonomous vehicles there are.
Most traffic is caused by humans doing really irrational things, like following each other too closely, making unnecessary lane changes, and driving at inconsistent speeds. It's very rare that traffic congestion is actually caused by roads' lack of bandwidth.
The sorts of Driving behaviors you mention are not a significant cause of traffic congestion. Traffic congestion is caused by there being too many cars on the road with limited space for cars.
Anything that will increase the amount of cars will quickly dominate inefficient driving behaviors, and self driving cars will massively increase the amount of cars on the road, as it would drive down the cost of driving and enable many many people who currently can’t drive to get onto the road (eg. Disabled, elderly, unlicensed, children).
> "Traffic congestion is caused by there being too many cars on the road with limited space for cars."
I mean in some sense this is always going to be true, and removing cars will always help alleviate traffic. But it's not true that driving behaviour is not a significant cause of traffic congestion: it's the cause in most cases. [1]
For an extreme illustration, there's a well-known phenomenon where adding extra lanes of traffic to a highway or freeway actually worsens traffic: the added bandwidth interacts in weird/subtle ways with driver behaviour. [2]
Yes these are real phenomenons that exist, but focusing on this marginal issue is fiddling at the edges of the real core problem.
Roads are fixed in size, cars are very big and take up too much space.
Fundamentally baked into the design this is an inefficient way to move people around and there's no fix to be found working within a car oriented framework. The only solution is in removing cars from the road. This is why self driving cannot fix traffic as it will only add more cars.
Self driving buses, now that would be an improvement, but self driving cars no.
On top of this, a future where every car on the road is self driving also lends itself to a future where every car is in constant communication with the cars around it. A car 10 spaces ahead is slowing down? Your car already knows and is prepared to handle it with grace rather than slamming on brakes. Cars behaving in a predictable and communicable manner, as much as possible, alleviates traffic by a large factor.
This is obviously largely ruined by one single human driver acting unpredictably, so it's a very "all or nothing" resolution.
The metric to judge assisted driving technologies is primarily safety. Driving is not safe and I suspect will appear insane to a future generation. Which generation and how soon is another question. However, what the US needs to massive investments in rail, which address both safety and traffic.
It's not that driving is unsafe, it's that cars are unsafe. Even if all cars were self driving, accidents happen, even with computers. A world without cars at all would be safer, albeit unrealistic. So the question is, would self driving put more or fewer cars on the road? I argue more. A lot more. So much more that it would offset the increase in marginal safety per vehicle.
It would also discourage the true solution around urbanization and continue to fuel more and more sprawling madness.
This only appears true because we've made our entire world so safe that we can call one fatality every 100 million miles dangerous. Given everything we use cars for and the immense utility of them, being driven almost exclusively by amateurs ... cars are remarkably safe.
And if you could find a way to reliably remove the 1% that cause most of the problems, it would be even safer.
My sad related factoid to this is that suicide overtook car accidents as the leading cause of death of teenagers in Colorado. This is because although suicide is a worsening problem, crash safety and driver education have improved much faster than the suicide rate has been rising, causing them to flip for the first time.
Allowing FSD (in the next 20 years) to drive is unsafe, in my eyes. Investment in alternative transportation is the answer that doesn’t require driving from human or an algorithm.
People always bring up this idea that future generations will think it is “insane” to drive your own car. Will everyone in the future be very scared children in a far left monoculture? I just think people are more complicated than that. I think if people are really like that we will have very fragile humans.
We already do drive more than we need to because of how society is designed in America, but if the costs were lower as promised and you basically had an always-on-demand car with no driver, you could work or something while in the car, and take a nap wouldn't people take a lot more road trips or trips otherwise? I don't see how this wouldn't increase usage. That being said, increase in usage doesn't necessarily correlate to increased traffic.
> FSD is not by itself going to compel people to drive more than they need to.
What people "need" (or "want") changes: if you live in the city and want to partake in Nature, it may be a hassle to get out of the city. But if you can 'just' hop in a FSD car and have it drive overnight (while you sleep) to a nature reserve/park, people may do that. Whereas previously they may not desired it enough to deal with traffic in getting there.
Imagine FSD recreational vehicles (RVs, UK: campers), or those live-in vans that are somewhat popular nowadays: go to sleep in one place, wake up in another.
Great, but not relevant to this story. HN is a worse place when commenters use tangentially-related discussions as thin pretenses to spam their pre-written manifestos. Please stop.
> Same thing that happened pre-COVID with "ridesharing"
Fifteen minutes in an Uber is far more pleasant, productive, and I’d wager healthy, than ten minutes in gridlock. I can focus intensely on a task or thought or even take a nap.
We should also build rail. But self-driving cars look more likely to complement rail than detract from it. (One can imagine multi-modal trains, onto which self-driving cars park so their owners can enjoy privacy and a custom interior alongside the economics and eco-friendliness of rail.)
Traffic is _people_. "Induced demand" just means that more people are able to get around in a presumably less miserable way, rather than you (who are presumably already driving if gridlock is a problem to you) getting a better experience.
And no, buses or rail won't provide a good experience until they're so frequent and have so many stops that they basically end up being ridiculously overbuilt taxis in all but the highest-density areas.
> Just think about it - what happens when everyone has a self driving car? All of those cars that are otherwise parked are now on the road.
Why would that be the case?
I would keep my self-driving car the same place I keep my normal car now. Why would it be driving around when I’m not in it? That would waste electricity. It would drive me somewhere, park, and wait — just like I do with my current car.
When you have a self driving car, it should be able to pick you up--driverless-- at the airport when you are coming home from a trip. But logistically that's a challenge for the airports.
If it wasn’t my own car doing that, it would be a friend using their own car, or a taxi, so I doubt I’d be increasing the total number of cars on the road.
You might rent it out so that it's earning money for you when you're not using it. What the GP fails to understand is that if and when that happens, your car won't add to the traffic burden, it will replace another car that would otherwise be on the road.
Also, if we ever get our act together and make these things talk to each other and to the infrastructure, we will have effectively built trains that can operate without fixed rail. The upside is obviously considerable.
Bottom line is that FSD won't increase demand, but it has the potential to help us manage demand in ways that are completely unattainable with human drivers.
> Just think about it - what happens when everyone has a self driving car? All of those cars that are otherwise parked are now on the road. Traffic. Same thing that happened pre-COVID with "ridesharing".
Explain this logic? To my understanding, ride sharing increased traffic because it increased demand for car travel. They did this by being a better choice than other modes of transit. But I believe the idea of ride sharing itself took 2-3 would-be individual riders, and put them in one car. The flip side was that the car would sometimes be empty, driving to its next customer, but that wasn't nearly as common as having multiple customers at once.
But holding constant demand for car travel, I don't see the logic for why FSD ride sharing would increase traffic. It would even free up an extra seat to add an additional passenger. And in theory, it would be able to drive more efficiently if it was an entire network of FSD vehicles.
you should read the study I linked. generally at a minimum self driving removes the need to drive, by definition, making the time more productive and removes downsides of driving, thus encouraging it.
You’re correct, self driving cars are not going to reduce traffic. A car has an average of 1.5 passengers in it while driving and it spends some 95% of its time parked.
Because they’re low/single occupancy vehicles and everybody is trying to get to and from work at approximately the same time you still need to have capacity for every single person at peak time. Any temporary reduction in traffic is immediately filled up by an increase in demand.
Just look at the last 50 years of urban development of the United States and tell me adding a few more lanes it’s gonna fix anything. Lyft is basically a self driving car if you squint, and it did nothing.
Throw in a couple trains and some busses people actually want to take, and you might actually solve the problem.
I have yet to see a plausible explanation for how self-driving cars are to reduce traffic.
> Or we can continue burning tens of billions on self driving personal vehicles.
Your comment operates on the assumption that autonomous buses provide greater net utility than a "manned" bus - sure, without a driver's seat you might add a row of seats, but are two more seats worth the R&D money to design the bus, and all the cameras and sensors that have to be added to the bus and then maintained for years to come?
To me, autonomous driving features are a luxury item - a convenience reserved for those who can afford it, until it becomes cheap enough to find its way into more affordable vehicles.
There is no evidence to support this theory, and in fact the opposite is likely true. Full self driving would lend itself tremendously to increased public transit options. And tremendously lower car ownership.
Ride sharing is already “self driving” from the end user perspective and has increased congestion. The actual technological full self driving function is an implementation detail.
Only if humans are banished altogether from the roads. In which case the current attempts at FSD are pointless anyway and we should focus on building the infrastructure for simpler automated cars. As it stands today, every current automated car in existence is quite a lot more timid than a typical human and will cause gridlock very quickly as a result. An awful lot of driving relies on behavior assumptions that computers aren't good at making yet.
What makes you think that the self driving technology being developed (by, you claim, "burning tens of billions") is only for "personal vehicles"? The same technology is what can enable the large fleets of self-driving minivans that are necessary for what you describe. Cities can impose a charge per mile driven (that fluctuates depending on current traffic) and people can then decide whether they wish to use a "personal" vehicle or order a self-driving Uber/Waymo minivan pooled ride.
The roads will be so gridlocked no one will drive on them anymore /s.
Seriously though there are ways of reducing traffic on the roads like increased registration taxes, raising the price of tolls, gas price hikes (or equivalent dc fast charging price hikes). Gridlock won’t be a problem.
DC fast charging price hikes will be especially effective at reducing robotaxi usage as unlike privately owned EVs, robotaxis will primarily rely on fast charging.
Also don’t forget about how much traffic flow will improve because self driving cars don’t get impatient and tailgate.
Car traffic has very different physics than that of people walking in a popular area.
The flux of the number of vehicles, as you increase the number of vehicles on the road, starts out going up as one would expect. But at a certain point, adding more cars on a road starts slowing down the movement. And then as more cars come on the road, the flux goes down, there is less road capacity as it get more crowded. This is the fundamental diagram of traffic flow:
So once cars on the road exceed that peak amount of flux, there's a phase change and cars start backing up.
So with cars, gridlock really does mean that far far fewer people get serviced by a chunk of road. Whereas with people in a busy area, there really are more people there.
> Seriously though there are ways of reducing traffic on the roads like increased registration taxes, raising the price of tolls, gas price hikes (or equivalent dc fast charging price hikes). Gridlock won’t be a problem.
Yes, this is why traffic is so minimal in the top 10 MSAs in America.
It just means prices are not high enough. Make it $100 to travel at rush hour on a particular stretch of road, and congestion will go away very quickly.
There are societal reasons why this won’t be accepted by the public. It’s sort of analogous to allowing the rich to pay for a higher spot on an organ donor waitlist, in theory it is economically efficient but in practice it is unfair and will cause outrage.
Yes, the political challenge of getting massive wealth redistribution passed in conjunction with public opposition to giving up the luxury of personal cars means the probability of the US ever getting its transit situation in order is very low until energy prices rise high enough to force a change.
It also means energy prices will not rise high enough until the government becomes incapable of subsidizing them, if you (s your argument does) neglect the possibility of political shifts.
I want some of those subsidies. Aside from a few particular oil-rich nations (and some very poor ones), where are these fuel subsidies I keep hearing about? A gallon of gasoline largely costs the same in the US as it does in Europe, if you strip away the taxes.
> Aside from a few particular oil-rich nations (and some very poor ones), where are these fuel subsidies I keep hearing about?
Flowing from the US to oil-rich nations in form of security assistance to buy policy, or in the form of US military going to enforce policy preferences, as well as:
> A gallon of gasoline largely costs the same in the US as it does in Europe, if you strip away the taxes.
Accepting negative externalities without internalizing them (whether through direct liability for externalized harms or pigovian taxes) is a forced subsidy to the activity creating those externalities by the rest of society.
Reducing the congestion isn't the goal though. Increasing throughput is. Cars simply have worse throughput than a train, or a bus for that matter. Simple as that. Charging more money would be counterproductive unless that money served to build out transit since you would increase the congestion on the non-toll roads.
There will be no political will to invest in quality public transit unless there is lots of pain in using comfortable and convenient individual vehicle transit.
Paying $100+ for a trip will make people support bus and train and bicycle infrastructure investments. Keep driving costs down at $0.60 to $1.00 per minute or per mile, and it makes sense to use your own car, which then means it makes sense to support initiatives that are contrary to public transit, such as mandated parking lot minimums and giant roads.
in practice this doesn't make sense since people are broke. you're be punishing them for no reason, they would continue to drive on more and more congested roads, take longer to get to their destination and you wouldn't raise enough to build out public transit. everyone loses. well, actually, I guess rich people paying $100 a ride on underutilized roads would win. great for them I guess.
Make use fees explicitly progressive: pay a fee to use the road, and then at the end of the month, the bottom 25% get back 150% of the average fee, middle 50% get 100%, top 25% get 50% back.
Drive less than average and you're making a profit regardless of where you are on the scale.
> the retort comes up because charging egregiously to use public roads that they're all paying for with their tax money makes no sense.
So, stop paying for the roads with tax money. Fully internalize the costs with user fees, directing tax spending elsewhere, and the argument about fairness of tax spending goes away. In fact, raise the user fees high enough that the road users are subsidizing other public projects, specifically mass transit, as well as paying the full cost of the roads.
The idea that a single person is paying a specific part of infrastructure makes no sense. My taxes do not pay for a specific road, they pay for the operations for society as a whole.
Sometimes, that would mean my taxes are going to pay for something which hurts me in the short term, if I want to get to a future where individual cars are not used as much and walking/bicycling/public transit are possible.
The problem is this is just some leftist fantasy and the idea that to solve all of the consequences is to just do wealth distribution is not surprising. You would need to reform the government into a dictatorship to keep these legislations in place. And this is to solve “gridlock”? People already experience gridlock and are not clamoring for a Maoist state.
It is not to solve “gridlock”. It is to change infrastructure to better use society’s resources so that people spend less time in cars and more time living life.
The solutions that worked when there were x number of people living in an area, with y number of miles traveled in z number of individual cars simply might not work if you double or triple x, y, and z given the hard boundary conditions of the world such as space, pollutions, and supply of materials and energy.
I'd rather work in a slow car than be crammed into a full bus. You miss how self driving is changing the economics. A major component of the appeal of public transportation has been that you can do other things while using it. This made up for its (sometimes extreme) inefficiencies with regards to the time it takes you to get to your destination. In fact self driving cars will have an advantage there because you won't have to switch vehicles.
you will simply kill the businesses and housing for anyone living nearby as a side effect. It's not an effective solution because it ignores actual second order effects. Just like toll roads, it will push traffic to other paths while making people who rely on that stretch of road end up paying more.
I live in a city where they decided parking downtown was too cheap because people were parking for too long. The effect was massive drop in business EVERYWHERE downtown. Why would I travel there to pay $25 to park while I shop? I'll just go shop elsewhere, where parking is free, and too bad so sad to the shops that go out of business because policy makers don't understand second order effects.
Now downtown is filled with empty parking and empty stores.
OK, so how do you keep the politicians in office that enact such legislation and if you say you want to create a Maoist state either how is this preferable or how do you stop rebellion?
Wealth inequality is a separate problem with separate solutions. It is possible to simultaneously redistribute wealth, and implement measures to curb personal car usage (which is basically the cause of congestion).
>what happens when everyone has a self driving car? All of those cars that are otherwise parked are now on the road. Traffic.
Why? The number of trips is governed by the demand for trips, not the supply of cars. There isn't sufficient demand for trips to put all the cars on the road all the time, or else they already would be. The inelasticity of demand for trips is right there in the data you cite.
>TNCs led to increased road congestion in terms of both intensity (by 0.9%)
Tens of millions of people live in places where frequent bus service simply isn't practical. It's probably reasonable for society to make choices that overall incentivize against living in those areas, but that isn't going to change where people live anytime soon.
And you can just use taxes to control any induced demand that comes from lower costs or new behaviors. The vehicles have sophisticated tracking systems and telematics built right in.
Buses are as large as they are because the driver's time is valuable. If self-driving cars were cheap you could a bunch of 8 person van/busses on the road that give you most of the environmental/traffic benefit of a bus but could really expand frequency and reach.
lol I did this once with Lyft's cheaper service. One cranky old boomer killed the whole vibe bitching about hte driving "listening to his machine (GPS...)" instead of listening to his own, supposedly superior directions. Wish I'd sampled it.
Adding transit makes people live around the transit - if you let them by not kneecapping them with zoning rules. All this is an emergent property of a system designed to prevent new construction.
Though you're not wrong, in a hypothetical world where most vehicles on the road were self driving busses, the frequency could be dynamically set and reasonable for everyone.
I don’t want a car. I’d happily share one. Uber where I live (Johannesburg) has kinda forced conversation and I always feel the need to tip. But honestly the drivers here are dangerous, speed around, ride their clutch, go through orange lights, don’t want to take short trips. I’d love a Uber that just does the job, doesn’t want to talk and drives sensibly so I can zone out. One my kid could ride in.
Is this true even if you account for the fact that cities can reclaim lanes and buildings dedicated for parking?
I would think that a lot of congestion is related cars having to be parked near a specific venue such as a restaurant or an arena (especially in the latter case where there is a gathering of 1000+ people).
The thing about arenas is those 1,000 people all want to show up and leave at the same time.. so those 1,000 cars will all have the exact same demand and cause roving gridlock from wherever they are parked on the way to the arena.
One of the killer feature I expect to see in the future when we have FSD is that cities will take direct control of traffic in a central way. If cars can be distributed evenly over all roads with each driver getting an average distance/time-to-arrival the overall speed would significant increase, with car owners loosing the control to pick the shortest route in order to be allowed to use the roads in the city.
Centrally controlled FSD cars would also allow for much faster intersection since all cars could accelerate and break simultaneous. Both would require quite tight control and coordination that only computer controlled cars could do.
I expect people to feel like this (and the downvotes was thus expected).
In concept it is no different from getting into a bus or taxi. You specify your destination and choose when to drive, but the passenger do not decide the exact route. A FSD is in concept no different from having your own personal taxi chauffeur.
City engineers often uses one way roads to spread traffic out on multiple roads. By preventing people from using the shortest route they can force traffic to spread out and thus reduce congestion. An other common trick is to prevent left or right turns, so that the shortest route become artificially longer and thus force some part of the traffic to use alternative routes which is less congested. When there is some really nasty recurring congestion there is also other more extreme tools that get deployed to really discourage drivers from all choosing the same road, like excessive road tolls or speed bumps. Cities has a long history of using hostile design to shape traffic in order to address congestion.
Naturally giving cities more direct control to shape traffic is something that will create new problems, among the biggest being privacy. The car has also very long been the symbol of freedom, so reduced congestion will likely take a while if people choose to do that trade. It is a steep price but the rewards are less congestion and replacing hostile design (which often also cost money to construct) with software.
Indeed. Sure is interesting to hear some folks talk about all the things they'll do to other people when given the authority. Of course, in their head it's all justified for the greater good. What is it they say about the path to hell...
>cities will take direct control of traffic in a central way...with car owners losing control to pick the shortest route in order to be allowed to use the roads in the city
I can't see that happening in the United States.
I don't know a single person who would be happy with the government controlling when and where they can drive, in the manner you describe.
I think if we were going to see this, we'd already be seeing it with map apps. Central planning only relies on being able to give directions, not on it being a computer following them.
> Though it can be easily fixed, particularly given the lack of user control over cars with FSD: For self driving cars all public roads are toll roads.
I disagree. Politically it would be hard to convince the majority of people (which would be the default state until some inversion) that a vehicle that is capable of self-driving should have to a surcharge per mile, given that it could also be driven manually. The only recourse would be to make all roads toll roads, which obviously won't happen.
Depends on the public's perception of self driving. If it's perceived as some tool of the rich making roads unusable for the rest it might not be so hard to make the case. We already have the situation where EV's aren't paying road taxes that everyone else pays.
Induced demand is a weird edge case that means the road network causes each trip to cover more distance. Usually what they mean is latent demand, which is the number of people who already wanted to go somewhere, but gave up and stayed home due to traffic on roads that haven’t kept up with population.
>Mainstream full self driving would just create massive traffic and gridlock
Can you show your work here?
> States are better off acknowledging that sprawling car dependent cities are not sustainable. High frequency busses in the interim with rail transit is the real solution.
No, see, mass transit is pointless since it would just create massive demand that would overwhelm the system and thus fall apart. Citation: trust me bro
California was the first state to take autonomous vehicles seriously.[1] DMV licenses them. There's three levels of license. The "learner's permit" is testing with a safety driver. The autonomous license is testing with no safety driver. Deployment is carrying real customers for money. Waymo and Cruise have deployment licenses, which requires testing and evaluation by DMV. You can see their cars running around San Francisco with nobody in the driver's seat. That's self-driving.
Tesla has never been able to get beyond the learner's permit. So, they can't sell a fake self driving product. It's that simple.
Waymo and Cruise deploy in extremely limited public situations, including only in the middle of the night, and have been in the news a lot recently for a serious of crashes and software failures, some of which caused a pile up of around a dozen autonomous vehicles just blocking the road that had to be removed with tow trucks and employees. The scope of what Tesla is doing is vastly larger, more future proof, more reliable, and much more difficult to achieve.
Additionally, Tesla's model right now is to have a driver in the seat. They aren't trying to deploy a car without anyone in the driver seat. There's no need for them to have any license other than the one they have.
> Waymo and Cruise deploy in extremely limited public situations
This is provably wrong for Waymo. It may be geo-fenced, but it's robotaxi service is unlimited in SF and Phoenix.
> The scope of what Tesla is doing is vastly larger, more future proof, more reliable, and much more difficult to achieve.
If it's more reliable, then getting the license from CA should be a breeze. The required reporting and publication of disengagement metrics seem to be giving Tesla cold feet, I wonder why
> Additionally, Tesla's model right now is to have a driver in the seat.
Great! So they should appropriately name it as a driver assistance technology.
> Tesla's model right now is to have a driver in the seat. They aren't trying to deploy a car without anyone in the driver seat. There's no need for them to have any license other than the one they have.
Don't they need to get past the driver in seat level to make Musk's "1 Million Robotaxis by the end of 2020" promise come true?
> The scope of what Tesla is doing is vastly larger, more future proof, more reliable, and much more difficult to achieve.
I'd say the scope of what Waymo and Cruise are doing is vastly larger. Full Self Driving taxi service, in which they have achieved small milestones.
> Additionally, Tesla's model right now is to have a driver in the seat. They aren't trying to deploy a car without anyone in the driver seat. There's no need for them to have any license other than the one they have.
Well yeah, but they have been promising "full self driving" for a while where the driver won't have to do anything. They also want to have robo taxis in the future for which they should be testing with a license now.
Uber predicted it was just a few years away in 2016.[1]
Google's Sergey Brin predicted driverless cars by 2017.[2]
BMW predicted we'd have driverless cars by 2020.[3]
Supposedly sober industry analysts predicted 2019.[4]
It's also worth pointing out that many of Elon's predictions were not predictions of a finished product, but rather predicting when Tesla would have a system "capable" of self driving, not necessarily a system capable of driving consistently and reliably. (In much the same way, I am capable of hitting a target with a crossbow, but not consistently and reliably.)
To be clear, I'm not defending Musk's predictions, only pointing out that such predictions were widespread and Musk was just one of many, many, many people who turned out to be wrong.
Uh huh. And did they take money from customers with the promise of full self-driving being just around the corner? Did they take money from customers for a product making the specific claim that it has the "hardware needed for full self-driving capability at a safety level substantially greater than that of a human driver":
Tesla has backed itself into this corner for no good reason, which is an irony considering that Teslas aren't good at backing themselves into anywhere else:
> Musk will continue to lie about it because he thinks that without full self-driving Tesla is worth basically zero
Lots of people who buy Teslas don’t care about the self driving at all. They want the best-designed, best-driving, fastest electric cars on the road with the best charging network.
> They want the best-designed, best-driving, fastest electric cars on the road with the best charging network.
Well, Europe's charging infrastructure is the best at the moment and "best driving" is subjective. So do you mean they want a Porsche Taycan in Europe? Or maybe the Audi e-tron GT? Or the Mercedes EQS? Or perhaps the BMW iX?
Many entities have released cherry-picked marketing videos of autonomy appearing to work perfectly under controlled conditions. About as convincing as TV commercials for supermarket deodorant that link use of their product to sexual desirability.
Whether Tesla is ahead or far behind others, they're the only ones exposing their technology to unsupervised assessment by independent third parties in uncontrolled, adversarial conditions. On the other hand Waymo have real, productive, actually driverless cars operating within geofenced areas right now. How do you compare the two? Anyone who claims to know which of the two is "ahead" is either smuggling in unstated qualifiers, or they're blowing smoke.
Claim two sounded a bit suspicious so I checked the quote:
" He promised, despite unhappy faces from his engineers, that it would take fewer years than he had fingers on his right hand before they were available to everyone – although the price wasn't mentioned."
I don’t understand why this requires a California law to address? Why hasn’t the FTC intervened? A quick search reveals that it is ‘on their radar’ as of this past June, but as other commenters have pointed out, the claim has been used for more than 5 years. Perhaps they need a more sensitive radar.
Being "pretty clear" in the fine print doesn't help when your product is called "Full Self Driving".
> But Musk bought Twitter.
The bill was introduced in mid-February and was presumably drafted over the preceding months. Musk made his very first Tweet about buying Twitter in mid-April. Not sure what connection you're trying to make here.
> Being "pretty clear" in the fine print doesn't help when your product is called "Full Self Driving".
The likely chose the FSD branding with good intentions. Tesla just had some over-optimistic projections of what it will be in the near future and they weren't the only ones being optimistic in 2015-2016 when it launched (you can find plenty of Youtube interviews from other car companies at the time projecting <=2025).
The plan was to always call it in "beta" and an early ~$2-10k(?) investment for the future offering that wasn't yet ready.
The only problem is when those projections turned out to be heavily optimistic after ~2-3yrs in they should have rebranded it... but chose to stick with it for whatever reason.
Another, which I find more likely when there's lots of money involved, is that there was deception to feed the hype machine which caused profits of millions.
If it was simply ‘self-driving’ there would seem to be more wiggle room, but the inclusion of ‘full’ to the claim makes a much more specific promise of the degree of autonomy drivers should expect. Adding more specific detailed language elsewhere should explain a marketing claim, not contradict it.
FSD is a Boondoggle. (Collision/accident avoidance and driver assist systems are not)
I don't think we will see a legitimate L5 automation in our lifetimes. Take my 1.5 mile journey to do the morning school run, in a small UK town:
- Getting off the driveway, there is a wall that creates a blind spot behind it, often other parents are walking small kids who are shorter than that wall. I can't see behind it. I check down the street as I get into the car, watch for subtle reflections, and aware that an adult further down the road may have a kid closer by on a small bike.
- Turning out the end of the road, there are parked cars blocking my view, it's safer to stop slightly further back as you can see slightly better.
- Another turning onto a busy road, at rush hour you can't get out without a considerate driver pausing and waving you out.
- Cross roads that is slightly out of alignment. The collision avoidance on my Audi often incorrectly triggers braking thinking that we are going to have a head on crash.
- Delivery vans blocking a lane, heavy on coming traffic. Considerate delivers on the other side flashing to let you around.
- A right hand turn (uk driving on left). It's better to pull slightly across the centre line letting people going straight on pass on the left, while you wait for a gap across the road. Plenty of space even for the largest trucks on the other side when you do this. Keeps traffic moving and everyone happy.
- Another right turn, no road markings and 4! different exits off the road in under 8m. Often blocked by cars so have to go further forward by a car length and double back.
- School car park! Children walking everywhere, large cars, teenagers who have just passed their driving test, parents dropping their kids out in the middle of a moving queue.
I just don't think it's possible for AI FSD to understand the nuance of all that.
(Yes, a 1.5 mile journey to school, we should be walking)
In the US going around a vehicle turning out of the traffic lane is illegal. Of course people still do it all the time.
The cooperation stuff is going to be harder (though if you have automatic drivers in a majority of vehicles you can imagine the systems planning together).
I know you are making a joke, but my point is I don't believe we can train these AI system to work collaboratively with humans (maybe with other AI though) and understand the subtlety of a human environment.
I agree, we have two choices, cycle on the busy roads with the 4+8yo or on a footpath through a park and down the pedestrianised road. Second option sounds perfect right?
No, there are big signs saying "no cycling" because I live in an F-ing stupid Tory middle England town where that sort of thing is not aloud...
Once the 4yo is a little bigger we will be breaking that stupid rule.
Yes and when they do they are ticketed or otherwise charged. Who holds liability when someone purchases a package that is supposed to do this for them.
You don’t really need it to work 100% of the time for it to be helpful. Users of FSD Beta report not driving manually 90-95% of the time, which sounds amazing to me.
I can’t believe Tesla has gotten away with selling a feature that doesn’t exist, and will never be delivered, on so many cars for so long. Tesla began selling “full self-driving” in October, 2016 and a month later marketed it with (now known to be heavily controlled) video of a car driving itself for two minutes. Owners of those initial cars have probably paid them off by now and still don’t have a feature they were sold.
In fairness I know people with "FSD" who went into it well aware that it's not really FSD and they still like it. That said, the marketing is clearly fraudulent.
This has always confused me. I knew FSD wasn't feasible, I knew Tesla was selling a lie. I also knew it was an incredible investment opportunity. I passed on it because I envisioned a bunch of lawsuits destroying the company.
Obviously, I was clearly wrong. While the stock is way down from peak it's clearly been an amazing investment for those that invested early. 2020/1/10 $31.88, 2021/11/30 $381.59 (peak I think) and around 122.81 at close Friday.
Normally, I can look back at an investment in mistake and the issues are obvious in hindsight. With Tesla, I still have no idea why the company still hasn't succumb to lawsuits.
Those updates will be cheaper than the 15k price. And for those who already opted in, it‘s a small expense to pay for the upgrade. Tesla can easily afford it.
To nitpick your metaphor further: A 2x4 plank starts off as 2" by 4", and it is shaved down to 1.5 x 3.5 side to straighten it out. A Tesla "FSD" car has never been in its life fully self driving.
Im one of those people, I have it and love testing it out, but I know it's not FSD, I really do get the big deal about the name. Anyone who buys it thinking it's really FSD is an idiot.
The difference between me and apparently everyone else in this forum is I do think TSLA will build a video model that is L4/L5 in the next few years. I live in a rural america and majority of my drives do not require interventions.
Musk said that by 2020 there would be a million robotaxis on the road. To say that anyone who bought it thinking that FSD wasn’t real is a very, very stupid statement.
Before I spend $10K+ on something, I do more research than listening to the claims of some CEO who has a history of being "optimistic" about timelines at a minimum.
The title of this is "Full Self Driving Hardware on All Teslas". The hardware is in fact on all Teslas. What part of that is fraud? Nothing about this video was misleading and the car did in fact drive itself through that route.
Calling it hardware for "Full Self Driving" is fraudulent if FSD doesn't exist, is nowhere close to existing, and the video has been deliberately edited to defraud viewers into thinking the hardware is capable of something that doesn't actually exist.
What sort of semantic game are you playing here and why? I don't understand how you can not see this as anything but competent fraudulent, particularly with the editing that happened to make this video. I'm truly bewildered, this is not rhetorical or exaggerated at all!
The name FSD intimates an SAE Level 5 experience. It’s not partial self-driving or any other sort, but full. If there was any name to suggest it, this would be it.
Imagine going to an emergency department with chest pains/trouble breathing/thunderclap headache, and the triage nurse saying “oooooh yeaaaaah, I know emergency is in our name but we don’t really deal with that, you’re on your own”.
Things that are critical to life-or-limb should be appropriately named without the use of weasel words or asterisks.
You know there are lots of ERs that can’t handle lots of emergencies, right? There’s even a whole “Level” system to describe which emergencies they are and are not equipped to handle.
Even when a rural ER receives a CTAS 1-2 patient, there is a chain of survival that’s followed to ensure continuity of care and transfer to an appropriate facility and accountability.
There is no safe failure mode in FSD that doesn’t rely on human intervention that’s intimated in the name.
Any Kickstarter is obliged to be very explicit about the risks. They're required to have a section in the pitch dedicated to what might go wrong, and Kickstarter itself has text everywhere reminding you that rewards are not guaranteed. Tesla, on the other hand, never acknowledges the very real possibility that your extremely expensive vehicle never sees full self driving.
The sale page is pretty explicit about what you’re buying and has this note:
> The currently enabled features require active driver supervision and do not make the vehicle autonomous. The activation and use of these features are dependent on achieving reliability far in excess of human drivers as demonstrated by billions of miles of experience, as well as regulatory approval, which may take longer in some jurisdictions. As these self-driving features evolve, your car will be continuously upgraded through over-the-air software updates.
Also, at this point, my experience with the FSD package in my 2020 Model 3 is that it is not misleading to call it a “FSD beta”: in specific predictable situations it has problems with maneuvers but, over the last month or so, it’s continuously gotten better at making full trips without disengaging.
Somehow we got used to companies selling products named "Definitely Does Thing [X]!" and then adding fine print that says "this product absolutely does not do thing [X]." I think it would be much better if we stopped accepting that practice.
The stuff above this says exactly what the FSD package is. All this fine print is adding is that you will get new features as they are released. In 12/2022 this is a very good description of what the package does (and, IMO, it undersells the capability: autosteer on city streets works really well for me with the new FSD package)
> Full Self-Driving Capability
$15,000
> All functionality of Basic Autopilot and Enhanced Autopilot
Again, to repeat my point (which I didn't think was very confusing), the things described in the small-point text are all fine. But they cannot in any reasonable world be summarized by the product name "Full Self Driving." Insofar as California is encouraging Tesla to name their products more accurately, that seems like an absolute good.
This point holds even if I don't go into the fact that both autosteer/autopilot (phantom braking problems) and FSD-beta (requires active driver control, disengagements every < 1 mile) are both kind of a mess. Or that the claims from Musk regarding capabilities and timing have been completely inaccurate.
> they cannot in any reasonable world be summarized by the product name "Full Self Driving."
I think it’s fine to name a placeholder for a feature that’s in development according to the final state. Especially if the feature gives you access to the most recent developments towards that feature.
As for the other stuff, the FUD online about FSD is completely inconsistent with my daily experience of using it. It’s by no means perfect, but it does a good job for my normal driving around SoCal
This note makes it very obvious that customer's evaluation of the expected timeline might impact their purchase decision. In my non-expert opinion Tesla puts itself on notice that any false statements that might impact that evaluation would be fraudulent. If my opinion matches the legal consensus, then the doctored video would be such a false statement.
Hardly surprising that action has been taken over FSD, especially the robo-taxi claims and the deceptive advertising over this alleged safety-critical system proven to be unsafe and putting other drivers on the roads at risk. As I have previously said before. [0]
So not a surprise here that many are realising that FSD is a complete scam and is being investigated by regulators and lawmakers for its false claims and misleading advertising.
It is pretty trivial to disprove that: A human could no doubt "remotely" drive using the current Tesla cameras. So the sensor suite might be more challenging than alternatives, but obviously not impossible.
You could however argue that the current computer isn't powerful enough to run the software required for vision based L3/L4.
>A human could no doubt "remotely" drive using the current Tesla cameras.
I have doubts about that. I wouldn't want to remotely drive that without much better cameras at the very least. Advanced planning to be in the right lane or avoid obstacles is hard when the details are just a couple pixels.
It’s a trivial claim, advanced without evidence. There’s no reason to believe that a human could drive a car based only on those limited visual inputs. Humans use all our sense to drive.
And even if it were true, so what? Last I checked Tesla didn’t have software which can perfectly emulated a human agent.
> And even if it were true, so what? Last I checked Tesla didn’t have software which can perfectly emulated a human agent.
GP was arguing against:
> The feature is impossible to implement with the sensor suite that comes with the car.
_For that statement_ it's immaterial whether Tesla does something right now, but rather whether it's in principle possible. Modulo your previous objection humans provide a counterexample to that (very general) statement.
This is wrong. Waymo uses both lidar and radar, which Tesla opted to remove. Also, look at Waymo's cars. Those things have sensors strapped all over in ways that give them lots of range and visibility. Tesla will never do that because it compromises the aesthetic.
They're different in practically every way. It's easier to enumerate the things they have in common: NNs borrowed the idea of using a connected network of functions whose outputs feed into each other's inputs.
That's it. That's the total resemblance between the two. The brain isn't just an NN implemented in biology, it has whole systems that aren't accounted for in digital NNs, like hormones and neurotransmitters, and even the system of connected neurons doesn't work the way digital NNs implement it.
Neural networks model the brain exactly as well as objects in OOP model cells: not very well at all. They're inspired by biology, nothing more.
Neural networks as used in AI are inspired by the brain in much the same way that OOP was inspired by the way cells work--neither one is an attempt to faithfully recreate the actual operations of an extremely complex (and only partially understood!) biological system.
The problem with driving in video games is using a keyboard/mouse or controller. Driving with a steering wheel and pedals is pretty easy. Even more so if you have the monitors to give you a realistic field of view.
driving a car in a video game with a steering wheel is easy because it's an experience designed from the ground up with that interface in mind. driving games happily do shit like change the fov to make the user think they're going faster, etc.
being easy to drive with a steering wheel in a driving game, and being easy to drive a real car with a steering wheel (with internet level latency and packet loss at play, mind you) are very different things
Pretty much all simulator games have to cheat. Because most of us can't hop into a Formula 1 race car or the cockpit of an F-22 and drive/fly it with no real training. (Much less race other drivers or fight in a dogfight.)
>being easy to drive with a steering wheel in a driving game, and being easy to drive a real car with a steering wheel (with internet level latency and packet loss at play, mind you) are very different things
Sure, but I don't think it's that different than driving a real car looking out a monitor sized windshield. You might lose some braking and cornering without your inertial perceptions. But you'd still be able to driver around easily.
Either way I think a speedometer is an assumed input for a self driving car and from that you can calculate almost everything needed related to proprioception.
Traction control software uses touch (knowing how much grip tires have via slip), proprioception (knowing the steering angle), equilibroperception (accelerometers).
If you are referring to humans, we have a couple orders of magnitude more connections, our neurons can achieve the same functions at 1 order of magnitude fewer numbers, and our brain has better inductive biases.
By which metrics are human eyes outperforming “any camera that can be put in a car today”? This seems unlikely in almost any domain, given the much wider dynamic range camera sensors can capture - cameras can see into spectrum we simply can’t (infrared, UV…) and operate at much lower levels of light than a human eyeball while retaining full color vision using really cheap tech. They also don’t get tired or worse with age, or forget to wear their glasses, which is nice.
This strikes me as a pretty odd statement to make, personally!
“There is no real comparison” - for the benefit of the less informed, please make the comparison, assuming you are able.
The human eye can perceive 21 stops dynamic range, much better than regular cameras. Event cameras might solve that issue, but they're not used other than in research at the moment.
Maybe in a single still capture? Let’s not forget cameras can easily combine multiple exposures into a single capture to substantially increase dynamic range, and can do so at high frame rates, and can go well beyond 21 stops in doing so. The human eye is stuck with the same ~21 stop range regardless.
If you use a pair of digital sensors with a 15 stop exposure offset between them (seems fair - humans have two…), thats ~30 stops in a single shot if we assume best we get is 15 from a digital sensor. Again though with high-speed exposure blending and one sensor this is not really necessary in a lot of cases.
The practical reality is digital capture can exceed 21 stops and you don’t need particularly fancy equipment to do it. Two decent cellphone grade sensors (~12-14 stops) would be enough if you don’t want to do single sensor blends and would work well for real-time video applications.
They can also "see" much further than the average human can [1].
The only way I could imagine that we are superior to phone cameras is stabilisation, something that could be resolved with vertical integration that informs the sensors and image processing units about forces being applied to the vehicle (though this is coming from somebody outside the field so take it with many grains of salt).
The biggest difference between human vision and cameras is the fovea. Half of our optic nerves are concentrated in the visual area the size of our thumbnail with an outstretched arm. To replicate human vision you have to have a high resolution camera, downsample the image and then grant the AI access to high resolution imagery when requested.
The claim was impossibility with the sensor suite. It may well be impractical. In the long run, there's no better way to be wrong than claiming impossibility.
In which case he’s even more wrong. Humans use a whole mess of different senses while driving, including hearing, the inertial sensitivity in the inner ear, touch to feel vibrations and from the car and the wheels on the road. Plus we have a huge amount of contextual information about the meaning of what we are seeing from life experience outside driving, which no Tesla that currently exists can ever have.
It’s a clever bit of snark, but absurdly wide of the mark. If that’s actuary what the Tesla engineers think, no wonder they’re failing by their own criteria so completely.
The claim isn't that humans can drive as well using the Tesla cameras as they can in person, just that they can. That seems obviously true.
The (not explicitly made) sub-claim is that an AI can make up for the lack of audio, etc. by being smarter than a human and faster than a human, better able to multi-task than a human, and completely non-distractible. That's debatable, but not impossible.
...And furthermore, the neural nets and cameras Tesla uses are vastly inferior to our brains. Just because you can argue that a neural network of some kind uses the same basic structures as our brains doesn't mean that it can come within a light-year of what our brains can actually do.
Edit: I'd love to know why I'm being downvoted. Tesla cars guess depth with a neural net. Humans have the hardware for getting this data directly. Unless you either have lidar, radar, or dedicated stereoscopic cameras, you don't have real/accurate depth data. And depth data like that stops your car from plowing into white trucks.
This is true. But most of the things perceived while driving are outside of the stereoscopic depth perception ability of humans. IIRC that stops around 20 meters or so.
Unless I've been driving very differently from you, most of the things that I care about the precise distance of are well within 20 meters.
And moreover, it's lucky at best if my Model 3 acknowledges a car 20m away on the preview. It struggles with cars a few feet from me on the diagonal at stoplights.
I think it’s substantially unlikely that all of the cars sold with FSD will be capable of using whatever Tesla settles on — between processing power and sensor differences in their lineup, feels like they’d have to hamstring the software a fair amount to make it work fleet wide. Maybe not?
What a windfall for Tesla! They get to walk back their absurd claims while pinning the blame on someone else. They're still probably going to face massive lawsuits when they fail to ever deliver full self driving, but this might at least limit the size of the class.
This was as easy as it gets for the federal government to tell Tesla to drop it, but for some reason neither Dem or Repub administrations are acting on it.
A recent candidate for the U.S. Senate from California built his entire campaign on exposing the truth about Tesla FSD and spent millions doing so[0]. I must have seen his campaign commericials on TV at least 20 or 30 times during the campaign.
Laws probably shouldn't be passed in ultra fast fashion to give democracy some room to happen (people discussing the pros and cons and given room to actually accept or decline rather than being pushed into force prematurely).
California has a slow regulatory body. Other states can act even quicker here. For example just see how quickly abortion regulation was made post Supreme Court ruling.
Other states maybe can react quicker but they choose not to. California is quick to have the political will to act.
California’s regulations on health, safety, and consumer protections are usually a good preview of what other states (or the feds) may consider in coming decade(s).
even that's misleading honestly and Tesla is battling in courts over that label here in Germany. In my opinion the correct label in both the technical sense and informing the average consumer correctly is 'driving assistance'. A system that requires you to have the hands on the wheel at all times and be fully alert and ready to intercept on average every few miles is not an autopilot either.
How much could the recent Musk meltdown be related to FSD just not arriving? It seems like we have reached a point after many years of FSD being “next year” when it’s becoming clear that it’s further away than anyone thought.
I’ve thought the difficulty was being vastly underestimated for a while. A reusable Lunar and Mars shuttle system is probably a lot easier than actual FSD.
I worked full time in a computer vision company (non-automotive) for several years prior to 2018. I joined an automotive company (not Tesla) in 2018 and very intentionally steered clear of the autonomous efforts going on there. Not because my skillset wasn’t best applied there, but because I didn’t believe the roadmaps the industry was laying out at the time. “Level 3 by next year, Level 4 by 2020, Level 5 by 2022” is what they were saying back then. I called bullshit (silently) and worked on viable software instead.
Lots of people were doing the same. This isn’t surprising to a whole bunch of people that work in the non-automotive computer vision/ML space and realize how absurd it is to think the these models will converge any time soon.
I think 6 or 7 years ago, many people thought we would at least have something like Waymo's self-driving taxis in Phoenix in several cities, more publicly available, really starting to change transportation.
We're not too far behind on that estimate. My guess is in less than 3 years - Waymo will be publicly available in large parts of Phoenix and probably another major metro - and within 10 years most of the south west.
But a lot of uninformed people thought we might have FSD everywhere by now. I think there's just less uninformed people now.
My guess is Tesla won't have anything resembling FSD to cover the entire US within 15 years. Though Musk and Tesla fans have shown that FSD is subjective, and they seem to think they already have it now, so I guarantee in 15 years they'll be arguing even louder...
If they change everything about their approach, and/or lease tech from Cruise or Waymo - then maybe they'll have something resembling FSD in large parts of the southwest. Predicting 5 years into the future is tough - 15 years is almost impossible.
Anyway, the pendulum seems to have swung the other way now. Since FSD is so far behind Elon the marketing-hype-stock-pump-Charlatan / the louder voice in the room - now the mainstream idea is - we're never going to have FSD.
This is ironic, since we basically already have it in a few large metros. It seems like the newspapers famously publishing that we might have manned space flight within 10,000,000 years - literally the day before the Wright brothers first flight...
People thought the car would never catch on, because cities were designed for people and bikes and horses. The car was so useful, we redesigned cities. Self-driving will be the same way. Once it takes off in some places and the value is realized, cities will redesign to make it work more broadly. It might take a long time, but it will happen...
Unless the nature of transit changes and individual vehicles are phased out for whatever reason first - though I don't see that happening without FSD busses / street cars.
Self driving isn’t nearly as evolutionary as going from horse to car, or boat to flight. We aren’t likely to redesign entire cities for self driving. We might see self driving only highway lanes with faster speeds but that’s all I’d envision over next few decades.
Today we have roads that aren’t good for self-driving cars and we think “how can self-driving cars be allowed?! They’re so dangerous!” but after self-driving reaches some critical threshold, people will look at the situation and think “why do we tolerate roads that are hazardous for self-driving cars?! That’s so dangerous!”.
I don’t think Tesla wants to release a “FSD” that only works in part of one single city. Phoenix is easy-mode for self driving vehicles which is why Waymo is able to do what they do.
Tesla attempts to solve self-driving by generating a sufficiently-accurate world model in the moment, relying very heavily on vision. Waymo solves self-driving with a centimeter-level precision pre-scanned map of all the places it can drive, relying very heavily on point clouds and LIDAR.
The latter model doesn’t scale, and can’t handle surprises.
Or who knows, maybe it could scale if you had a global fleet of specialized vehicles and gigabit/s full-duplex internet connections in all cars. But that’s a different approach even to Waymo’s.
No no, only musk was stupid enough to imply level 5 autonomy within x years.
Everyone know how hard it is, realtime accurate position, machine perception of roads, & accurate prediction of cars/lorries/other are all very difficult problems. Then to hobble your team by saying "blah blah blah no lidars" because they are expensive and someone from the ML team convinced you that they could estimate depth reliably, was the clincher.
Level 4/5 self driving is more of a matter of scale. if you have a 99.5% safe rating, then a lot of people over a country the size of the states are going to have a bad day.
Let’s not pretend Apple, Uber, Google and countless others weren’t spending billions trying to make this a reality. They didn’t burn that money because they were smart enough to know it wasn’t going to happen. Everyone thought it was imminent. Many still do.
Billions of dollars for Apple and Google are within the realm of no-nonsense moonshots given that their core business is very solid.
Tesla had to use all their weapons including fraud, exagerations and lies because their core business is the worst business in the world: manufacturing automobiles.
So they have essentially nothing to fall back on if FSD fails.
I'm still mystified that tesla isn't mired in lawsuits over this: Some tesla dealers were telling customers that within a few years their self driving tesla would earn back its purchase price from acting as a robotaxi.
Somewhere unrealistic ambition has been mistaken for stupidity. If everyone was “realistic” our progress would move so much slower. Breakthroughs happen because of people like Elon.
Promising that it be ready by 2018(or whenever it is) == unrealistic ambition
Chucking it out the door, branding it as FSD, charging for it, and not actually putting in the right sensor suite to make it work == stupid, or, perhaps dishonest.
If only because you're probably doing two dozen moon missions a year TOPS - probably far less than that - but FSD needs to handle millions of trips flawlessly.
But yeah, at some point he should have wised up and stopped saying that FSD was gonna be ready Any Day Now.
He went so far as to say that it was financially irresponsible to not buy a Tesla, as it was the only car that was fully hardware capable, and that had FSD software coming in about a year. We are very far from the touted fleets of Tesla robotaxis, independently-operated or otherwise.
The hard thing about FSD is related to the scale. It’s the long tail of situations it has to deal with, many of which require some kind of model or understanding of the world outside the car.
It’s the worst sort of problem, one where 10% of the work gets you 90% of the way there and the other 90% of the work is needed to get you the other 10%. Problems like this are really bad for creating this mirage that you’ve almost got it when in reality you are far from the end.
I dinged you for this in a different comment too, but it's clearer here: you're doing exactly the conflation of "Full Self Driving" the product with "Full Autonomy" the concept that the linked law is trying to prevent!
The former has value and people will buy it even if the latter can't be delivered in totality. Making Tesla rename their product does not mean their cars magically stop driving themselves, and especially does not mean that there isn't a market for people (that's me, hi!) who want cars that can drive themselves.
FWIW, as far as "will the car drive itself?", FSD is already here. Call a friend and get a ride. The beta lets people in relatively rapidly now, and there are hundreds of thousands of enabled cars on the roads. It's pretty amazing. Now, it's absolutely true that it's not at a 100% supervision-free level. It's still timid, it has trouble with lane selection in some areas and misses turns, etc... But for months and months intervention-free drives in my area have been the rule and not the exception. My car, for sure, drives me around. It's pretty amazing.
But that's not regulatory approval, and coverage of the issue tends to conflate the two. So you have people like you who read this stuff, but don't get rides from friends, who think somehow it's a disaster and a failure. But it's an actual product people are buying and using. Add that to all the folks who have a vested interest[1] in Tesla failing, and... the discourse around this subject is pretty toxic.
All I can say is, again, call a friend and get a ride. There's a lot of hate to push through, but the cars are amazing.
[1] Because, to be fair, the CEO is an asshole burning Twitter to the ground as we write.
Other companies have cars one public streets without drivers. Calling what current Tesla’s do Full Sell Driving is so misleading it’s probably just fraud.
As an owner in the FSD beta, I think the most misleading thing is Elon's BS time lines. When I bought in August 2019, they were basically telling everyone they have it ready to be a taxi next year with a flip of a switch. Implying they already have it figured out when they didn't. I bought my car with FSD and didn't get into the Beta until a year after they were suppose to be ready for use as a taxi. At this point for me, it's not really good enough to use on city/rural streets without frequent interventions. I pretty much only use self driving on the highway, which is currently using a separate stack than FSD Beta. They are suppose to start using the same stack for highways soon.
Is there any consumer car that is anywhere close? My understanding is that no other car has things like automatic lan change and route following. Let alone stopping at stop signs / red lights etc.
Yeah, nothing else does that stuff. Which is I think the big disconnect in this discourse. People who want to take the anti- side can lean on "It's not complete!" as an argument, because it isn't. But in terms of capability, it's way ahead of the competition[1] And some of us want this feature and want to pay for it! And it works!
So... I'm actually fine with renaming it. People shouting "fraud" on the internet want somehow for a different name to magically make the product bad. But the product is great.
[1] Yes, including Waymo and Cruise, who work only on carefully calibrated geofenced areas. My Tesla will take the highway when it needs to and honor a five-way stop sign. Waymo won't even make an unprotected left turn.
Other driver assist/hands free driving systems do lane changing and still clarify they aren’t self driving.
The most important metric for self driving is the distance a car can travel before it needs assistance from an occupant. That’s what self driving is. Tesla is very far behind the leaders here.
Also, Tesla might not in theory be geofenced, but in practice it will refuse to drive on many roads making it effectively geofenced.
>The most important metric for self driving is the distance a car can travel before it needs assistance from an occupant. That’s what self driving is.
That's reductionist to the point of meaninglessness. A car limited to a specific track could keep going autonomously until it breaks down, thus having an infinite metric on the distance a car can travel before it needs assistance from an occupant, yet practically it could easily be useless as a self-driving vehicle.
Being able to handle self-driving without being restricted to carefully chosen "good spots" and instead being able to generalize is at least on par in importance with distance the vehicle can travel.
Also, the vehicle determining that it can't handle a certain road well enough yet is a desirable feature and isn't really comparable to a bunch of people deciding if a certain road is safe for their vehicle before allowing it to self-drive there, as the latter simply isn't practical at scale.
Self driving trucks at mines have very limited routes but they are legitimately self driving and quite useful.
There are under 4.1 million miles of roads in the US. ~70 people could map every mile of that every year assuming average speed of 30mph and 2,000 hours of work per year. It’s literally trivial compared to building a self driving system.
Tesla’s insistence on good road markings is a far larger limitation than simply pre mapping roads.
> Self driving trucks at mines have very limited routes but they are legitimately self driving and quite useful.
One of the problems with this discourse is the propensity for people to talk past each other. No one is interested in mining automation here, no matter what level of SAE autonomy they qualify for.
> Tesla’s insistence on good road markings is a far larger limitation than simply pre mapping roads.
And another problem is people's insistence on arguing from bad information and ignoring the testimony of those who have good info. I honestly don't know what you're referencing here, FSD works fine on unpainted roads (I live on one).
> no matter what level of SAE autonomy they qualify for.
Yet, it clearly disproved your point. A self driving Taxi or Semi needs to be fully autonomous to be useful, coverage area is a useful feature only after you can get rid of the driver.
I’ve seen a FSD Tesla have issues with bad road markings recently. That’s a complete failure from an autonomy standpoint, because if they can’t predict and thus avoid bad situations the system is completely dependent on a human driver.
And that’s the problem, Tesla has been stuck at at the cool tech demo stage for years and adding features alone isn’t progress because something that generally works as long as a human can take over is so vastly easier they can just gloss over the hard bits. Eventually they will run out of new feature that sort of work and the cars will still still be unable drive anywhere by it’s self. Perhaps they start trying to level 4+ in a few years, but their going to need to start over getting each feature to actually work without that safety net.
FSD was never real. TSLA was the product the whole time. Just like solar roof and tesla semi and tesla bot… if it gets an idiot futurologist wet then it will lift the TSLA price. Schadenfreude, long over due comeuppance, etc, etc.
Good. Back when I believed in Tesla and its owner, this excited me. But the more you look into it, the more you experience it, the more you know about it and its delays, it's clear we'll be lucky to have actual FSD in 10-20 years. I think even that's optimistic.
I'm glad this is being recognised at legislative level. Objective people on the ground have known this for at least 5 years ...
One regret is that, while Tesla have forged a place for EV's in the world market, the waste of talent and time spent on FSD (and competitors trying to immitate and catch up on the promise of FSD specifically) has been tragic. I can only hope the secondary outcomes of that time are worth it.
Honestly, decent auto stop-start in very slow traffic and excellent lane-holding cruise control would have been enough for me ... and that's been available for a decade.
How times have changed since Tesla was untouchable here. What happened and why does everyone talk like they've always hated Tesla (and Elon)? Surely it's mostly the same crowd as when Tesla was loved and Elon a superstar.
It might not be. I've never liked Elon, a view that was truly cemented during the "pedo guy" incident. In the past that meant I just ignored fawning articles or comment threads about this modern "Tony Stark" as I knew any critical comments would result in a pile-on from his supporters.
Conversely, herds are gonna herd, and Musk fans are probably a lot less vocal these days compared to the days of peak Elon hero worship.
HN is a forum with a lot of people in it with different opinions.
Speaking as one of those who's always disliked Tesla and been turned off by Musk, one explanation is that the skeptics now feel more empowered to speak out than they did before. Also, depending on how far back you're talking, you may be overestimating how much unadulterated love there was for it before: I've been following conversations on Tesla here for years, and plenty of other vocal skeptics have been here for as long as I have.
I think the biggest shift is less the emergence of skepticism which was always there, and more that the people who would previously counter that Elon had always shown impeccable judgement and laser focus on big decisions, that we should trust him the promised FSD would be delivered soon or that he truly believed in the cause of free speech and would completely transform social media for the better tend to have gone a bit quiet.
And the experience is like stepping out of an elevator that a football team had farted in and you'd been holding your breath for a decade. Let's enjoy this moment while it while it lasts.
We'll back on another hype train of tech "making the world a better place" before you know it :)
Tesla has been hated on HN for years. Current sentiment is just a continuation of that long trend.
I've kept a record of HN comment threads that keelhaul Tesla and Musk since around 2015 or so. I really believed in the company's mission and execution throughout those years (and still do), and wanted to keep a record of the very widespread opposing view. This in order to document it in case I was wrong, or in case I was right and someone claimed in retrospect that success was obvious and inevitable.
I don't believe there has been a material change in sentiment.
I definitely perceive a shift in some people who were very pro-EV, pro-Tesla due to environmental concerns but became very anti-EV, anti-Tesla as it became apparent that Musk was not part of their political affiliation.
In RL or on HN? If on HN, you can be pro-EV and anti-Tesla (which is where I fall), so don't mistake a commenter taking that position with someone changing their mind.
On social media. Agreed that you can be pro-EV and anti-Tesla; I am remarking about an observed change in tone from pro-EV, pro-Tesla to anti-EV, anti-Tesla.
I think a lot of people got caught up in a sort of mass mania induced by media portrayals of a billionaire genius with a mission to change the world for the better.
Recent events have just given people permission to climb down from the collective psychosis.
There are a lot of profit incentives for corporate media and established players in a lot of industries to smear to Elon. This results in a ton of bad media coverage, and people are a product of the information they consume. Over time this style of coverage embeds itself in people's minds, and we see what we're seeing now: posters that have never met elon, never spoke to him, and never worked with him stating ridiculous opinions about his motivations, his health, and mental status as fact.
He's on display on Twitter every day all day. No need for conspiratorial thinking, people can make their opinions based on years of lying about FSD, being an absolute asshole to (former) Twitter employees, and generally flailing around since he bought Twitter.
What does meeting the guy in person have to do with anything?
It looks like my first public comment on Musk/Tesla was in February 2021 [0].
I remember getting hooked on the FSD fraud story sometime in 2019 and seeing many negative Musk/Tesla comments since that time.
There’s always been some fanatic Musk supporters leaping to defend/apologize, but at least to me, it feels like the skepticism has been strong here for quite some time.
Twitter hasn’t been exhibiting “odd behavior” for years. Systems don’t fail overnights, but I would expect some signal to emerge from the noise by now.
I don’t know that they were overpaid. I also don’t know that all the people fired had to deal with day-to-day uptime operations. If they did, then they probably were over-staffed but probably they were not and I think we all know that. Now, it seeend quite likely they were over staffed to an extent but I don’t know about this idea that “Twitter is still up so those people were useless.”
Also, Musk fired some people for saying stupid things about their boss on his platform, that is just fair play. I don’t see why people like NYT are up in arms about it (well, I do). However, read Musk’s Tweets from November. He was behaving like a little child trolling on the Internet.
He is selling waaay less ads and way cheaper then used to be. And while it has many reasons, him firing people who were actually selling those ads is one of them.
Imagine thinking that firing 50% of the staff that built an enormously complicated piece of software, and having the service stay up and running means that the people who built it were overpaid or overstaffed.
It seems that recently, it is become trendy to hate Elon Musk because of the news where as I'm only critiquing FSD as I did a year ago against the fanatics [0], since that is obviously full of snake oil and false promises and a beta 'safety critical' product deployed to hundreds of thousands of drivers on the road. I don't see that level of deception yet with SpaceX, Starlink or Neuralink.
Tesla cars themselves are also fine until you're talking about FSD which that is the problem and the scam that puts the lives of other drivers on the road at risk.
Neuralink experiments are based on old experiments and nothing new. We have to ask ourselves why didn’t the research lab that originally did those experiments move forward with them?
But the idea that this has any implications for healthy people is ludicrous. They can read a thousand neurons close to the brain’s surface. So? The understanding of the brain is incredibly limited and Musk won’t be able to access much of it anyway.
Did anything change? I was downvoted for calling out Musk as dangerous (...) unstable and unethical sociopath with so called "self driving" car a year ago[1] just as I was downvoted for pointinug out FSD is a myth just one day ago[2].
There's a reason tech is seen as the job that's "too plum to unionize." We are the most aligned with the beliefs and values of management and capital. We wouldn't have our nice jobs if we weren't.
All that's happened here is that one paranoid delusional billionaire has a reactionary conspiracy theory that the most elite bootlicking workers in the economy are secretly all trotskyites in league with the woke marxist antifa at Ford and GM and other major advertisers etc..
The collateral damage of this has caused mass cognitive dissonance for people in tech who are either going to realise that maybe robber barons don't have our best interests at heart (shock!) or just double down on the bootlicking to see if that makes papa stop hitting us and love us again.
Good. I love the features they have but you can't say it's not false advertising. Most people I know assume my car drives itself, which is not true, even when I was on the beta.
I live in Boston. You know all those stereotypes about Boston drivers and Boston traffic? Largely accurate. If they ever do succeed in creating an AI powerful enough to drive a car safely in this Commonwealth, I want that AI curing cancer and inventing time travel, not bothering with trifles like driving.
What strikes me about this is that the area he's driving has to be the easiest area to drive within like 20 miles. I'd like to see what happens if someone tries FSD say, around Powder House Square rotary in Somerville in the evening rush hour on a November night with rain pouring down. But I wouldn't like to be in the vehicle in question.
We were in San Mateo recently and took a Lyft to a restauraunt. It was a gull wing model Y, which was curious to see in operation, and also curious to see as a Lyft car. Anyhow, we got in, and were pulling out of hotel parking lot when the driver unprovoked said "I love my Tesla, check out its self-driving!" and takes hands off the wheel, right before a 90 degree turn near the hotel. It did it, but it did it weirdly, not like real driver. My wife and I were both really freaked out and later agreed that that wasn't what we got a driver for and we wished he kept his hands on the wheel.
It is very much against the prevailing HN narrative to have an optimistic view on autonomy in general, and Tesla's efforts in that space in particular. The approved HN narrative is that autonomous driving is far too complex for computers and cameras, and Tesla's FSD is endlessly crashing into everything. Oh and it's somehow more dangerous to be near a Tesla on FSD than any other driver on the road, data be damned.
Color me shocked that you're completely ignoring the merit of the post. Stay on target, Red 5. This is about Tesla selling dangerously flawed software and misnaming it. This doesn't fall under what lawsuits would refer to as "puffery" (world's greatest coffee!). Tesla is literally misleading customers in a way that can lead to death.
The idea of describing an opposing view as derangement is disingenuous because it implies you know the mental state of the person, and it also completely avoids having to discuss the merit of the argument by derailing it. It is also questionable because your post history is full of dozens of pro-Tesla/Elon posts which are mostly trolls or memes. That makes me wonder what you mean by "derangement" because I consider doing that on a discussion forum that seeks knowledge as antithetical to its existence. But since people aware of news coverage for the past decade, this is a term that was popularized in the US during the 2016 election cycle, it has been used to shut down any rational debate. It is effectively name calling. Which is the opposite of the purpose of HN and makes me wonder why you are even here. To troll or to learn or to help people learn. Again from your history, it seems to be the first option.
Whether or not someone took donations from an opposing industry, and using that to dismiss their argument is one of the classic fallacies of reasoning.
To me, naming a deadly product something it is not seems worthy of regulation, because it is lying. If you prima facie consider me deranged because of that, I do not understand your world view and would like to understand it better.
Mayra Flores says the "oil and gas industry is critical to the success of our State and Nation" and that she "will always work to help the oil and gas industry and stand firm against radical policies that would undermine it": https://www.mayrafloresforcongress.com/issues/
Being more charitable, I am not shocked by the following independent scenarios:
1. The law on self-driving was sponsored by a legislator on the Transportation committee
2. The Transportation Committee chair is heavily lobbied by - and received donations from - the oil and gas industry. The pil and gas industry is a major user of, and supplier to transportation as a whole.
Edit: I'm failing to corroborate oil and gas PAC spend on her - seems to be FUD. Her total clear-money campaign expenditure to date is under $500k, so the claim of a multi-million dollar PAC with no evidence seems suspect.
No it's not. That a different product. FSD is still what they call the "full autonomous" mode. Enhanced Autopilot is just the highway auto lane change and some other minor gimmicks.
20 years ago, the Linux kernel version was 2.5, we had the Intel Pentium III, Apple just released the Lamp, Nickelback was Top 1 with "How you remind me?" for over 16 weeks, Warcraft 3 got released and I bought my first Palm Tungsten.
This reads like another world because it was and I suspect the world will change even faster in the next 20. We will have cat like AGI before 2042.
Your examples work against you, I think. So, in 20 years we have some incremental improvements. I think ‘82 - ‘02 was a bit more dramatic than your example. AGI requires crossing a barrier that we don’t really know is realistically possible with the type of technology we have currently.
There are a lot of pop crap radio now, WoW just released another expansion, the latest iPhone has some pretty forgettable features compared to previous iterations, etc. It’s not like the world is exponentially different in 20 years.
The jump from 134,000 (82) to 9.5 million (02) transistors was a significant one, but it's nothing compared to the leaps and bounds made when transitioning to 20 billion transistors (2022). It's a staggering difference that truly demonstrates the magnitude of the advances made in the last 20 years. The 80s and early 90s were computational winters.
I think you should go back Patterson’s book. The benefits of Moore’s law ended and then what happened? I think what you are trying to suggest flies in the face of all conventional wisdom of computer hardware.
Moore's law ended? Because Nvidias CEO said so to have an excuse for their dud of a generation? ASM, TSMC, Intel, Apple and AMD are saying it's still going strong.
All things considered, there isn't much of a difference between 2012 and 2022 consumer computing. 4/8 cores, all pretty much clocked the same, maybe 8GB/16GB of RAM and 500GB of storage. At best, we optimized the process, optimized edge cases, but not it is certainly not an exponential leap.
Between 90-00, we went from 40MHz CPU to 1+ GHZ, 1GB to 100+GB storage capacity. A factor 25 for clock speed, and 100 for storage. By the 90-00 pace standard, in 2022, we would have 25GHz CPU, 1PB disk, but we topped at a few TB at best, consumer. D instruction per core went from 0.33 (1992, Intel i486DX), to 4.1 (2002, Pentium 4 Extreme Edition), to 8.46 (AMD Ryzen Threadripper 3990X), so in 20 years, we had less advancement (x2) than in 10 years (x12).
Beside that, Moore's law stated doubling of "metric" every 2 years, 20 years is 10 doubling period, so we should have scaled by a factor of 1000. So yeah, Moore's law is long dead.
Moores law has nothing to do with performance and the metric is transistor count and like my example in the previous post we exceeded your 1000 figure...
Many believe that AGI will develop independent of our understanding of the human brain, cognition, and consciousness.
I personally find this quite strange — as if intelligence and cognition can be brute forced without having a good understanding of those concepts and how they apply to humans.
Anyway, I’m hoping to do some reading on this topic in the future. If anyone has suggestions for books and papers of much earlier works please post them here. I’ve enjoyed the Chomsky interviews and I’ll probably read some of his references.
I think the rate of change over the next 20 will be slower. Mostly just because the rate of change in computing power has slowed down. AI training compute has been doubling every 6 months. But how long can that continue without fundamental improvements in hardware? At this rate would need to increase compute by 1000x by 2027.
Ever gone out in a small town or village? Not much changed in the last 20 years, it was mostly at home. Take away the flatscreens, smartphones, tablets and calendars (doh) and nobody could tell you if it's the 00s, 10s or 20s.
A lot of comments here seem to be extremely pessimistic of self driving (not just Tesla, but in general), claiming that it's "decades away".
As someone who has been following fairly closely the advances in self driving, I think most people are underestimating the current state and rate of improvement of these systems.
For Tesla, it's currently "L2+". Here's an example[1] of how good or bad it currently is, in a fairly easy environment. Just a year ago it was an absolute mess, even though it still has a long way to go, the improvement rate is really good. In places like Manhattan it's a mess, but I don't think FSD needs to be able to handle Manhattan to be useful, because not everybody lives in Manhattan.
For Waymo, it's currently L4 (operating without driver), however available only in very select locations. Here's an example[2]
Regarding vision vs lidar, anybody in the field knows that both approaches are viable for self driving[3], it's just a matter of which can get you there sooner (and LIDAR is basically agreed upon to be an easier but more expensive approach).
I used to think Tesla FSD was doomed to fail, computer vision good enough for self driving was more than a decade away, but I've changed my mind since. I now think vision based is viable, and really important because it will enable self driving for mainstream cars (not just expensive robotaxis)
From what I've seen, Tesla's FSD perception is quite good already, and the majority of the times FSD messes up it's because of it's planner, not because of perception. And that's considering it's cameras and computers are quite old already, when they decide to upgrade them it will lead to a pretty large improvement to perception.
My current estimate is that Tesla will achieve L3 in 2 years, and L4 in 4 years. It will require a HW upgrade (cameras and computer).
Edit: Instead of just downvoting, I'd appreciate if you left a comment pointing out what you disagree with. I'm very interested in this technology and other people's thoughts.
I just wish Tesla would focus on this stuff without hyping an AI feature that can never happen.
It's like what if somebody invented the world's greatest backpack? It's bigger on the inside than the outside, and it incorporates negative gravity so anything you put in it becomes 50 lbs lighter. And all that was true; it actually worked? But for an extra $10k, the company will add on the "anti-grizzly bear" feature which makes you invulnerable to grizzly bears while you're wearing the backpack. Except most of the people who buy the anti-grizzly bear feature end up getting eaten by grizzly bears. Whoops! Just quietly stop advertising the anti-grizzly bear feature and you'll still have a great product.