วีดีโอ

It's like saying 20 years ago "Smartphones are impossible because no one did it yet!!!", but like, is it tho?... Waymo has a driver too, it gets rescued constantly by a remote driver when it gets stuck.

@@Zripas it's not like saying that at all. It's saying that Tesla does not have a fully autonomous vehicle. Because they don't, and haven't actually shown that they do. FSD requires permanent in person human supervision, just like all of their cars.

@@dillchives Correct way to say it would be that mass produced cars which are being sold to regular customers doesn't have access to Level 3+ autonomous system, this doesn't mean that Tesla doesn't have one. Robotaxi unveil at 8/8, so its less than 3 months away which would indicate that Tesla atleast have autonomous prototype done at minimum. Your point is really weird and quite pointless. And FYI autonomous driving would be where car drives on its own, this is not dependent and doesn't require for people to not supervise it, it can still do autonomous driving even if its being supervised by 50 pro race car drivers at same time. What you are trying to do here is to split hairs and to say that just because people have been inside car while it drove autonomously that this now doesn't count for some reason... Well, Waymo has supervisors over internet too and a passenger in the back seat, does that mean that Waymo did 0 autonomous driving too?...

@@Zripas Tesla doesn't have an autonomous vehicle, and has NEVER shown they're capable of producing one. There will be no robotaxis, and FSD does (and will continue to) cause an accident on nearly every trip it takes if it is allowed to drive by itself. It's a poorly thought out attempt, and incredibly unsafe execution of it.

@@dillchives Again... How do you define "autonomous vehicle"? Is it one who does everything 100% and never fails at its task? By basic usage, car which can get you from point A to point B involving atleast one turn can be labeled as autonomous. But it looks like by your standards if car can't go across entire world without a failure then its not autonomous?... Like, how high is your standard here? Robotaxi will be unveiled at 8/8, this year. So there will be one. At best you can only argue on how good/bad it will be, but there will be one, just like Waymo has one, Tesla will have one too. Its not really up to a debate here... Causes accident on nearly every trip? Then how the hell I can see dozens if not hundreds of videos with FSD from same person if this car gets into accident on nearly every trip?... Something doesn't add up... how do you define "accident"? Doing something you didn't liked? Crashing into wall at 100km/h?

How? majority of entire production industry is automated, majority of things you own is because we have automation, replacing taxi drivers with robotaxi is a good business model as people still need to get from point A to point B without owning a car and for company to not have a need to pay wages for all those taxi drivers is a huge money saver.

@@Zripas because you can't even automate car making itself, when it is a fixed process, stays in place, and easily stopped if something is terribly wrong. Driving is mobile, highly situation dependent, and interactive. Manufacturing is partially automated, while you can't partially automate a taxi ride. And frankly, an end user should not take a waymo ride. When it goes wrong, one's life is in danger. For manufactured goods, the problem is dealt with before it reaches the end user.

@@CO8848_2 You are trying to make some really weird comparisons here. Manufacturing is automated and flawed products can reach end user which would threaten their lives too... Autonomous cars are coming, there is no rational reason to think otherwise. We already have cars which can drive from point A to point B on their own, and its basically just a beginning of all of this. In 50 years people will look back and wonder why someone wanted to drive cars manually when you can just hop into autonomous one and get to your destination with zero problems. There is no way to stop progress, autonomy will be a thing, it's only question of when.

@@Zripas manufacturing is not automated. That's why it has moved to Mexico. It's called labor cost. Neither can driving, that's why waymo can only work in very specific locations. Flawed product reaches consumers, but these products are not pretending to be human. But self driving is pretending to be human. That's why you fail to see the problem. You think driving is just like a calculator, while it is patently infinitely more complicated.

@@CO8848_2 Again... extremely weird attempts to do comparisons here... Rootaxi is a service, not a production. Waymo cars could be made in different countries to save on production costs, but actual robotaxi costs basically the same amount. And its limited to specific locations because Waymo is technologically limited, not because its limited on a cost side... You sound like someone who never in their entire life watched video of AI performing things... There is a video of 2 AI's talking to each other, google that one out, watch it and try to realize that in few years (if not even now) you will not be able to recognize that person you are talking with isint actual person, heck, I could be an AI and you would never know... Go watch Tesla FSD v12.3.6 videos, it already tends to behave human like, you can't even imagine what it will do in another 5 years.

Because speed limits are bullshit

Because there is a difference between rule books and real world. Just look at that full stop at stop signs rule, its stupid, if you have clear view to all directions and there are no cars around, you will just slow down and keep moving across, while by the rule book you have to do full stop at it. Same with speed limits, we have hard speed limit of 50km/h in a city, but not all city roads are equal, some require less of that speed, some could benefit from more of it, as a human you will just go with the flow of traffic which can be faster than actual speed limit.

When I rode In Waymo In LA It was actually quite fast and way safer then any human!

@@blessguy5330 Lies, there is no going fast in LA, especially with Waymo driving

"Waymo is built from the ground up to be a self-driving taxi service" This sentence makes no sense when you look at cars Waymo is using... They are using existing mass produced cars and retrofitting those with their things, its not doing from ground up, its retrofitting to existing things.

@@Zripas I take your point, and concede that the Waymo cars are based on a Jaguar I-Pace they are so heavily modified and have at least $100,000 of "extras" and modifications that they are effectively a different car; just as the Audi R8 and Lamborghini Huracán shares the same platform and engine. Waymo vehicles are therefore designed to be full time self-driving RoboTaxis and never private cars. Whereas Tesla, have always promoted their mythical RoboTaxis as primarily private cars that can be used as RoboTaxis. Inside and out, you can instantly differentiate between a Waymo, from a private Jaguar I-Pace, in theory you would not be able to instantly differentiate a Tesla RoboTaxi from a private Tesla.

@@martincday007 But that's the thing, at the moment Tesla FSD was made to be used in a generic Tesla cars as additional feature, while this year 8/8 we should get reveal of actual Tesla robotaxi which should be easily recognizable from any other car as that one will actually be built from ground up to be robotaxi and not some weird looking retrofitted as Waymo currently does. " just as the Audi R8 and Lamborghini Huracán shares the same platform and engine." Difference is that those are properly built as a cars which should do, well, car things. Waymo current cars contains front row seats and steering wheel which are not needed for this car to be a thing, aka its a crude retrofit. All of the sensors on those cars are just glued or bolted on... For example ZOOX car is made from ground up to be autonomous, this is what Waymo should have done. " in theory you would not be able to instantly differentiate a Tesla RoboTaxi from a private Tesla." Is it good or bad thing tho? What I would agree on is that cars which are being in autonomous mode should have some sort of light indicator, like a blue or green one or something which would even indicate what type of autonomy is being engaged to have a better idea how it will behave on the road. But its just a convenient thing to have, not a requirement.

HW4 is in emulation mode dummy

Same thing, in general cameras can see better than humans in dark environments

its night ride share videos out there in san fran too!!! its pretty dope

Except the Waymo needs a remote driver on demand. So Tesla could do the same thing to get out of sticky situations, assuming that situation isn’t a crash.

That’s a good question. Let me look.

It went off because it was driving too fast at a car that was still turning

Tesla FSD is made to work in any place on any car with basic cameras, and it does work quite well, is it perfect? No. But its getting there. While waymo isint even scalable option to begin with... Its a car running on virtual train tracks which is limited to specific few cities and specific tiny areas in those cities. Why? Because it doesn't scale...

@@Zripas It is a fact that Tesla FSD does not work. No one is saying it does - not even Musk. After 7 years of failure, Tesla FSD is still pathetically stuck at Level 2. It's not getting anywhere. In Chuck Cook's very latest video, it dangerously, potentially fatally failed over and over because Tesla FSD does not work. You are also wrong about Waymo. It has been an actual working and safety authority approved robotaxi for over 2 years and has already scaled to multiple cities while Tesla FSD still does not work anywhere. Tesla FSD will also never be approved by any safety authority because of its complete lack of alternative back-up sensor suites like the successful, working Waymo system has. Musk's own engineers told him this years ago and they have been proven 100% correct. It's not a question.

@@phillgordon7523 " It is a fact that Tesla FSD does not work" In what sense? Because Tesla FSD can get you from point A to point B without any interventions from you. It's not a 100% consistent, but it does work. "Tesla FSD is still pathetically stuck at Level 2" You do understand that those autonomy levels are only indication of who takes responsibility in accidented and how much attention passenger/driver needs to give? You can have level 5 capable system and label it as level 2... No one is preventing that... You could even have level 1 capable system and label it level 5... It would ruin your company in no time, but you could label it that way... That level isn't indication what it can do, its only what company decided to label it as. Like we have Mercedes joke system labeled as level 3 which can ONLY drive on highways with lead car and under 40mph with dozen other limitations. Are you saying that Mercedes system is more capable than FSD which can navigate entire cities?

@@Zripas Completely wrong. It is a fact that FSD cannot "get you from point A to point B with no interventions". There are many, many videos from Tesla fans that prove your claim completely false. These videos, from fans like Chuck Cook, can be found in under 30 seconds with 3rd grade computer skills and show FSD failing constantly. You are also completely wrong about the levels of driving automation. The level must be demonstrated and proven to work and be safe to safety regulators like the DMV before any company is authorized to operate their system on public streets. That's why Waymo is authorized as a robotaxi and Tesla FSD can never be, because Tesla FSD fails constantly. That's why Mercedes received authorization above Tesla. Tesla's constant failure keeps it a Level 2 system. It is not Tesla's choice in any way. Musk depends on people too ignorant to comprehend these facts to keep believing his lies.

@@phillgordon7523 "There are many, many videos from Tesla fans that prove your claim completely false." And there are videos showing Waymo being stuck in empty street... What's your point? If it can't do something with 100% consistency then it can't do it at all? Because you would be disqualifying Waymo straight away... And just try to read what I write here. Tesla FSD can get you from point A to point B without any interventions. Is it 100% consistent? No. Does that negate my factual statement? No... It's like having a person who says he can do a backflip, you ask him to do 100, he does 99 and fails one and then you like "Aha!!! I knew you can't do backflips!!!!" do you see tiny issue with your logic here? "That's why Waymo is authorized as a robotaxi and Tesla FSD can never be" I guess you are the person from the future who knows how all this ends... Sorry, I thought that you are just a regular one, but you are an actual time traveler!!! Tesla never asked to be evaluated as Level 3, so I have no idea what point you tried to make here. "That's why Mercedes received authorization above Tesla." Yes, because anyone can pass adaptive cruise control as level 3... Mercedes joke system is literally adaptive cruise control with start/stop enabled with additional loooong list of limitations. Here, I can list you some of those ● It operates only on geofenced, precision mapped, limited access highways ● It operates only at speeds of 40 MPH and below ● It operates only in the same lane-no lane changes ● It operates only with a lead vehicle within 100 meters ahead ● It operates only in clear weather-no rain or snow ● It operates only with no direct sun on the cameras ● It operates only if the driver doesn’t avert his gaze from ahead for more than five seconds ● It operates only outside of tunnels ● It operates only with an internet connection and yet it counts as level 3... Because Mercedes simply decided to take responsibility for accidents if they happen, that's it. Yet it can do less than any other modern adaptive cruise control system. "Musk depends on people too ignorant to comprehend these facts to keep believing his lies." Then please, explain to me how Mercedes joke system got level 3 pass while any other adaptive cruise control is level 2 at best even tho those can do more?

Tesla hasn't learnt anything, yet. The very latest FSD is a simple Level 2 assistant that makes mistakes every day. Level 5 Full Autonomy is years away. Maybe decades.

A privately owned vehicle costs about 15 dollars per day, over 15 years. Robotaxis will take me round trip to work every day, plus trips to the health club, restaurant, grocery store, and all of that for 15 dollars per day? Not a chance! And will robotaxi take me on weekend trips, 100's of miles away, for 15 dollars per day? NOPE.

100% agree!

@@rcmgllc Cameras are HORRIBLE with depth perception, all tesla struggle with potholes. Safety is #1.

Tesla cars aren't actually able to drive themselves without causing an accident though. They've never shown they can produce a car that is able to avoid this. That's the major difference.

@@dillchives waymos stability/reliability is much better than Teslas - in the pre-scanned sandboxes and cherry-picked cities they have been working on for 13 years - outside that they still require safety drivers. But they still need far too many rescues from humans and other help to be presumed flawless. To your point, Waymo has also not built such a car. they are not doing anything remotely close to Teslas scale while playing it safe by keeping to extremely well-vetted areas - where they still screw up. It’s about interventions per mile, where Waymo leads due to their mileage being the same spots. Let’s see them drive city to city where Tesla goes, then we will see how Waymo fails. If Tesla wanted to pre-scan some city-blocks and run a taxi they could. They aren’t going for crumbs though, they are going for the big pie.

@@rickkay9548 Tesla doesn't have any cars that can drive themselves at all. FSD would get itself into an accident on its own virtually every time out. They have never demonstrated they have the capability of producing an autonomous car that can operate successfully or safely in any way, without requiring a human driver to be present and at the controls (who also has to take full responsibility for everything that happens).

That’s a great idea.

It is default legal to operate autonomous vehicles in most of the U.S, and explicitly legal in some other states, though I'm sure regulators will want to see lots of data, to keep it that way. What I want to know is whether Tesla will have its own remote operators, at least for a while, and how long it will take to get it good enough in most of the country. I can tell you that, for all the progress in CA, it has not improved here in WA in years... Also, what limits will be placed on operation in the active areas? They currently limit it in rain and keep throwing up warnings at night. It needs more work in snow too. What will they do when conditions are really bad?

Sure. You can keep your Ford. I would use a Tesla every day of the week ending in a y.

Tesla has a one month free trial currently.

Your right, but everything he said is true.

You are right, FSD Level 5 means 99.99 percent competency. Right now. FSD is a simple Level 2 assistant that is maybe 97 percent capable. Just do the math.

@@DerekDavis213 probably more like 99.99999 assuming one incident per 100,000 miles is the reliability target (which is still pretty relaxed)

@@self-drivingscientist512 99.9999? Yes you are probably right. Many humans drive every day for 20 years without an accident. The latest FSD cannot drive 1 hour on a challenging route, without mistakes that require *human* *intervention* . FSD is a joke.

@@DerekDavis213 I apologize, it should have been 99.999% assuming a reliability target of 1 incident per 100,000 miles. (1/100,000 is 0.001% not 0.00001%)

​​@@DerekDavis213FSD level 5 is superhuman and not 99% competency

There is no competitor on the road. It's not 10 years. It is as far as the regulators allow them to be.

@@wolfgangpreier9160 When the regulators drive a Tesla today with FSD, they will see it is simple minded and makes many mistakes, on any challenging drive. Regulators will *not* approve FSD, in its current Level 2 form.

@@DerekDavis213 Thats perfectly ok. Regulators do not drive FSD anyway. That is not how this works.

@@wolfgangpreier9160 Regulators are going to approve or deny something that is potentially dangerous without trying it themselves? Surely you are mistaken? Today, the very latest FSD software is a simple minded Level 2 assistant that needs human intervention every day on the road. Remember what Elon promised 8 years ago? FSD in April 2024 is such a let-down, such a disappointment.

@@DerekDavis213 Regulators work in their offices and study the data they get from the manufacturers and testing institues. Those testing institutes like Dekra, TÜV and many others test the manufacturers claims and if they are following the laws and UN/ECE regulations in Europe or NHTSA regs in USA etc.. Not one single regulator ever tries one product themselves. At least not officially. That would constitue a conflict of interests and is a perfect example for corruption. Imagine the BMW and VAG (VW,Audi,Porsche) people who are sitting in those regulator workgroups - and if you did not know, the manufacturers regulate themselves, just like Boeing - drive a Mercedes Level 3 and declare it for fit for purpose and it can use the roads as the manufacturer claims it can. Then an accident occurs. A insurance company finds the culprit that mercedes actually did an error in some weird edge case where kamikaze pigeons attacked the car and it did react poorly. Next the regulators are asked why such cars were permitted on the roads. Next someone finds that the regulators USED the cars. That is a perfect example of corruption. That is precisely what happened with Dieselgate. If you can remember. I don't care what Elon promised or not. I measure companies not by their marketing - and Elons promises are exactly that - marketing - but by their products and what they actually achieve. No one on the www - the whole wide world - is able to do what Tesla does. Will it result in success and a universally working FSD? Yes, i hope so. Does it result in a working Bot? Maybe. Does it result in a working robotaxi for the US market - i bet it will. Does it rsult in a working robo taxi for Chinese cities - Oh yes it will. Does it result in a working robo taxi for the rest of the world? Maybe, but i would not bet on it. BTW: Besides Tesla not one other manufacturer offers a level2 autonomous drive everywhere. NOT A SINGLE ONE OUT THERE. Try to find one. Go ahead. Good luck.

Just take a look at a existing Tesla taxi.- We have a few of those running in our country. The seats hold about as good as every oher manufacturer.

@@wolfgangpreier9160 That is what i meant. It is not the cars fault that the seats are dirty when used as a taxi, and especially when used as an autonomous taxi. IF (will not happen though...) Tesla will have their autonomous taxi working, then the same thing will happen there.

@@Xanthopteryx Yes, true, FSD is a hoax and Ford will prevail with their solution.

​@Xanthopteryx I doubt that if tesla owns and operates the robotaxis that they will allow it to be very dirty. If they are licensed to third parties then there will definitely be no difference.

@@robinheider414 Of course they can get dirty. How do you think you earn money? By driving people or by cleaning the car? How close is the closest service place for the taxi? They are expensive to run so not too close but you can not have too far away. Maybe 100 km? Then drive one, drop off, drive next, drop off, drive next, drop off, ... You normally drive people as much as possible until battery needs recharge, and then you go back for cleaning. Each person might go 25 km. Then travel maybe 25 km between each customer. That is 50, and you might have a range of 400, leaving 10 for buffer, so 300 left. That is 300/50 = six customers at least. And it is enough that the FIRST have spilled something and the 5 after will think the car is dirty. Right? So, it is not easy to keep things spotless because we deal with people. And also have logistics to take into account. And the more you drive in the city the more people will go in the car before charging/cleaning. And if you would operate it by yourself (as Elon lies about), then YOU have to deal with all this. YOU have to deal with disengagements and go out to "save" the car or operate it remotely, and YOU have to clean the car, and charge the car, and so on. See any problem with this?

Check your settings. Mine on offset goes slightly over usually - the speed of safe traffic flow

To help with the slow speed I generally have to press the accelerator a lot to get the speed up to my liking, but it then will usually stay there, but not always. If you haven't tried yet, press the accelerator to get up to the speed you want to be at while FSD is engaged.

There is NO comparison. Tesla FSD is scalable, while Waymo IS NOT. Tesla vehicles, potentially can drive anywhere in the U.S., while Waymo can only drive where the roads were scanned. It is Geofenced in to those areas ONLY. To scan the roads all over the U.S. is very prohibited. Those road conditions and situations needs to be monitored 24/7 or it will become a nightmare for Waymo.

I live in the middle of Europe. When will Waymo technically be able to drive me about 140km to the next airport? And when will Tesla be allowed to do it?

@@1flash3571Exactly my point. There is no comparison.

I think the point is that if Tesla is able to develop end to end driving properly, the overhead cost is not going to be as much compared to Waymo, making it more profitable from the get go.

@@trotmanCE More profitable for who? And where will Waymo and Tesla and all other competitors be allowed to operate? On certain streets? Only on highways? Only between the city and the airport? Only in the night? Are there no limits? What happens to the taxi drivers? Will they all go bankrupt? What will all those Uber drivers do? Accept their defeat and do something productive or fight against Tesla?

Tesla has *what* in the bag? Waymo can drive alone, without a human in the car supervising. Tesla cannot! And this was a very easy daytime route. Waymo is full of advanced sensors, which are much better than Tesla at night, or in the rain.

@@DerekDavis213 Clearly shows that you haven't done your homework Derek. tut tut. Waymo costs 6-10 times as much a Tesla, uses radar and lidar, which are hideously expensive and available in very low numbers. Waymo have human 'supervisors' to remote to control the Waymo when they get into an edge case they cannot get out of. Waymo can only operate on an area which has been pre-mapped to cm detail (basically they work on virtual 'rails'), this is not a scaleable or profitable solution. Tesla uses cameras a a neural brain, same as 100' of millions of human drivers across the world have proven works, with the right training. Teslas can drive anywhere which has a public road or car park (still need training a little more here granted), including highway which Waymo cannot. Teslas drive perfectly well at night and in light rain conditions. I believe they are still training to drive better than the average human in poor weather conditions, including rain, snow, fog, high winds etc. Come back to me with any further education you might need. Thanks

@@BongoWongoOG You talk about all the Waymo limitations as a robotaxi. But Tesla has MORE limitations! Waymo is operating a robotaxi business today. Tesla is not. Waymo has the advanced sensor suite working TODAY. Tesla only has the simple minded cameras. Waymo is trusted by customers, while Tesla is definitely not.

​@@BongoWongoOG I'll bet you that Tesla won't have a robotaxi by the end of 2025 and 2026. There's a million things they have to get right that they're not getting right at the moment, up to and including perception. I've seen v12 almost kill WholeMarsBlog by running him into a highway divider and in the same video misrecognize a Chinese New Year decoration as a stoplight (LbCGAN6Pk_c). It can't see chain-linked fences, tries to drive right into them (tRGoEN0O5K0 @ 46:50) . I've seen videos of v12 randomly turning into stationary cars . It can't do Chuck Cook's unprotected left reliably, it pushes its nose out into traffic dangerously as it just plain can't see without doing that and Mr. Cook disengages (--qBme0ddng). Pre-HW4 Teslas have no compute redundancy. A bug that crashes the software will ... not be good for anyone ... and I've seen videos of FSD bugs crashing the software entirely (supervised this usually turns out ok). Let's hope HW4's redundant node will not be cannibalized like they did with HW3. It's pretty bad with road debris (GbiQ_uopCXI). Also, any robotaxi will need remote supervision and support for the near future. There are too many odd scenarios these cars get stuck in.

@@DerekDavis213 Waymo operates a limited service in limited locations, supervised by people (remotely), using 3rd party vehicles equipped with expensive unnecessary sensors and is completely unprofitable, losing approx 600 million last year. I’m not seeing the advantage? Can you elaborate how Waymo will compete against a flood of Tesla Taxis entering the market at scale? Good luck with that

Then you are feeling very happy today ;)

" Drop a Tesla in an unfamiliar city and it will just self drive as usual" Could be interesting if they drive on the left.

If you look at the rate of 'fixing stuff' that has been happening in V12 I suspect that within a couple months the mistakes will be down drastically. They are hiring like a 1000 people whose job it will be drive wherever there was a mistake - so the new data generated will fix that.

@@michaelnurse9089 That subscription and overall price drop for FSD could be due to lack of specific videos of specific use cases, just to get as many as possible as they are not longer compute restricted, which means they lack actual videos to be processed at the moment.

@@Zripas It is for all those EDGE Cases, and unique situations.

@@1flash3571 To get enough videos for last 1% can be tricky and in some sense, they need those as soon as possible or they will lose money

it would have been nice to have actual map on the side with icons where each car is, as it's really hard to see anything what's going on and how each car handles specific spots.

Very hard to see much of the road in either car's video. Frustrating since it's a nice concept, comparing multiple semi autonomous systems on the same route.