I have used Tesla FSD since 2018. It was scary in city driving but doing good in freeway. And it improved about 1 to 2 percent bettter with every update but still not good enough for me to drive in city street confidently until the last major update which they switch from commmand code to Artificial Inteligent technology. And this is a game changer, it drives really good in city street right now and i use it everyday confidently.
It will never be good enough to drive in downtown conditions. Just thinking about the construction zone I drove through yesterday and the orange cones were so ambiguously placed.
@chm2 Instead of spending all this time "thinking" why don't you take a ride in one through the cones to see for yourself how it reacts. That's exactly what people with FSD do everyday with new construction sites. They aren't hiding from them. Plenty of videos showing it navigating construction cones. There are people submitting videos of failures to help train the AI for weird scenarios it screwed up in. It learns from example.
@@someuser7501 No thanks. I don't want to pay to be their part-time employee. If they want me to risk my life to test their product, I should get paid for it -- not the other way around.
I test drove a Tesla last month and the Tesla rep put me in FSD and took me to a destination and back to the sales center and I was sold. Now I am waiting for my Model Y delivery. And now with ASS, it's getting better. I can see myself using FSD everyday and can't wait to watch it get better with each update. Looking forward to 10/10 event and taking delivery of my MY.
I used to hate FSD. More importantly, my wife hated it even more because it felt too jittery. However, the most recent update, 12.5, has been fantastic for both of us. It’s the first time where we both feel comfortable having the car do its thing.
@@Chris.Davies 12.5 is a huge step backward on hw3. Stopping on green light, stopping in the middle of the street. It’s overall horrendous again like all prior FSDs before 12.3. 12.5 is unusable unless you have it on to tinker with it.
I am very happy with FSD supervised. Last weekend, I had the car drive me from my home (long Sault, Ontario) to Syracuse, NY, and back. I only took control at the border crossings. My 2024 Model Y (with FSD ersion 12.5.4) drove me back all the way to my driveway. And I used summon to park it in a tight parking shelter. Without FSD, at my age of 75, I couldn't do the trip in one day. Yes, I had to touch the car controls every now and then when I was reminded by the car to pay attention. But hey, I was relaxed, and the car was handling all the lane changes, the steering, the distance, and the speed keeping. It was doing it all by vision. My return trip was all at night. I had some messages that FSD was reduced because cameras were ocluded but, FSD still worked and brought me to my home safely and confortably. Thank you Elon, and thank you Tesla. Keep the good work, and keep forging ahead. One day, we will have real FSD.
@@user-ns7ip8qy9l Do you live the History channel? Firstly most owners like myself pay a $99 subscription. Btw the rrecent updates are amazing and since 12.5.4 and as long as you Tesla has a cabin camera then no hands are required at all.
I have this feature but I almost never use it. I feel like it's more stressful to supervise FSD than it is to just drive. One major issue is potholes in NYC. But tbh I haven't really tried the latest update. The pervious ones made too many errors to be trusted. I have to learn to trust it again.
Cruise, Waymo, Zoox, Wayve, Mobileye, Nuro, all get better with every update. The question is how far they are from being able to go hundreds of millions of miles in all conditons, 24/7, exceeding human safety levels by about 10x. FSD is a long way from that.
No it doesn’t, I’m in forums / groups specific for FSD users and testers and plenty of people say the latest version is a big regression from previous versions they were on.
@@Destorrrrr Exactly have shills for tesla and tesla isn't even testing anything to prove it. Unless you call human testers as testing like in parking lots. lol. sad. Meanwhile, china kicking our ass and they won't allow their cars here that are better.
Sure thats expected. But its also only one part of the problem. As the video mentions Tesla is missing several key things that software can't fix if they are actually going to try to create a robotaxi Uber like service. They mention several of them in the video. And the sad thing is that current drivers will not have hardware to make doing this possible. Despite having paid for it already. So they are either forced to pay for an expansive update or buy a new car.
I tried the free trial of FSD on my 21 Model Y, even with the older hardware chip and cameras, it did surprising well with very little interventions. It has a long way to go but with MOJO coming online, the progress will speed up. Remember, all the other companies are having issues too and can only drive on pre-mapped areas, Tesla FSD can drive anywhere, a MUCH harder task!
I have had FSD and I'm surprised with the amount of optimization with every update. After seeing so much improvement and safety I do not doubt that in a few years, the system will be safer than 90% of all human drivers on the road.
I avoided expressways for over a decade due to panic attacks. This is the reason why I bought an EV just over a year ago. I decided on a Tesla Model Y and couldn't be more happy. I am back on the expressways and save 40 minutes of drive time to work a day. It is impressive the improvements I have seen in FSD in the past 12 months. It is now picking me up in front of grocery stores and other public places. The look that I see on people's faces is priceless.
But all these idiots who haven’t experienced say it doesn’t work. Ignorant moron trolls. It’s off the charts amazing. Took my cousin on a 75 mile drive over the weekend and he was freakin out the whole drive. He couldn’t believe that the car drove the whole way without me touching the wheel once! He even texted me twice after still in utter disbelief. Now calls my car Christine.
@@twothbeave The average driver goes 120,000 mile between accidents. Tesla Data shows 360 miles between "Critical intervention". So according to Tesla's own data the system is 333 times worse than a human.
It is one of the most challenging endeavor. Why not respect and Be happy an American company will most likely succeed at it first. And once again take on the world. 😊
Took delivery of a Model Y the end of July. Thought FSD was a gimmick. It’s not. As someone in their mid-60’s, I’m done with driving if the car can drive itself. I push the right roller wheel, tell the car where I want to go, pull down the right stalk and relax. The only problem are things where other drivers don’t do what they’re suppose to do. Like yesterday, an ole boy pulls on the interstate, I’m in the left lane to pass him but he forgot to turn off his blinker. The car doesn’t know what that car is doing. I completely understand the logic being on the spectrum myself the car thinks exactly like I do about things😏
That’s the thing, if only Everyone was using a self-driving car, we’d be fine. Plain and simple think about it. The cars could communicate and drive Around each other better than stupid humans ever could.
@@AznPrzsn I 100% agree. I am surprise there is no legislation in place to have a common language where all manufacturer must use so the machines can communicate effectively with each other.
Yeah one of the most annoying things about fsd is when youre stuck behind someone who forgot to turn their blinkers off. And there are multiple other scenarios where human drivers do things fsd does not react well too. Honestly i dont think they will ever do.
@@f50elite He is literally building the Semi truck factory as we speak after Pepsi showing impressive results fro the trucks they are operating. Hyperloop is not where he is focused.
@@f50elite hyperloop was a concept, he didnt say "we're building it soon" or anything like that. it was mostly an idea he put out there for someone to hopefully build.
YOU SHOULD TRY THE 12.5.5 branch. It's completely hands free, barely requires you to watch the road and is simply amazing. Don't bet against Elon and Tesla
Lol bro. I have it as well and its not "there". Cope harder. Its impressive, but not there at all. Which is probably why they have built a dedicated platform for it. Which means YOUR car will likely never be a "robo taxi", which is something I've assumed all along, but rather it will have an impressive driver assistance feature.
@@ultrastoat3298 You are taking this wery lightly but many will not! Supervised is not what was promised and what people have paid a lot of money for! I knew from the beginning they were lying, but many take everthing serious what comes from Elons mouth! They will sue Tesla!
Just got a Tesla, the new fsd software is great. It moves over slightly when you pass a semi which is something I do naturally. It slows down slightly for a railroad track. The little human things we do when we are driving is being built into the software. We only intervene to avoid bad drivers, but it’s getting better at dealing with that too.
FSD is incredibly impressive and with the latest updates I use it often. But you absolutely have to be ready to take over. It does make silly mistakes 😊
I’m so excited to see genuine self-driving cars in the near future. I think it will be an unbelievable technological achievement. My wife and I could own one car if my car could drive back home after dropping me off at work in the morning.
Tesla is using vision based tech for a good reason. They are building a general vision framework that can be implemented into all AI products that are required to navigate around the real world, such as humanoid robots. One system that will ultimately allow AI (and not just cars) to see.
Sounds pretty negative reporting but I have FSD and the latest versions are amazing to me. This is the hardest software application ever attempted to conquer and it takes time.
I've owned a Tesla for over 5 years w/ FSD. It's definitely not perfect, and I hesitate to use it much in the city, but on long trips it makes the drive so much more relaxing. That alone is worth it.
I have 2 teslas and FSD is a lot better now than myself driving, and the good thing is it continue to improve. Book a drive test now and see for yourself.
@@Kodakcompactdisc as a person that is new to EV, I have FSD version 12.5.4 and I have to say FSD drives safer and better than most drivers on the road.
Yeah well they said these things like tesla couldn't be successful as an EV company and here we are. Of course they must protect their investors and instil doubt into consumers. FSD is great and forces the market to shift in attempt to keep up. But we all know there is no real competitor here. Tesla is leagues ahead.
Like it or not, the FSD technology is miles ahead of the competition in offering autonomous driving in a passenger vehicle available for purchase today. The latest update doesn't require you to hold the steering wheel anymore either as long as you keep your eyes on the road.
Tesla still requires you to supervise the car, and while they have made advancements over the years, people still have to take over way too often to simply say that this is ready for an autonomous vehicle. And as mentioned in the video: no, current Teslas are not designed to be autonomous taxis. Even if they could drive autonomously one day, you are still lacking so much functionality.
I was out in silcon valley a few months ago. I finally test drove a tesla. It was super fast, very quite, very comfortable but no life and no soul. I prefer gas. I did try the auto pilot, it worked perfectly. I was very amazed at its ability to completely drive without any interaction from me. I even changed the address and it did a legal u turn drove me right to what ever address I added. Amazing vehicle. Unfortunately I love the soul of gas. Not for me but my daughter loved it and of course she prefers the car to drive for her.
To expect CNBC to report anything with full transparency and honesty is impossible - legacy media at its finest. I just got back from Abbotsford all the way to (Yaletown) Vancouver with 0 intervention. Yes, it's supervised (for now), but the feeling of enjoying the ride is something that can't be explained to anyone. It's exactly like trying to tell a friend the funny scene in a TV show; no matter how a person tries to explain the scene, it doesn't have the same effect. It to be experienced to fully appreciated.
I’ve been driving a Model Y for the last 3,000 miles, using FSD 98% of the time, starting from v12.3.6 to v12.5.4. I’m completely addicted to FSD. Its quick reactions and 7-camera alertness make me feel extremely safe. It’s a dream come true - low-stress driving. 🙏🏽 I’m definitely an Elon fanboy. He’s probably an alien: SpaceX, the boring company, NeuroLink, X, and Tesla
CNBC obviously does not agree with Elon politically. This video has a noticeable negative slant. I own a model-Y. It’s awesome and FSD 12.5 is amazing. They are very close to Robo-Taxi. Also factor in Tesla Energy, humanoid robots, FSD licensing deals and this stock will 10X in 10 years.
Not at all, MIT professor point out that FSD is only SAE level 2 that should not be called robo taxi. Current camera on Tesla models are really unreliable at night or dusk . Without LiDAR or ultrasound, how Tesla could achieve Full self driving is mystery.
@@chang-kp9sp have the sun right in front of you, low above the horizon. Good luck with vision only. One of the first fatalities with Tesla's Autopilot was because of such an incident, but what was worse was that that car was equipped with radar and Autopilot decided to ignore the radar and give priority to vision which to this day is mindboggling to me. If one source says: there's an object in front of you and the other says there isn't. Why not play it safe?
Although recent, your examples are version 12.4.3 and current is 12.5.5, much improved. You edited a quote from Elon such that it completely changed what he said. Elon stated that he hopes Tesla will apply for robo-taxi license this year in a few States, but then you cut to another statement where he says they are applying for permission to market FSD (Unsupervised) in Europe and China, and he expects to get approval early next year. He did NOT state when he expects to get approval for robo-taxi. You make it look like Tesla is applying for robotaxi certification in Europe and China and expect quick approval. You also quote Reuters' statement that the low cost car has been cancelled, but that was denied by Elon. The denial was within 24hours and well known, but you deliberately ignored that. I do agree that it was foolish of Elon to think that autonomous driving would be easy to solve 6 years ago. They still have a way to go, 2025 could be the year FSD is certified for SAE Level 3 (private) and 4 (robotaxi). Level 5 is many years in the future.
I took my Tesla with FSD out last week and my street was blocked off with 5 signs because the street was closed due to an fair that was going on. The care advanced close to the signs, turned left, turned right, turned left, turned right, and I had to disengage. I also have yellow sidewalk lights in my neighborhood and when I stop for stop signs, the car wont start again on its own. Until these issues are resolve, it is not street safe without a driver.
@@leothepeo He does not tell us the year make and model of car he supposedly owns or the software number he is using.. A 2014 Tesla with Mobileye FSD would have those problems. I have a 2024 TM3LR with 12.5.4 It is amazing. I live in STL Missouri and it does 99% of my driving. All over the area. Several trips a day. Very rarely have a problem. When I do, it is usually something silly like not cutting off people to make an exit that requires you to cut people off to make. The system wants to be a good driver. Not possible. The road design does not allow it. His "I can't use it because of one edge case that nobody else in the world will ever have." is just BS. That is why, bad driver. Clearly he does not understand what FSD is, where it came from, or where it is going. Amazing. Every day it is Amazing.
total side note, somewhat a nerd about this but why are they showing the really dated versions of the FSD/autopilot in the beginning of this video? I get in the start of the video it was based on the past but when they moved on, they still were showing old UI and such? They should've showed the new FSD and A.S.S.
Are you telling me the the "new" version is better at killing the owners of the car better than the old one. Musk ODOR is all about monies and they used everyone to that end just like his love DADDY the Orange BLOB!
@@WrathChild-NZ Waymo operates SAE Level 4 robotaxis in geofenced areas, which is the best Tesla would be able to do as well. Tesla currently operates Level 2 everywhere and robotaxis nowhere.
As someone who recently had 30 days of FSD as a free trial, I am extremely curious to see how Tesla will make FSD work on a taxi yet not function well on its cars and trucks. I was seriously underwhelmed by my Tesla FSD trial, it was mostly annoying and did not meet my expectations at all. I expect some nasty backlash from Tesla owners who paid big money for this ability to see it accomplished on Tesla's robo-taxis but not their personal vehicles.
FSD is impressive, my problem with Elon is the massive overpromises and massive underdelivery. Been following the man for a while and FSD is far from the only thing he's either overstated the capabilities or completely lied about them. He did a Ted talk in the last year or two where the host was probing about the robot. He was asking about random household tasks and wether the bot could do it, all yeses from Elon. Can it do the laundry? Yes Can it serve Grandma her tea? Yes Just like how he has been claiming since the semi reveal that it'll be cheaper than rail, and it's something "we can do today" Idk how he can keep getting away with Therenos levels of deceit.
Because he's surpassed all his skeptics' expectations, repeatedly, in delivering breakthrough technological advancements in multiple fronts. Not saying it's right for him to get away with deceit. But that's why.
Even his underdelivery is more than what other companies and people can accomplish. Just look at the promise of interplanetary travel, while still a promise and not reality, SpaceX is still impressive and the only company that can get people safely off the International Space Station.
If you’ve never tried FSD you have no business talking about its capabilities. I like it - it’s not perfect but it’s far safer than the usual distracted or drunk driver. Robotaxi is a matter of when, not if
I do think Tesla made a mistake by removing ultrasonic sensors and radar. They should be adding more sensors and cameras, not removing them. But I do also think Tesla FSD has improved significantly throughout the years. Maybe not full robo taxi, but it has improved and less scary to use.
Good objections: Items left in cab (Tesla will notify via app), doors not closed (will be required to "end trip" I am sure), ....or notify passenger to CLOSE door, "Charging".....well, there is already Tesla and other companies that has wireless charging of EVs - which I am sure will be shown 10/10.
If it were up to people like you we would have never gotten past the stone wheel. I have a refreshed Model 3 Tesla and it drives better than I do, and I am an excellent driver.
When Musk needs money he overpromises to inflate the stocks, and eventually when he underdelivers the stocks plummets but a bit above the initial value, obtaining a net gain.
@@DavidDedmon-pw8jz So stating a FACT about Musk automatically means that you align with a particular candidate? So problematic on many levels. I miss the days where we could give opinions without automatic political associations. Believe it or not, you can have a positive/negative opinion on someone and it can have absolutely nothing to do with who you vote for. Shocking in 2024, I know.
The majority of my interventions are for the timing of lane changes and to increase speed through intersections which are not safety critical but more for optimal route planning and speed. The lane choice is mainly a mapping and memory problem. Once they solve this problem, we are basically there
Lol. Of course not. All hype designed to pump stock. For it to increase Tesla’s revenue significantly, it would have to squeeze even more money out of people which just not realistic. Ride hailing just another tech bro con. It is not profitable nor are people making a lot of money on taxis. Anyone claiming otherwise, just doesn’t know what they are talking about.
I love it when people say others don't understand what they are on about, then fail to understand the huge leap autonomy will make. If you think this is simply about Taxis, you are seriously mistaken.
@@daydreamer8373 Obviously you know nothing about transportation. If you did, you would know AVs won’t be anywhere near as beneficial as those hyping them are saying. You need to realize as with most tech much of it is just a con to get money from investors and the government.
Is everybody poor at CNBC and driving old Pintos? Why do you need to go out and find a Tesla owner?? The Model Y is a top seller. Next are you going to go out on the streets and look for an iPhone owner to find out how they work? There has to be someone at CNBC with a Tesla you can pay $99 to have FSD turned on and you can play with it for a month.
So many unqualified people debating an engineering debate on whether or not VSLAM is better than LIDAR (the “larger array of sensors”). In engineering, counterintuitively, *less is more*. And a sensor suite that uses AI for depth perception (contrary to what the Guidehouse Insights guy said) that *costs less* is obviously more scalable and more likely to win market adoption because it simply costs less to put in the road
The issue is that the depth perception from stereo is more computationally demanding, less accurate, and also fails to offer redundancy. Software glitches, camera blockage, and damage from an accident can render cameras useless which is why LIDAR is useful to compensate for this.
@@TheSpartan3669 Tesla are proving it working, and are making huge strides towards solving automation. I have no doubt Tesla are very close to solving self driving, when they do, I don't see how Waymo can compete.
@@WELVAS. That Tesla's FSD is falling behind others. When in reality it's the closest to FSD and the most scalable. I wish I had the ability to properly express my next point, but I'll do my best. Forgive me for being a bit drawn out with this. I hope it makes sense. When you hate or dislike something/someone, it's easier to take a video like this at face value. Some of it is true, so it's not entirely a bad thing to do. But the problem you run into is, people tend to stop at what they're shown. Conversely, to any fan/follower/user of the product, it's immediately obvious how much context or facts might be missing. Think of something you know extremely well. Could be something related to your job or a hobby you love. If someone else comes up to you and criticizes this thing you're knowledgeable on, it's immediately obvious to you what they're missing or that they don't understand some of the deeper moving parts. This is what it's like for me and others watching something like this. It's immediately obvious how shallow it is.
@@alexhein6054 The rate of improvement in 12 alone suggests it will be much sooner. That is, the rate of improvement compared to V1 - V11. It's kind of insane...
Musk saying lidar much too expensive. But price keeps falling rapidly. And if on robotaxi and over 10 yrs of use, the original cost is next to nothing in per mile cost.
@@jbenny9851 but the governmental bean counters on approval will require for next few yrs for approval for use on robotaxi. Yes is sad but that’s the government with this same for new medical drugs. Many hurdles. Tesla should include and can eliminate in few yrs. Yes range somewhat less but at least cars on the road fast for approval.
All these pieces about being “late”…. They are working on SELF DRIVING CARS people! I think the real angst should be directed to companies who have made the same basic cars for 50 years. At least Tesla is moving innovation forward.
Not the con man saying "your car will appreciate in value because it'll have FSD by 2016. I mean 2017. Or 2018. Okay, maybe 2020. This time for sure: 2022. No wait, 2025". Its fair to say he is selling cars and at best has no idea what he is talking about.
Waymo operates in geo-fenced areas. And how much does their hardware cost per vehicle? How are on earth will they ever scale? Tesla currently has all the hardware on their vehicles necessary for self-driving. And they currently make a profit on these vehicles. How isn't anyone picking up on that?
You might want to update this "opinion piece" with the fact that the shareholders trying to sue Tesla for Autopilot / FSD promises had their case dismissed. I think the "journalist" who put this together misses the point about SAE levels completely. Tesla is intentionally staying at L2 for now for the FSD (Supervised) product, since it has way less compliance and requirements. They are busy collecting the intervention data and when ready, will request L4 / Robotaxis approval in certain cities and jurisdictions. Notice how Waymo / Cruise are slow to roll out their Robotaxis service? It is because they are incredibly brittle and need specific training from each city they operate in, let alone the millimeter accurate HD maps to give it a "cheat code" to help place itself in the scene.
I am not worry if Tesla Tobotaxi is immediately become available after10/10, but i am more concerned the doubters being slapped in face once FSD suddenly proves itself works much safer than human drivers in near term
I agree with CNBC Robotaxi experts on here WayMo is 10 to 15 years ahead of Tesla’s Robotaxi in development and that maybe in 2040, we see Tesla’s 1st Robotaxi on the streets of Austin, TX. WayMo will dominant the Robotaxi marketplace in 2030 capturing 85% of the rides.😊
Tesla has some huge advantages over the competition, anyone thinking Waymo is ahead are not seeing the bigger picture. FSD is well placed to dominate in this space.
@@daydreamer8373CNBC Robotaxi experts do not even think Tesla is in the Robotaxi game now! Has no lidar in the vehicles yet! Tesla is completely lost according to the experts here and on Bloomberg News!
@@Agent77X Well I'm not sure who these so called experts are? But FSD is improving at a rapid pace. Experts said Elon would never compete in the car industry. Now Elon has the best selling car in the world.
Overwhelmingly negative? Trying to ground expectations following what Elon promises in tech conferences and interviews is not overwhelmingly negative. It is the appropriate journalism when a person who regularly overpromises and delivers way late (or not at all) tries to convince the public of the next miracle technology. I personally think Elon is a total hack, an egomaniac, and WAY too idolized. But that bias pointed out.. this journalism is fine.. they aren't bashing Elon at all. They are taking his promises and pitting those promises against all the realities and external hurdles.
Elon Musk knows nothing about technology at all! Everybody knows that! The why all the news outlets are alway negative on him because he does not know anything!😂
Promised and missed the time - it's much better than to put something dangerous to the real road. So, "not ready yet" is not a bad thing. This is not a topic for rush.
FSD 12.5 is an improvement. They fixed the biggest issue I had was the car driving too fast around curves. But now, I have a NEW problem. The car gets confused at intersections it used to handle without any issues. It will pause and the steering wheel will rotate fast back and forth like it doesn't know whether to turn left or right. It is bizarre.
I have FSD and it cannot drive 7 miles to my workplace without making a mistake. I love my Tesla Model 3 but FSD still is not good enough in an urban environment.
FSD works fine in the Boston area..even through Harvard Square on Mass Ave which is full of high IQ morons who don't look both ways when crossing streets while walking/jogging/biking
LOL. Look, I'm a big Tesla fan and love my AP1 Model S, but if you have any real expectations of something substantial then you are a fool. If its a vehicle of some sort, it'll be little more than a prototype just like every other vehicle "release" has been. If it's a software update, then what exactly do you think has changed that hasn't been released in Beta form that will be giant leap? Musk gets hated on, but lets face it - ALL car manufacturers claim what they want with no repercussions. Every single manufacturer claimed that they'd have Level 5 by 2020 and here we are in late 2024 and nowhere close. They claim that all of their new vehicles would be electric and no more ICE but there's not a single manufacturer that is even close to that. You claim what you want and make excuses when it doesn't even close to happening.
For me with a 2 month old Model Y, the latest hardware, the over air update which included summon is incredible. Tried it the other day at the movie theater. Stood outside and watched the car pull out of my parking space and drive to me. That’s got to be part of this new era. Tesla can “summon” your car if one has signed up their car to participate.
@@kirkhilles523 Wife is the same but she is type AAA personality. I was a test pilot for years so I’m use to that sort of thing and still have really fast reflexes.
All these self driving companies are chasing a wild goose. Cities need to be redesigned from the ground up, literally. There are too many incompatible systems conflicting with each other and they need to be kept separate unless you like to live in a chaotic environment, which is what we have today…
While vision alone may not be as good right now, if they ever do figure it out, it will basically put their competition out of business because they will be able to do the same thing with a much cheaper product to produce. People can do it with vision alone, so i do think it's possible.
Dan Ives is in for the surprise of his life... In 2025 Tesla won't get approval for FSD (still decades away), won't have robotaxis, won't have a Model 2, won't have a $61,000 Cybertruck and won't have growth. During the recession, we're expecting the stock to drop under $100.
Ives will be saying in 2025 "FSD will be ready in 2026, for sure! He's deep in the Elon tribe, with no clue about the real world of autonomous vehicles.
Even if everything goes right, is the market for robotaxi, FSD, AI, etc, really big enough to justify the enormous investment so far? I’m beginning to think that they’ve overspent this long ago. The returns will come but not as high as they hope.
FSD goes way beyond Robotaxi. Building an ie capable of understanding the environment it is operating in, and making decisions about the options available, will have huge ramifications for their Tesla bot, and offer a huge leap forward in capability. Solving general ai will be a massive step.forward.
If AI becomes that good it'll be way more than just FSD. Think cure for cancer, nuclear fusion. Everything humans aren't smart enough to solve the AI will
If Musk pulls this off Tesla stock will 10x. He will pull it off it’s just a matter of time. The big advantage Tesla has over their FSD competitors is instant scale and the ability to pivot this technology to their Optimus robots. It’s a multi trillion company in the making.
@@TheSpartan3669 are you for real ? Look what he has done to the space industry. Look what he has done to the car industry. FSD is just time - something that he is prone to being a bit too optimistic about.
I have FSD. It drives better than I do. Amazing. Long time coming, but it's here now. It really is. I recommend giving it a try (the new version, not what existed 6 years ago).
We have two Teslas both with FSD "Supervised" both cars have it turned off because we don't use it. You can save it with all of the try the new version nonsense. We purchased FSD in 2019 with the expectation that the car would be able to go out and make money for us all day rather than sit in the garage all day. Until it can do that, it is not what I paid for. It also seems like the hardware in my car wont support the final version if there is one.
I tried it on my MYP 24 with the free trial. It bugged out on a long curved turn at night on a street and I had to take over. On the highway it was too slow for my taste. Not worth the subscription at least for me yet. Autopilot is good enough tho, drove from LA to Oxnard and it saved me the hassle of stop and go traffic when we caught an accident on the 101. For $100 a month it might be worth it for a long road trip and just cancel afterwards. Might consider it if I do Coachella again next year to assist on the tiring drive back home.
@@AlexJee7601 The system is not very good mathematically. A human goes 100,000 miles between accidents if you took a 1,000 miles of FSD driving how many interventions you think you would have?
I don't remember which self driven cabs, but the amount of people they had in their hq would be enough to just put them in the cars and drive them themselves, and it would be a cheaper and better service. People want self driven cars, but if you don't want to drive we already have subways, taxis, buses, trains..
6:15 Cameras do have depth sensing, just like humans. If you put 2 cameras together you make a stereo camera, that is able to see depth, just like Tesla does and how humans see depth. Even if you have a single camera, there are algorithms, that render depth, based on the movements of the camera. I have a lot of experience with SLAM and I even met the Estonian president for my achievements.
The only way to do FSD is if all cars r connected to each other and the city has cameras everywhere. Everything has to connect to each other so the cars can see everything
I'm surprised so many people haven't caught on to his scams. Tesla Q1 2024 earnings tanked and he immediately made the robotaxi announcement. I'm convinced the cybertruck is the last vehicle that will be in production since they are betting big on vapourware products like robotaxis, robots and fsd.
Dan Ives is stupidly wrong when he says Tesla will be giving robotaxi rides in 2025. Dan doesn't understand the challenge of AVs, where 99% of the drivng is easy, and the last 1% is 99% of the challenge to exceed human levels of driving. FSD is now at 99%, with a nice demo, and 99% of the hard challenges yet unsolved. There won't be a Tesla public robotaxi service in a major city by 2030.
Then why are they building those massive AI training centers? If it's going to take that long, Tesla is doomed because they will literally run out of money. Your "There won't be a Tesla public robotaxi service in a major city by 2030.." comment is declaring Tesla as easy picking for vulture like hedge funding establishments.
I have been following self driving for years (not just Tesla) And from what I am seeing, it would not surprise me to see Tesla operating autonomously somewhere in America next year. The rate of Progress since switching to neural net has been rapid, every update adding more and more capabilities. It might not happen next year, but I would not be surprised if it does.
I've said for years now that using only cameras is just dumb. I understand the motivation, it's cheaper, but cameras are still nowhere near the efficiency of human eyes, especially in low light conditiions. Even if you can ultimately do it, radar/lidar still provides more safety. Just because *humans* can only see with their eyes doesn't mean we have to limit self-driving vehicles in the same way.
@@dijikstra8 Companies are in the process of dropping Lidar due to an AI explosion. The cost of these Lidar cars is up there with Lamborghinis and Ferraris. 10 times more expensive than a potential robotaxi. Visual only is far superior for so many reasons and it's not even a conversation any more. It was a good argument before the chatgpt moment and now it's just obsolete. The quality of cameras, AI chips and available data is getting better every 6 months.
@@dijikstra8 I have no idea why you think the human eye is superior to cameras? The camera can not only see all around the car, but contrary to your beliefs can see far better than humans in the dark. Going camera only was not about saving money, it just made the most sense. Extra sensors does not make a safer car, it just adds cost and complexity.
I have used Tesla FSD since 2018. It was scary in city driving but doing good in freeway. And it improved about 1 to 2 percent bettter with every update but still not good enough for me to drive in city street confidently until the last major update which they switch from commmand code to Artificial Inteligent technology. And this is a game changer, it drives really good in city street right now and i use it everyday confidently.
Totally agree. Before the AI things were 2 steps forward and 1 step back.
It will never be good enough to drive in downtown conditions. Just thinking about the construction zone I drove through yesterday and the orange cones were so ambiguously placed.
@chm2 Instead of spending all this time "thinking" why don't you take a ride in one through the cones to see for yourself how it reacts. That's exactly what people with FSD do everyday with new construction sites. They aren't hiding from them. Plenty of videos showing it navigating construction cones. There are people submitting videos of failures to help train the AI for weird scenarios it screwed up in. It learns from example.
@@someuser7501 No thanks. I don't want to pay to be their part-time employee. If they want me to risk my life to test their product, I should get paid for it -- not the other way around.
I test drove a Tesla last month and the Tesla rep put me in FSD and took me to a destination and back to the sales center and I was sold. Now I am waiting for my Model Y delivery. And now with ASS, it's getting better. I can see myself using FSD everyday and can't wait to watch it get better with each update. Looking forward to 10/10 event and taking delivery of my MY.
I used to hate FSD. More importantly, my wife hated it even more because it felt too jittery.
However, the most recent update, 12.5, has been fantastic for both of us. It’s the first time where we both feel comfortable having the car do its thing.
Agreed! It’s awesome. I was in the video and I wish we could reshoot it with 12.5.4
12.5 is horrendous. Bring back the 12.3 or quickly patch 12.5.
@@MarvinFontanilla $1 million to have it drive your kids 100 miles without you in there? Put your money where your mouth…
It works perfectly, right up until it doesn't.
@@Chris.Davies 12.5 is a huge step backward on hw3. Stopping on green light, stopping in the middle of the street. It’s overall horrendous again like all prior FSDs before 12.3. 12.5 is unusable unless you have it on to tinker with it.
I am very happy with FSD supervised. Last weekend, I had the car drive me from my home (long Sault, Ontario) to Syracuse, NY, and back. I only took control at the border crossings. My 2024 Model Y (with FSD ersion 12.5.4) drove me back all the way to my driveway. And I used summon to park it in a tight parking shelter.
Without FSD, at my age of 75, I couldn't do the trip in one day.
Yes, I had to touch the car controls every now and then when I was reminded by the car to pay attention. But hey, I was relaxed, and the car was handling all the lane changes, the steering, the distance, and the speed keeping.
It was doing it all by vision. My return trip was all at night. I had some messages that FSD was reduced because cameras were ocluded but, FSD still worked and brought me to my home safely and confortably.
Thank you Elon, and thank you Tesla. Keep the good work, and keep forging ahead. One day, we will have real FSD.
You are so full of it!😂😂😂😂😂😂😂😂😂😂😂😂😂
@@ahmedlasfar3290 tell yourself whatever you want to justify your waste of 15k
What does Full Self Driving Supervised mean? That’s an oxymoron.
@@user-ns7ip8qy9l Do you live the History channel? Firstly most owners like myself pay a $99 subscription. Btw the rrecent updates are amazing and since 12.5.4 and as long as you Tesla has a cabin camera then no hands are required at all.
I have this feature but I almost never use it. I feel like it's more stressful to supervise FSD than it is to just drive. One major issue is potholes in NYC. But tbh I haven't really tried the latest update. The pervious ones made too many errors to be trusted. I have to learn to trust it again.
12.5 is amazing, that's the thing with Tesla... it keeps getting better with every update.
Cruise, Waymo, Zoox, Wayve, Mobileye, Nuro, all get better with every update. The question is how far they are from being able to go hundreds of millions of miles in all conditons, 24/7, exceeding human safety levels by about 10x. FSD is a long way from that.
No it doesn’t, I’m in forums / groups specific for FSD users and testers and plenty of people say the latest version is a big regression from previous versions they were on.
"getting better with every update" - that's the bare minimum requirement for tech
@@Destorrrrr Exactly have shills for tesla and tesla isn't even testing anything to prove it. Unless you call human testers as testing like in parking lots. lol. sad. Meanwhile, china kicking our ass and they won't allow their cars here that are better.
Sure thats expected. But its also only one part of the problem. As the video mentions Tesla is missing several key things that software can't fix if they are actually going to try to create a robotaxi Uber like service. They mention several of them in the video. And the sad thing is that current drivers will not have hardware to make doing this possible. Despite having paid for it already. So they are either forced to pay for an expansive update or buy a new car.
I tried the free trial of FSD on my 21 Model Y, even with the older hardware chip and cameras, it did surprising well with very little interventions. It has a long way to go but with MOJO coming online, the progress will speed up. Remember, all the other companies are having issues too and can only drive on pre-mapped areas, Tesla FSD can drive anywhere, a MUCH harder task!
its DOJO, but MOJO is also cool!
I have had FSD and I'm surprised with the amount of optimization with every update. After seeing so much improvement and safety I do not doubt that in a few years, the system will be safer than 90% of all human drivers on the road.
only if those humans are drink driving or speeding or fiddling with their mobile
👍
@@turbodewd1not sure where you live but I am perplexed at the amount of trust you give Humans to keep driving
Isn't that 90 percent of drivers?
If your video is talking about a software version that's been out for more than 7 months, its irrelevant. FSD v12.5.4 is basically robotaxi.
100%. It’s my car. I wish I could do it now but we shot this in July 2024
I avoided expressways for over a decade due to panic attacks. This is the reason why I bought an EV just over a year ago. I decided on a Tesla Model Y and couldn't be more happy. I am back on the expressways and save 40 minutes of drive time to work a day. It is impressive the improvements I have seen in FSD in the past 12 months. It is now picking me up in front of grocery stores and other public places. The look that I see on people's faces is priceless.
who are these "experts"? 😭
fanboy butthurt
exactly, whoever knows about Tesla know they are full of BS
They are on CNBS which gets money from ICE. They are lying.
noones
@@htc334similar to Kamala Harris 😂.
I just drove my brother's model Y 200 miles on Full FSD. Amazing.
Dirt lanes to Super highways. No intervention required.
But all these idiots who haven’t experienced say it doesn’t work. Ignorant moron trolls. It’s off the charts amazing. Took my cousin on a 75 mile drive over the weekend and he was freakin out the whole drive. He couldn’t believe that the car drove the whole way without me touching the wheel once! He even texted me twice after still in utter disbelief. Now calls my car Christine.
@@twothbeave what I liked was the driving at night. More relaxing especially now that my eyes are getting old.
@@twothbeave ignore the trolls, instead profit off of their ignorance. Tesla stock will skyrocket when no driver is needed.
@@twothbeave The average driver goes 120,000 mile between accidents. Tesla Data shows 360 miles between "Critical intervention". So according to Tesla's own data the system is 333 times worse than a human.
I'm actually dumber having listened to CNBC's "experts."
This report is dated. They used version 12.3.6. What a poor reporting job CNBC.
Unfortunately it was filmed in July 2024. It’s my Tesla. The new version is incredible
@@niteshnmehta Thanks for clarifying. The problem is this is a misleading, inaccurate report and CNBC should fix it.
Typical CNBC crap😂😂😂
It is one of the most challenging endeavor. Why not respect and Be happy an American company will most likely succeed at it first. And once again take on the world. 😊
Make TOYOTA 🗾 Great Again 😎 Scotty Kilmer
Took delivery of a Model Y the end of July. Thought FSD was a gimmick. It’s not. As someone in their mid-60’s, I’m done with driving if the car can drive itself. I push the right roller wheel, tell the car where I want to go, pull down the right stalk and relax. The only problem are things where other drivers don’t do what they’re suppose to do. Like yesterday, an ole boy pulls on the interstate, I’m in the left lane to pass him but he forgot to turn off his blinker. The car doesn’t know what that car is doing. I completely understand the logic being on the spectrum myself the car thinks exactly like I do about things😏
That’s the thing, if only Everyone was using a self-driving car, we’d be fine. Plain and simple think about it. The cars could communicate and drive Around each other better than stupid humans ever could.
@@AznPrzsn I 100% agree. I am surprise there is no legislation in place to have a common language where all manufacturer must use so the machines can communicate effectively with each other.
that is a good use case youcited about not turning off the blinker
Driving is an inherently human activity. Not knowing what other drivers are actually doing is a limitation of the tech.
Yeah one of the most annoying things about fsd is when youre stuck behind someone who forgot to turn their blinkers off. And there are multiple other scenarios where human drivers do things fsd does not react well too. Honestly i dont think they will ever do.
I have FSD version 12.5.4 and use it everyday, 90% of my drive, it is comfortable and safe. FSD makes driving relaxing and fun.
So literally 10 years late by his 2014 comment 😂
@@d.r.656 Elon may be off with his predictions, but he always delivers.
@@daydreamer8373oooof.. that is.. not true. Hyperloop? Semi trucks? Truly free speech on twitter? Lol 😂
@@f50elite He is literally building the Semi truck factory as we speak after Pepsi showing impressive results fro the trucks they are operating. Hyperloop is not where he is focused.
@@f50elite hyperloop was a concept, he didnt say "we're building it soon" or anything like that. it was mostly an idea he put out there for someone to hopefully build.
YOU SHOULD TRY THE 12.5.5 branch. It's completely hands free, barely requires you to watch the road and is simply amazing. Don't bet against Elon and Tesla
Ha ha ha. Wait for it!
Can't wait to rent your car on Uber in 2020!!
problem is their sales are under serious pressure from BYD...
@@turbodewd1BYD is never coming to America and Tesla is the top selling EV in china
@@TheGuillotineKing ha ha ha
As someone using FSD 12.5.5, its basically there
*basically
Lol bro. I have it as well and its not "there". Cope harder. Its impressive, but not there at all. Which is probably why they have built a dedicated platform for it. Which means YOUR car will likely never be a "robo taxi", which is something I've assumed all along, but rather it will have an impressive driver assistance feature.
So when can I book your tesla on Uber?
Never after he finds out Tesla will take 75% of the fare and he will get $4. 😅
@@ultrastoat3298
You are taking this wery lightly but many will not!
Supervised is not what was promised and what people have paid a lot of money for!
I knew from the beginning they were lying, but many take everthing serious what comes from Elons mouth!
They will sue Tesla!
Just got a Tesla, the new fsd software is great. It moves over slightly when you pass a semi which is something I do naturally. It slows down slightly for a railroad track. The little human things we do when we are driving is being built into the software. We only intervene to avoid bad drivers, but it’s getting better at dealing with that too.
FSD is incredibly impressive and with the latest updates I use it often. But you absolutely have to be ready to take over. It does make silly mistakes 😊
It is Simple Inevitable....
I’m so excited to see genuine self-driving cars in the near future. I think it will be an unbelievable technological achievement. My wife and I could own one car if my car could drive back home after dropping me off at work in the morning.
Can’t wait for 10/10 long live Robotaxi 🚖
Tesla is using vision based tech for a good reason. They are building a general vision framework that can be implemented into all AI products that are required to navigate around the real world, such as humanoid robots. One system that will ultimately allow AI (and not just cars) to see.
Sounds pretty negative reporting but I have FSD and the latest versions are amazing to me. This is the hardest software application ever attempted to conquer and it takes time.
I've owned a Tesla for over 5 years w/ FSD. It's definitely not perfect, and I hesitate to use it much in the city, but on long trips it makes the drive so much more relaxing. That alone is worth it.
I have 2 teslas and FSD is a lot better now than myself driving, and the good thing is it continue to improve. Book a drive test now and see for yourself.
No offense, but if that's true I would question your baseline. Instead of assuming FSD is good, have you considered maybe your driving is not great?
You obviously work for tesla
@@Kodakcompactdisc as a person that is new to EV, I have FSD version 12.5.4 and I have to say FSD drives safer and better than most drivers on the road.
@@PreschoolDropout673 I am new to EV and have FSD version 12.5.4. I have to agree that SFD drives safer and better than most drivers on the road
It isn't even close as good as Waymo.
Yeah well they said these things like tesla couldn't be successful as an EV company and here we are. Of course they must protect their investors and instil doubt into consumers. FSD is great and forces the market to shift in attempt to keep up. But we all know there is no real competitor here. Tesla is leagues ahead.
Like it or not, the FSD technology is miles ahead of the competition in offering autonomous driving in a passenger vehicle available for purchase today. The latest update doesn't require you to hold the steering wheel anymore either as long as you keep your eyes on the road.
If you can't drive your own vehicle, then you should walks!
Tesla still requires you to supervise the car, and while they have made advancements over the years, people still have to take over way too often to simply say that this is ready for an autonomous vehicle.
And as mentioned in the video: no, current Teslas are not designed to be autonomous taxis. Even if they could drive autonomously one day, you are still lacking so much functionality.
@@tomnguyen9931 Why not let it act safer and smoother than you
@@BenefitOfTheDoubtInquiry but its not... it acts weirdly in traffic endangering everyone.
it is far from truth. Google and Biudu is leading.
I was out in silcon valley a few months ago. I finally test drove a tesla. It was super fast, very quite, very comfortable but no life and no soul. I prefer gas. I did try the auto pilot, it worked perfectly. I was very amazed at its ability to completely drive without any interaction from me. I even changed the address and it did a legal u turn drove me right to what ever address I added. Amazing vehicle. Unfortunately I love the soul of gas. Not for me but my daughter loved it and of course she prefers the car to drive for her.
I love cnbc documentaries they are my go to relaxation 😊
they are trash and full of bias information
Ignorance is bliss
It's a biased reporting. Do you remember that CNBC said VW is winning the EV race against Tesla?
Documentaries by news media should not be trusted, it's all the same propaganda
@@DanielL-ee7fe Agreed. VERY biased. These guys need to fact-checked all the time!
To expect CNBC to report anything with full transparency and honesty is impossible - legacy media at its finest.
I just got back from Abbotsford all the way to (Yaletown) Vancouver with 0 intervention.
Yes, it's supervised (for now), but the feeling of enjoying the ride is something that can't be explained to anyone.
It's exactly like trying to tell a friend the funny scene in a TV show; no matter how a person tries to explain the scene, it doesn't have the same effect. It to be experienced to fully appreciated.
FSD is a good demo now, and impressive driver assist. It is a very long way from solving full autonomy everywhere.
CNBC who?😂😂😂
I’ve been driving a Model Y for the last 3,000 miles, using FSD 98% of the time, starting from v12.3.6 to v12.5.4. I’m completely addicted to FSD. Its quick reactions and 7-camera alertness make me feel extremely safe. It’s a dream come true - low-stress driving. 🙏🏽
I’m definitely an Elon fanboy. He’s probably an alien: SpaceX, the boring company, NeuroLink, X, and Tesla
what happens on the 2% of drives you don't use it?
CNBC obviously does not agree with Elon politically. This video has a noticeable negative slant. I own a model-Y. It’s awesome and FSD 12.5 is amazing. They are very close to Robo-Taxi. Also factor in Tesla Energy, humanoid robots, FSD licensing deals and this stock will 10X in 10 years.
Much of what you’re reporting is completely false
Mainstream media are just bunch liars. The country is going to be ruined by these liars
@@S13-w4k much what you are saying has absolutely no substance
CNBS
Not at all, MIT professor point out that FSD is only SAE level 2 that should not be called robo taxi. Current camera on Tesla models are really
unreliable at night or dusk . Without LiDAR or ultrasound, how Tesla could achieve Full self driving is mystery.
@@chang-kp9sp have the sun right in front of you, low above the horizon. Good luck with vision only. One of the first fatalities with Tesla's Autopilot was because of such an incident, but what was worse was that that car was equipped with radar and Autopilot decided to ignore the radar and give priority to vision which to this day is mindboggling to me. If one source says: there's an object in front of you and the other says there isn't. Why not play it safe?
Although recent, your examples are version 12.4.3 and current is 12.5.5, much improved. You edited a quote from Elon such that it completely changed what he said. Elon stated that he hopes Tesla will apply for robo-taxi license this year in a few States, but then you cut to another statement where he says they are applying for permission to market FSD (Unsupervised) in Europe and China, and he expects to get approval early next year. He did NOT state when he expects to get approval for robo-taxi. You make it look like Tesla is applying for robotaxi certification in Europe and China and expect quick approval.
You also quote Reuters' statement that the low cost car has been cancelled, but that was denied by Elon. The denial was within 24hours and well known, but you deliberately ignored that.
I do agree that it was foolish of Elon to think that autonomous driving would be easy to solve 6 years ago. They still have a way to go, 2025 could be the year FSD is certified for SAE Level 3 (private) and 4 (robotaxi). Level 5 is many years in the future.
No way you quote Ark's wood. Her stock is worth 30% of what it was. Even I have made more return than she has in the past 4 years
The stocks she picks are usually right. Her problem is she sells too early
I would trust Tesla any given day when compared to bull sh*t media like you guys 😂😁. CNBC you never disappoint me 😂
Elon Musk is the best con man in Silicon Valley.
buy my car. Its an appreciating asset!
@@randomnobody8770ha ha ha
He's a male counterpart of Elizabeth Holmes.
Tell that to Mark Zuckerberg 😮
If Trump moved to San Fran, musk would have competition. But now he is far and a way the biggest ConMan in the valley.
Imagine nobody does it (or fails) the way Tesla does - we have a problem!
I took my Tesla with FSD out last week and my street was blocked off with 5 signs because the street was closed due to an fair that was going on. The care advanced close to the signs, turned left, turned right, turned left, turned right, and I had to disengage. I also have yellow sidewalk lights in my neighborhood and when I stop for stop signs, the car wont start again on its own. Until these issues are resolve, it is not street safe without a driver.
Hahahaha. You got confused. Bad driver.
@@davidbeppler3032 What makes you say that these things are the drivers fault?
@@leothepeo He does not tell us the year make and model of car he supposedly owns or the software number he is using.. A 2014 Tesla with Mobileye FSD would have those problems.
I have a 2024 TM3LR with 12.5.4 It is amazing. I live in STL Missouri and it does 99% of my driving. All over the area. Several trips a day. Very rarely have a problem. When I do, it is usually something silly like not cutting off people to make an exit that requires you to cut people off to make. The system wants to be a good driver. Not possible. The road design does not allow it. His "I can't use it because of one edge case that nobody else in the world will ever have." is just BS. That is why, bad driver. Clearly he does not understand what FSD is, where it came from, or where it is going. Amazing. Every day it is Amazing.
They'll show one prototype, promise production deliveries next year... and we'll be waiting for it for another 5. By which time it will be irrelevant.
total side note, somewhat a nerd about this but why are they showing the really dated versions of the FSD/autopilot in the beginning of this video? I get in the start of the video it was based on the past but when they moved on, they still were showing old UI and such? They should've showed the new FSD and A.S.S.
Are you telling me the the "new" version is better at killing the owners of the car better than the old one. Musk ODOR is all about monies and they used everyone to that end just like his love DADDY the Orange BLOB!
12.5 is phenomenal
Agree, the nee FSD is amazing. I've watch so many videos of tesla owners being fully driven home lol
Is the B-roll really that important?
@Hans-gb4mv no, but why use old ass footage? This is supposed to be news
No mention that Waymo only operates in geofenced areas while Teslas can operate anywhere?
Yes, tesla can hit trains anywhere
@@alex_lllthey're even known to drive themselves off the road, truly revolutionary
@@WrathChild-NZ Waymo operates SAE Level 4 robotaxis in geofenced areas, which is the best Tesla would be able to do as well.
Tesla currently operates Level 2 everywhere and robotaxis nowhere.
#1. Nope. Telsa, today, cannot operate lev3 anywhere. Your statement is wrong.
#2 Tesla can't even do what waymo does in the geofanced area.
Also no mention of unprofitablility of Waymo
As someone who recently had 30 days of FSD as a free trial, I am extremely curious to see how Tesla will make FSD work on a taxi yet not function well on its cars and trucks. I was seriously underwhelmed by my Tesla FSD trial, it was mostly annoying and did not meet my expectations at all.
I expect some nasty backlash from Tesla owners who paid big money for this ability to see it accomplished on Tesla's robo-taxis but not their personal vehicles.
FSD is impressive, my problem with Elon is the massive overpromises and massive underdelivery. Been following the man for a while and FSD is far from the only thing he's either overstated the capabilities or completely lied about them.
He did a Ted talk in the last year or two where the host was probing about the robot. He was asking about random household tasks and wether the bot could do it, all yeses from Elon.
Can it do the laundry?
Yes
Can it serve Grandma her tea?
Yes
Just like how he has been claiming since the semi reveal that it'll be cheaper than rail, and it's something "we can do today"
Idk how he can keep getting away with Therenos levels of deceit.
With Elon, it's always coming next..next year...and next year. Just like cybertruck, by the time it's delivered...the price is doubled.
can you blame him? look at tesla market cap, you would lie if you can get away with it.
Because he's surpassed all his skeptics' expectations, repeatedly, in delivering breakthrough technological advancements in multiple fronts. Not saying it's right for him to get away with deceit. But that's why.
@@jakesiu7773that’s inflation for ya buddy
Even his underdelivery is more than what other companies and people can accomplish. Just look at the promise of interplanetary travel, while still a promise and not reality, SpaceX is still impressive and the only company that can get people safely off the International Space Station.
If you’ve never tried FSD you have no business talking about its capabilities.
I like it - it’s not perfect but it’s far safer than the usual distracted or drunk driver.
Robotaxi is a matter of when, not if
Tesla’s vehicles DO have the capability of closing the doors without the help of a human. Some people interviewed here are just plain ignorant.
Market research here was lacking. Such is the risk of free, spoon fed, information.
Absolutely, also I can access the interior camera from my phone. Believe me, you can see everything in the interior!
Litterally, only the model x can close all closes. No other cars can rn
@@ragnarox16 just a matter of a software update
You're actually incorrect. Only Model X's can they have door drive units in all doors. Every other model doesn't. I was a Tesla tech for 7 years
I do think Tesla made a mistake by removing ultrasonic sensors and radar. They should be adding more sensors and cameras, not removing them. But I do also think Tesla FSD has improved significantly throughout the years. Maybe not full robo taxi, but it has improved and less scary to use.
And they are all wrong. Tesla will do it.
Good objections: Items left in cab (Tesla will notify via app), doors not closed (will be required to "end trip" I am sure), ....or notify passenger to CLOSE door, "Charging".....well, there is already Tesla and other companies that has wireless charging of EVs - which I am sure will be shown 10/10.
We will be the conservative old people against mandated self driving tech in a few decades.
Lol you're white 😂
Tesla stock will be at 0 and we'll be riding around in Google waymos.
No to self driving vehicles.
They have killed and no one is responsible or held accountable.
I agree
If it were up to people like you we would have never gotten past the stone wheel. I have a refreshed Model 3 Tesla and it drives better than I do, and I am an excellent driver.
When Musk needs money he overpromises to inflate the stocks, and eventually when he underdelivers the stocks plummets but a bit above the initial value, obtaining a net gain.
Bet you are voting for Kamala.
@@DavidDedmon-pw8jz So stating a FACT about Musk automatically means that you align with a particular candidate? So problematic on many levels. I miss the days where we could give opinions without automatic political associations. Believe it or not, you can have a positive/negative opinion on someone and it can have absolutely nothing to do with who you vote for. Shocking in 2024, I know.
FSD has gotten really good. Would like to see the latest version get a little better on HW3 but sounds like HW4 is seeing better performance.
"Tesla way behind Waymoo" !!! What a BS ep.
The majority of my interventions are for the timing of lane changes and to increase speed through intersections which are not safety critical but more for optimal route planning and speed. The lane choice is mainly a mapping and memory problem. Once they solve this problem, we are basically there
Lol. Of course not. All hype designed to pump stock. For it to increase Tesla’s revenue significantly, it would have to squeeze even more money out of people which just not realistic. Ride hailing just another tech bro con. It is not profitable nor are people making a lot of money on taxis. Anyone claiming otherwise, just doesn’t know what they are talking about.
I love it when people say others don't understand what they are on about, then fail to understand the huge leap autonomy will make. If you think this is simply about Taxis, you are seriously mistaken.
@@daydreamer8373 Obviously you know nothing about transportation. If you did, you would know AVs won’t be anywhere near as beneficial as those hyping them are saying. You need to realize as with most tech much of it is just a con to get money from investors and the government.
@@daydreamer8373 you fail to understand that Tesla has been faking their self driving for years and has fallen behind.
you should buy an ETF that tracks tesla and shorts it and put all your money into it
@@apple1231230 Anybody stupid enough to short Tesla, deserve to lose their money.
I did 5 rides with Waymo last weekend while in SF and it was incredibly impressive
Is everybody poor at CNBC and driving old Pintos? Why do you need to go out and find a Tesla owner?? The Model Y is a top seller. Next are you going to go out on the streets and look for an iPhone owner to find out how they work? There has to be someone at CNBC with a Tesla you can pay $99 to have FSD turned on and you can play with it for a month.
So many unqualified people debating an engineering debate on whether or not VSLAM is better than LIDAR (the “larger array of sensors”). In engineering, counterintuitively, *less is more*. And a sensor suite that uses AI for depth perception (contrary to what the Guidehouse Insights guy said) that *costs less* is obviously more scalable and more likely to win market adoption because it simply costs less to put in the road
The issue is that the depth perception from stereo is more computationally demanding, less accurate, and also fails to offer redundancy. Software glitches, camera blockage, and damage from an accident can render cameras useless which is why LIDAR is useful to compensate for this.
@@TheSpartan3669 Tesla are proving it working, and are making huge strides towards solving automation. I have no doubt Tesla are very close to solving self driving, when they do, I don't see how Waymo can compete.
@@daydreamer8373 Do you believe 13 miles per intervention is close?
@@TheSpartan3669 I believe whatever the intervention rate. The rapid progress being made will make whatever number you quote irrelevant.
@@daydreamer8373 How do you objectively measure progress without disengagement/intervention rate?
Pivot? lol robotaixs is are a small business. Selling cars are big business. Also….Musk is always wrong always wrong about timelines.
There is so much misinformation in this..
Like?
Hardly any.
FSD tech isn’t truly here until 2030
@@WELVAS. That Tesla's FSD is falling behind others. When in reality it's the closest to FSD and the most scalable.
I wish I had the ability to properly express my next point, but I'll do my best. Forgive me for being a bit drawn out with this. I hope it makes sense.
When you hate or dislike something/someone, it's easier to take a video like this at face value. Some of it is true, so it's not entirely a bad thing to do. But the problem you run into is, people tend to stop at what they're shown.
Conversely, to any fan/follower/user of the product, it's immediately obvious how much context or facts might be missing.
Think of something you know extremely well. Could be something related to your job or a hobby you love. If someone else comes up to you and criticizes this thing you're knowledgeable on, it's immediately obvious to you what they're missing or that they don't understand some of the deeper moving parts.
This is what it's like for me and others watching something like this. It's immediately obvious how shallow it is.
@@alexhein6054 The rate of improvement in 12 alone suggests it will be much sooner. That is, the rate of improvement compared to V1 - V11. It's kind of insane...
Musk saying lidar much too expensive. But price keeps falling rapidly. And if on robotaxi and over 10 yrs of use, the original cost is next to nothing in per mile cost.
The cost of LiDAR is not just the hardware cost but the cost of energy. You will loose range with the extra power consumption.
@@jbenny9851 but the governmental bean counters on approval will require for next few yrs for approval for use on robotaxi. Yes is sad but that’s the government with this same for new medical drugs. Many hurdles. Tesla should include and can eliminate in few yrs. Yes range somewhat less but at least cars on the road fast for approval.
All these pieces about being “late”…. They are working on SELF DRIVING CARS people! I think the real angst should be directed to companies who have made the same basic cars for 50 years. At least Tesla is moving innovation forward.
Not the con man saying "your car will appreciate in value because it'll have FSD by 2016. I mean 2017. Or 2018. Okay, maybe 2020. This time for sure: 2022. No wait, 2025". Its fair to say he is selling cars and at best has no idea what he is talking about.
Waymo operates in geo-fenced areas. And how much does their hardware cost per vehicle? How are on earth will they ever scale?
Tesla currently has all the hardware on their vehicles necessary for self-driving. And they currently make a profit on these vehicles. How isn't anyone picking up on that?
You might want to update this "opinion piece" with the fact that the shareholders trying to sue Tesla for Autopilot / FSD promises had their case dismissed. I think the "journalist" who put this together misses the point about SAE levels completely. Tesla is intentionally staying at L2 for now for the FSD (Supervised) product, since it has way less compliance and requirements. They are busy collecting the intervention data and when ready, will request L4 / Robotaxis approval in certain cities and jurisdictions. Notice how Waymo / Cruise are slow to roll out their Robotaxis service? It is because they are incredibly brittle and need specific training from each city they operate in, let alone the millimeter accurate HD maps to give it a "cheat code" to help place itself in the scene.
Thank you for making the critical point that L2, L3 etc are not about capabilities of the system, but the paperwork that has been submitted.
I am not worry if Tesla Tobotaxi is immediately become available after10/10, but i am more concerned the doubters being slapped in face once FSD suddenly proves itself works much safer than human drivers in near term
They used 12.3.6? Why?
Because Waymo has won the race to full autonomy.
I agree with CNBC Robotaxi experts on here WayMo is 10 to 15 years ahead of Tesla’s Robotaxi in development and that maybe in 2040, we see Tesla’s 1st Robotaxi on the streets of Austin, TX. WayMo will dominant the Robotaxi marketplace in 2030 capturing 85% of the rides.😊
Tesla has some huge advantages over the competition, anyone thinking Waymo is ahead are not seeing the bigger picture. FSD is well placed to dominate in this space.
@@daydreamer8373CNBC Robotaxi experts do not even think Tesla is in the Robotaxi game now! Has no lidar in the vehicles yet! Tesla is completely lost according to the experts here and on Bloomberg News!
@@Agent77X Well I'm not sure who these so called experts are? But FSD is improving at a rapid pace. Experts said Elon would never compete in the car industry. Now Elon has the best selling car in the world.
These CNBC documentaries are usually really nice but anytime you all talk about Elon Musk, it's overwhelmingly negative. It's a bit much.
Overwhelmingly negative? Trying to ground expectations following what Elon promises in tech conferences and interviews is not overwhelmingly negative. It is the appropriate journalism when a person who regularly overpromises and delivers way late (or not at all) tries to convince the public of the next miracle technology.
I personally think Elon is a total hack, an egomaniac, and WAY too idolized. But that bias pointed out.. this journalism is fine.. they aren't bashing Elon at all. They are taking his promises and pitting those promises against all the realities and external hurdles.
Elon Musk knows nothing about technology at all! Everybody knows that! The why all the news outlets are alway negative on him because he does not know anything!😂
Promised and missed the time - it's much better than to put something dangerous to the real road. So, "not ready yet" is not a bad thing. This is not a topic for rush.
Tesla model Y was the Best selling carmodel on planet Earth in year 2023 🎉🏆🥇
Due to cnBS being biased to elon and tesla, they also fail to mention waymo costs $180K per vehicle while the average tesla only costs $30K...
FSD 12.5 is an improvement. They fixed the biggest issue I had was the car driving too fast around curves. But now, I have a NEW problem. The car gets confused at intersections it used to handle without any issues. It will pause and the steering wheel will rotate fast back and forth like it doesn't know whether to turn left or right. It is bizarre.
Been waiting since 2014 for it... Been 10 years and he JUST released a car model... Nevermind figured out the self driving...
I actually trust FSD more than other drivers out here in Vegas
OK then get into the back seat like you would a uber and drive around.
Thank goodness for the comments here to combat the click-bait reporting by CNBC. Hard to argue with personal experience.
I have FSD and it cannot drive 7 miles to my workplace without making a mistake. I love my Tesla Model 3 but FSD still is not good enough in an urban environment.
It'll never be ready until AI can achieve human level intelligence.. And when it does, you don't have to worry about going to work at all .
FSD works fine in the Boston area..even through Harvard Square on Mass Ave which is full of high IQ morons who don't look both ways when crossing streets while walking/jogging/biking
what version are you using?
@@MrKaneshiro1966 Version 12.3.6 in a 2023 Model 3.
you're probably on an old update then, it's really good now
6:18 Cameras can achieve depth perception, that is how 3D cameras work and even human vision.
LOL. Look, I'm a big Tesla fan and love my AP1 Model S, but if you have any real expectations of something substantial then you are a fool. If its a vehicle of some sort, it'll be little more than a prototype just like every other vehicle "release" has been. If it's a software update, then what exactly do you think has changed that hasn't been released in Beta form that will be giant leap?
Musk gets hated on, but lets face it - ALL car manufacturers claim what they want with no repercussions. Every single manufacturer claimed that they'd have Level 5 by 2020 and here we are in late 2024 and nowhere close. They claim that all of their new vehicles would be electric and no more ICE but there's not a single manufacturer that is even close to that. You claim what you want and make excuses when it doesn't even close to happening.
For me with a 2 month old Model Y, the latest hardware, the over air update which included summon is incredible. Tried it the other day at the movie theater. Stood outside and watched the car pull out of my parking space and drive to me. That’s got to be part of this new era. Tesla can “summon” your car if one has signed up their car to participate.
@@Skyking6976 Neat! I'd personally be so scared to do it, but I'm sure it works well.
@@kirkhilles523 Wife is the same but she is type AAA personality. I was a test pilot for years so I’m use to that sort of thing and still have really fast reflexes.
We'll put all this stuff right up there with the flying cars that I read about in Popular Science back in the 1970's.
I'm skeptical of Elon musk. He over promises and under delivers.
Skeptical? Are you stoned?
He promesis and delays over delivery Again and Again they are late to the party but they always executes unlike all their competition.
@@isaksidenius7059 Hyperloop, their rockets instead of a plane, he also promised electric VTOL supersonic jet...
@supee7541 you don't know what your talking about
FSD by 2016, 2nd mission to Mars by 2024, roadster 2 by 2020, all by a guy closer to the tech than anyone. He's terrible at predicting things.
All these self driving companies are chasing a wild goose. Cities need to be redesigned from the ground up, literally. There are too many incompatible systems conflicting with each other and they need to be kept separate unless you like to live in a chaotic environment, which is what we have today…
FSD is a good companion but not close to Robotaxi by any stretch. Robotaxi is a hype to prop up the stock valuation in the midst of slowing sales
From what I am seeing, Tesla are a lot closer to solving self driving than people think.
While vision alone may not be as good right now, if they ever do figure it out, it will basically put their competition out of business because they will be able to do the same thing with a much cheaper product to produce. People can do it with vision alone, so i do think it's possible.
Self-driving is raw as hell every job that humans do other than making robots is done by the end of this decade
Dan Ives is in for the surprise of his life... In 2025 Tesla won't get approval for FSD (still decades away), won't have robotaxis, won't have a Model 2, won't have a $61,000 Cybertruck and won't have growth. During the recession, we're expecting the stock to drop under $100.
Ives will be saying in 2025 "FSD will be ready in 2026, for sure! He's deep in the Elon tribe, with no clue about the real world of autonomous vehicles.
Even if everything goes right, is the market for robotaxi, FSD, AI, etc, really big enough to justify the enormous investment so far? I’m beginning to think that they’ve overspent this long ago. The returns will come but not as high as they hope.
FSD goes way beyond Robotaxi. Building an ie capable of understanding the environment it is operating in, and making decisions about the options available, will have huge ramifications for their Tesla bot, and offer a huge leap forward in capability. Solving general ai will be a massive step.forward.
If they bring seamless level 5 autonomy on 10/10, then yes, it will be ready
If AI becomes that good it'll be way more than just FSD. Think cure for cancer, nuclear fusion. Everything humans aren't smart enough to solve the AI will
You honestly believe they have level 5 now? 😂
If Musk pulls this off Tesla stock will 10x. He will pull it off it’s just a matter of time. The big advantage Tesla has over their FSD competitors is instant scale and the ability to pivot this technology to their Optimus robots. It’s a multi trillion company in the making.
I'm surprised that people still take Musk at his word for any of his predictions. What makes you certain that he'll pull it off?
@@TheSpartan3669 are you for real ? Look what he has done to the space industry. Look what he has done to the car industry.
FSD is just time - something that he is prone to being a bit too optimistic about.
I have FSD. It drives better than I do. Amazing. Long time coming, but it's here now. It really is. I recommend giving it a try (the new version, not what existed 6 years ago).
We have two Teslas both with FSD "Supervised" both cars have it turned off because we don't use it. You can save it with all of the try the new version nonsense. We purchased FSD in 2019 with the expectation that the car would be able to go out and make money for us all day rather than sit in the garage all day. Until it can do that, it is not what I paid for. It also seems like the hardware in my car wont support the final version if there is one.
So many pro Tesla FSD comments are suspiciously similar.
Or perhaps people that use it actually realize how good it is. Its the nay sayers that get me. It takes zero effort to be negative.
I tried it on my MYP 24 with the free trial. It bugged out on a long curved turn at night on a street and I had to take over. On the highway it was too slow for my taste. Not worth the subscription at least for me yet. Autopilot is good enough tho, drove from LA to Oxnard and it saved me the hassle of stop and go traffic when we caught an accident on the 101. For $100 a month it might be worth it for a long road trip and just cancel afterwards. Might consider it if I do Coachella again next year to assist on the tiring drive back home.
@@AlexJee7601 The system is not very good mathematically. A human goes 100,000 miles between accidents if you took a 1,000 miles of FSD driving how many interventions you think you would have?
I don't remember which self driven cabs, but the amount of people they had in their hq would be enough to just put them in the cars and drive them themselves, and it would be a cheaper and better service. People want self driven cars, but if you don't want to drive we already have subways, taxis, buses, trains..
6:15 Cameras do have depth sensing, just like humans. If you put 2 cameras together you make a stereo camera, that is able to see depth, just like Tesla does and how humans see depth. Even if you have a single camera, there are algorithms, that render depth, based on the movements of the camera. I have a lot of experience with SLAM and I even met the Estonian president for my achievements.
The only way to do FSD is if all cars r connected to each other and the city has cameras everywhere. Everything has to connect to each other so the cars can see everything
FSD can already see more than a human.
He’s just trying to pump the stock…again.
I'm surprised so many people haven't caught on to his scams. Tesla Q1 2024 earnings tanked and he immediately made the robotaxi announcement. I'm convinced the cybertruck is the last vehicle that will be in production since they are betting big on vapourware products like robotaxis, robots and fsd.
No FSD is getting very close to being ready. It is as simple as that.
Like a pump and dump situation?
Human driver can hear sounds, can use hands to communicate with other road users, how about robots driver?
I don't think people understand just how powerful ai is. There is literally no reason an ai cannot be trained to understand hand gestures.
Dan Ives is stupidly wrong when he says Tesla will be giving robotaxi rides in 2025. Dan doesn't understand the challenge of AVs, where 99% of the drivng is easy, and the last 1% is 99% of the challenge to exceed human levels of driving. FSD is now at 99%, with a nice demo, and 99% of the hard challenges yet unsolved. There won't be a Tesla public robotaxi service in a major city by 2030.
Then why are they building those massive AI training centers? If it's going to take that long, Tesla is doomed because they will literally run out of money. Your "There won't be a Tesla public robotaxi service in a major city by 2030.." comment is declaring Tesla as easy picking for vulture like hedge funding establishments.
I have been following self driving for years (not just Tesla) And from what I am seeing, it would not surprise me to see Tesla operating autonomously somewhere in America next year. The rate of Progress since switching to neural net has been rapid, every update adding more and more capabilities.
It might not happen next year, but I would not be surprised if it does.
@@daydreamer8373 FSD won't be driverless next year, unless they do it in an empty parking lot. It can't last one day driverless in a real city.
@@charliedoyle7824 Time will Tell. FSD is about to surprise a lot of people. If not next year, very soon.
@@charliedoyle7824 Time will tell.
Yes one more decade and we’ll be closer to level 3
I've said for years now that using only cameras is just dumb. I understand the motivation, it's cheaper, but cameras are still nowhere near the efficiency of human eyes, especially in low light conditiions. Even if you can ultimately do it, radar/lidar still provides more safety. Just because *humans* can only see with their eyes doesn't mean we have to limit self-driving vehicles in the same way.
somehow I dont bump into things without radars, weird
@@haroldroa1806 Because your eyes are far superior to cameras like I said, especially in low light conditions.
The size of the data set available for algorithm training is orders of magnitude larger if they only use cameras.
@@dijikstra8 Companies are in the process of dropping Lidar due to an AI explosion. The cost of these Lidar cars is up there with Lamborghinis and Ferraris. 10 times more expensive than a potential robotaxi.
Visual only is far superior for so many reasons and it's not even a conversation any more. It was a good argument before the chatgpt moment and now it's just obsolete. The quality of cameras, AI chips and available data is getting better every 6 months.
@@dijikstra8 I have no idea why you think the human eye is superior to cameras? The camera can not only see all around the car, but contrary to your beliefs can see far better than humans in the dark. Going camera only was not about saving money, it just made the most sense. Extra sensors does not make a safer car, it just adds cost and complexity.