Robotaxi / Cybercab Launch Confirmed for June

A cybertruck's self driving feature literally drowned it's driver while the fire department could do nothing but watch as they suffocated in real time.

if you mean the lady like 8 years ago yeah. But I've also seen a recent video of some chick tripping on the street and the car immediately avoided running over her head. The future will be interesting. The current cars in use now will literally stop on a dime if you even shift put your foot in the street when it's coming

can't find that online do you mean the lady who backed up into water

The problem with self driving is the last 1% of reliability is like 100x the difficulty of the first 99%. And this just doesn't scale with Musk's style of min maxing low level tech for linear gains. You are at like 1% of the way there to get a truly "fully" autonomous car. Its not like taking shitty electric cars in the 2000s and min maxing the fuck out of the design to just barely bump a few luxury models over the line of viability with gov subsidies. Self driving needs a new paradigm. Its not something Tesla is going to figure out. Similar to how progress on space travel stalled, you can make the best, most precisely controlled rocket, but its still a stick filled with explosives. Tesla self driving is still a rats nets of a bunch of low intelligence hysterics just stacked on top of each other. Needs some sort of undiscovered AI related innovation that is more likely to happen in academia.

Wait until tesla figures out cities are full of liberals that hate them

Heck off hebrew

Wait until liberals find out they actually can't just get away with destroying property and killing people

Kek Musk is a chud. His products now officially give the ick. Good luck selling this weirdo stuff ever again.

Austin has been completely ruined by these retards. I want to go back.

t. From austin, dad was from austin, grandfather was from austin, great grandfather was from austin, etc.

Why is this a two seater? Why the impractical show off doors?

2 more weeks until fully self driving cars, for real this time

It was 2 years ago but I guess 2 years is a long time. Seems like just yesterday to me.

How did you solve the ground filtering thing? Because as far as I am aware anything 12 inches tall or below is tire food.

And we had such a nice pump going on by Tesla drowning in bad news, and here comes some bright points to make it plummet. god i hate this clown market.

This, the innovation to make self driving a reality isn't technical either. It's legal.

You want a self driving truck? Prepare to eat repair costs, how does the truck register it's goods? How does the truck read defaced bum fuck nowheresville signs? How does the truck give permissions to the mechanic who is going to replace it's brakes. Who tf pays for the autonomous truck deciding it wants to play chicken with a train. Or the joy riding motorcyclist who was sitting in the blind spots. You guys invested in a smart trailer right? You guys put BillyBob's Custom trailer out of business right?

You can tell how brutal this market is by how narrow the scope of work has become.

Autonomous cars were shilled heavily half a decade ago, then it became "robotaxies" then... well, they sort of disappeared from the media after a few oopsies summer of 2023. And even then an autonomous vehicle was a meme because each required 1-5 technical staff on call to troubleshoot it all the fucking time.

I fully believe they are still cruising, but the paradigm shift is not technical stuff. It's legal, and failing legal, social. Maybe it can work in china where everyone is a bug person. But in the first world, where liability exists?

Investors look at self driving, and just laugh.

who are these sweet summer children?

They ask.

why don't they just hire a jeet for 5 bucks an hour? Instant self driving. Millions in liability offloaded.

when the marketing team gets priority, there is nothing to buy.

How's Boom doing these days?
boomsupersonic.com/

Also that "stop on a dime thing"

Doesn't that piss off the actual customer? A computer stopping on a dime doesn't feel like a human stopping on a dime. You know there are videos out there of people getting their skulls cracked open because they were standing up in ye old autonomous tram? It stopped on a dime and sent one of the dumb ass "engineers" head's into it's fire extinguisher.

Have you nigs figured out precise torque control yet? Last autonomous vehicle I was in had some undergraduate bang bang speed control clown getup. I could feel it, the company reps assured me that wasn't the case. But if they are not lying that simply makes it worse. My conclusion was that their low speed control engineers have cluelessly stacked fancy forms of control over each other to result in a control system that may as well have been written by a 20 year old intern.

And yes, I know it will stop on a dime if it sees something. Every civilian autonomous vehicle I have had the misfortune of observing behaves as a timid creature of prey. They all are, big fat wastes of money that will be slaughtered by the first lawyer to touch them. Potentially not even for anything they physically do.

Fix your perception "stack" and don't play like the car "just did that" you told it to and you forgot like the absent minded egghead you are.

And they will win the lawsuit because they bought the courts and government. Numbnuts.

autonomous driving is constrained by edge cases (the 0,0000000001% like a toddler dressed as a seagull crossing the road)
ML relies on statistical prediction
it has no answer for edge cases since it never learnt to deal with them and it cannot reason
but it MUST do something because it's a moving car
when Chatgpt fucks up it fucks up your excel file, when the robotaxi fucks up it kills your grandma
you won't see self driving in your lifetime because there is not even the start of an approach to solve this problem yet, very much like "hallucinations" in LLMs
Elon is not stupid and he knows this. The taxi thing will just be a clown show for investors on sunny days like what Google does with waymo
the tech has not advanced in half a decade and compute or even data are not the problem
the problem is that there is always on edge case more to solve and the stakes are too high to accept anything but perfect reliability

TSLA FSD is driving around on its own in Chinese roads using its neural network alone and you niggers are shorting TSLA. If it can drive around China without any issues than North America driving will be a doddle in comparison.

x.com/joehansenxx/status/1904801598477664489?s=46

When did right wingers become fans of EVs, and Elon Musk lmao
It's always been a leftie leaning le science type thing

looks like a garbage death trap for retards.

I don't think Austin could go back to what it was even if every tech bro left. They literally and metaphorically bulldozed so much of what made it weird and interesting. I think in some regards it's much nicer than like 2010-2016 because downtown and 6th is so built up now far extending into what was once the hood. But the weird people got priced out and the few remaining work multiple jobs to supplement their DJ job and lifestyle coach bullshit. Maybe the city will actually build real first world convinces like an actual metro and bike lanes but instead we get retarded robot cars for now lol

So does BYD at half the price. What this does not mean is autonomous taxis, and that is the one thing that could save Tesla.

First two google results for "China Tesla FSD":

Tesla FSD run two red light in 30secs in China

Tesla halts driving-assistance software trial in China, pending approval

autonomous driving is constrained by edge cases (the 0,0000000001% like a toddler dressed as a seagull crossing the road)

ML relies on statistical prediction
Correct, accidents will still occur. I'd bet less frequently than accidents caused by human drivers however, but it's more of a spectacle so will people tolerate that? Probably not, I agree with the guy who was saying that legal and social issues are the primary roadblocks.

the tech has not advanced in half a decade

This is just flat out false. As someone with a Tesla the current FSD is so much better than it was a few years ago that there's no comparison. I could not tolerate using it even three years ago, now I turn it on and have only taken over control from it a couple times in the past four months, and those weren't necessary, it was just being a bit annoying with how it was parking.
I've always been a nerd into scifi and pretty ambivalent between the two parties because I despised both and basically only voted at the local level. In my opinion the dems have raced to the left to the point that I for the first time in my life voted for a republican for president in 2024. I'm probably not the only person of this type, further I know a lot of former dems have switched. I'd probably still be considered a lefty by most people here, the democratic party has just gone kind of crazy as far as I'm concerned, some of which I may have overlooked as fringe issues without a new perspective from being a father.

I think you're underestimating how many technophiles are now voting for the "right-wing" party, from a combination of being pushed out, seeing the republican party shift, and simply aging and having an older perspective. It's not the majority, but it's a growing minority. Another area showing this that's attractive to people like me is how republicans are the main party pushing for nuclear energy production.

The dems are more centrist than ever. What the fuck are your smoking?

Let me guess, you don't have any kids.

let me guess, your kids are autistic

I bet you couldn't cite a single far left policy on Harris's platform.
Go ahead, take a look.

change the world just like Google Fiber!!!

I mean, it kind of did. %-/

Official platforms are always curated to seem completely non-objectionable. I can easily point to her stating support for social media censorship and the broader democrat supported policy of teaching LGBT material to elementary school children without informing parents.

You mean the drunk Chinese richcunt who backed her car (not CyberTruck) into a pond, then was too drunk-and-stupid to figure out how to open the door or windows to escape, while she called random people from underwater and chatted for eight minutes and nobody bothered to call 911 to rescue her until she was glub-glub-glub drowned?

bbc.com/news/world-us-canada-68622898

So what you're telling me is that you can't find a single policy. It's a 92 page document.

Tesla is the undisputed world leader in self-driving. They are the only company to use a full end to end AI model to solve the self driving problem.

There are people that use FSD every day now because they simply trust it by now, and it's a major convenience. It often drives more comfortably than a real driver, and I believe it has also passed the point where it is safer than a real driver.

FSD performs way better than the leading Chinese self driving software even though it hasn't been trained on Chinese road regulations yet. Take a look at any video on this channel:

youtube.com/@TeslaFSDinChina

What I'm telling you is that the official stated platform really isn't important. It's like pointing to a press release from a corporation dealing with a catastrophe gone over by lawyers several times and saying there's nothing objectionable in it.

Waymo is the leader, at least in the US anyways. They've had self driving taxis on the streets for ages now.

Harris is about as far left as Romney.

Who the fuck even mentioned Kamala Harris fucking retard

1. Waymo cars cost $100,000 each with all the hardware and retrofitting. This is not scalable.
2. Waymo must 3d map any location and only operate within a geo-fenced box. After offering rides for 4 years, they still have less than 1000 vehicles and only operate in one city.
3. Tesla does not need to map. For example FSD performs very well on unmarked dirt roads.
4. Tesla is built to scale: they can manufacture millions of robotaxis themselves, and FSD can be launched in any country without having to 3D map every single street and alleyway.
5. Waymo uses static algorithms to navigate, which is fundamentally not the solution to a generalized self driving solution, which is required to effectively scale.

That said, Tesla FSD has the ability to change the world, while Waymo is a boutique experiment that cannot realize the full benefits of its dream.

Weird, did today's directive say to talk about Kamala Harris or something?

this is what elon does
very small scale demonstration in 1 teeny tiny area
investors flood in
he dumps
the project stalls and eventually stops
repeat

1. Waymo cars cost $100,000 each with all the hardware and retrofitting. This is not scalable.

The lidar component is already down to $1k last I checked, probably even lower now

2. Waymo must 3d map any location and only operate within a geo-fenced box. After offering rides for 4 years, they still have less than 1000 vehicles and only operate in one city.

This is because they are using a cautious, gradual approach, not because it is technically limited.

3. Tesla does not need to map. For example FSD performs very well on unmarked dirt roads.

The mapping is to add additional data, not because it is necessary for self driving. Additional info is always useful, for example you can have laws in certain areas that you need to follow. It's not all just about vision.

4. Tesla is built to scale: they can manufacture millions of robotaxis themselves, and FSD can be launched in any country without having to 3D map every single street and alleyway.

The Waymo tech can be applied to any existing automobiles. They'll probably license them to existing car companies that already produce far more cars than Tesla.

5. Waymo uses static algorithms to navigate, which is fundamentally not the solution to a generalized self driving solution, which is required to effectively scale.

That's just plain wrong, Google is a leader in AI too. Even their robotics is ahead of Tesla.

This is because they are using a cautious, gradual approach, not because it is technically limited.

yeah I was gonna say, it sounds like waymo is doing it the correct way, whereas TSLA just wants to rush through everything loool

Show me one place in Nike's mission statement where it says they support the use of third world slave labor?

[This post was sponsored by Nike™]

The mapping is to add additional data, not because it is necessary for self driving. Additional info is always useful, for example you can have laws in certain areas that you need to follow. It's not all just about vision.

Ok, I gotcha. So lets say you dropped a Waymo car and a Tesla with the latest FSD in China. Which one do you think will perform better?

Waymo because they have more experience and lidar.
Tesla has already been forced to suspend their trial and to remove "FSD" from the name in China.

Tesla owners are now driving an average of 14.7 million miles on FSD per day

Waymo's autonomous vehicles drive an estimated 25,000 miles per day on public roads

Actually it seems like FSD is used 588 times more daily than Waymo. So I would easily give Tesla the most experience.

Tesla has already been forced to suspend their trial and to remove "FSD" from the name in China.

So, recently Tesla started humiliating the Chinese domestic self driving being developed. The corrupt government created some regulatory hurdles with the help of Tesla's competitors that will just delay the China rollout a bit. Honestly that's a very positive sign that the Chinese are afraid of Tesla taking over their market. It really speaks to how good Tesla FSD is.

They have more data, not experience. It's also lower quality data because it has fewer inputs. It's not even controversial to anyone working in the space, Waymo is way ahead of Tesla. Even if they were on par, Tesla's valuation doesn't justify it's price when there are so many competitors.