Tesla FSD literally makes turns (including u-turns), merges lanes, works on virtually all regular roads and highways, etc. It still makes mistakes sometimes, but nothing safety critical in my experience since it received a major update earlier this year.

And I say this all as someone who thought it was gimmicky and sucked a year ago. For $100/month it's a no brainer and reason enough to buy one. Shocked this tech is not being licensed by other companies yet.

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48095758)

MFCR, it's extremely impressive and it's now MUCH better than it was before. It's also only going to continue to improve from here.

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48097494)

CR who doesn't love their daily commute?! It's one of the true glories of each workday! lol

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48097503)

Another fatal and embarrassing mistake Elon has made: solving a "prole problem" that only the bottom ~90% of the US population actually has. What a fool.

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48097521)

"REAL MEN drive themselves to work, they don't let faggot 'software packages' do it for them!"

Wow you just blew my mind

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48097533)

Sorry that literally everything I ever predicted about TSLA / Tesla is being proven to be correct right before your eyes. Oh well, sucks to suck I guess. I tried to tell you haters what's up, and you guys just called me a dumb faggot and insisted that Elon is nothing but a scam artist charlatan. Somebody had to be wrong!

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48097546)

https://xoxohth.com/thread.php?thread_id=5590346&mc=7&forum_id=2#48060198

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48097556)

that's probably not a good thing, but it is a thing.

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48102526)

It's still "supervised" fsd for regulatory reasons, but it's still incredibly relaxing to lounge back and watch the road instead of actively driving.

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48095793)

It's incredible that these anti-Elon kikes have now been forced to move the goalposts all the way from "LOL ELON IS A SCAM ARTIST HE'LL NEVER GIVE US REAL FSD LIKE HE PROMISES" to "Ackkkkkkkkkshually, I *love* my morning commute... IMO people LIKE driving and don't WANT to be driven around!" in the span of just a few short years.

Oh man.... I guess Tesla's improved FSD package won't end up being wildly successful and insanely profitable for Tesla within the next few years after all, since we've established that nobody actually wants this service.

lol

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48097511)

HOWEVER, cars with HW4 chips (manufactured since basically late 2022/early 2023) get access to the latest versions and it's unclear if HW3 cars will be able to keep up in the future. Probably best to get cars with HW4 chips to be safe than to hope for some sort of retrofit being available.

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48095860)

they’ll have all the bugs worked out in 15 years but until then ljl

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48095992)

"If it's not perfect, I'd absolutely never trust my life with this shit!" - Guy who, statistically speaking, almost certainly drives less safely than Tesla's FSD will be operating >99.9% of the time within a year.

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48096011)

Which part of what I said was substantively incorrect, Ari? Please feel free to be specific tyia.

Sorry that the proliferation of Teslas are OBJECTIVELY making America's roads statistically safer for you and your family to drive on. Fuckin' Elon, man!

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48096039)

Quotemo?

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48097490)

Yet it is insanely impressive and also sort of scary for the future. The whole "we were promised self-driving cars and its never happened lol" line is gone now.

We are 99% of the way there, and if we were ambitious, we could probably have accident, traffic free highways in 5 years. It'll probably end up being more like 15-20 years, but we have the tools now.

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48096958)

It's almost like all of the libs who stupidly bet against Elon and then doubled and tripled down while calling be a "dickrider" and insisting that he's ackkkkkshually just a midwit scam artist have now had every single one of their holds thoroughly blown out.

B-b-but that can't be right!

lol

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48097493)

Sounds more stressful than just fully driving the car myself.

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48099201)

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48105652)

PS: Tesla's profit margin on software packages like this (which, once they are developed, can instantly be ported into any given car with the click of a button at no marginal cost to Tesla) is comically absurd. And of course, Tesla can not only sell them on an ongoing subscription model, but in the future they can also offer all sorts of upgrades for additional upcharges which a lot of people will happily buy for another $20 or $50 or $100 month or whatever (multiplied by millions of Tesla owners across the world over the lifetime that they own their cars...).

It's almost like TBF actually had a great point when he kept insisting that TSLA still remains a great buy at current prices because at its core, Tesla is actually a *software* company and it's only just getting started when it comes to leveraging that aspect of its business model.

B-b-but that can't be right... TBF is always wrong about everything!

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48097527)

Jewish shitlibs love to make threads calling me out for being a fucking dumb faggot who is always wrong about everything...

...almost as much as they love to tell me to shut the fuck up and stop being such an annoying self-obsessed queer ("nobody cares retard") when I pop into threads to point out that I was actually 100% correct about something.

It's almost like a "heads TBF is a dumb retard who loses, tails TBF is a dumb retard who loses" situation with (((these people))). But that can't be right!

lol

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48097707)

BANGANRAG!

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48099545)

1. Spend billions just on compute

2. Develop all the training infrastructure

3. Outfit millions of cars with the $1.5K hardware suites (cameras and chips) for data collection and testing beta software FSD.

Safe to say it would take years to replicate Teslas strategy and require feeding many different mouths, while Tesla manufactures their own cars, codes their own software, designs their own chips, develop and manage their own data centers, etc.. And even when they do catch up to where Tesla is now, Tesla will probably be deploying robocabs at scale for $1/mile (Uber’s are $2-$3/mile largely due to driver-associated costs). At that price point, many people probably won’t even bother owning cars.

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48097601)

Again, data is not the wall people think it is. It’s cheaper and more effective to just train on car data, but you can pre train on other video data to give the models substantial understanding of the world and then fine tune on car data. Companies like Alphabet are likely to be training very large multimodal models on games, text and video data in the near future because they want AGI. You don’t need massive amounts of car driving experience with a model that basically understands how the world works. There’s a reason why Google can give Gemini a book of a new language not seen in training and have it learn with human level sample efficiency. If you already have a conceptual framework, you can repurpose learned representations for a novel environment without starting from scratch.

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48097627)

CR, anyone could do this IMO: https://techcrunch.com/2024/08/03/tesla-dojo-elon-musks-big-plan-to-build-an-ai-supercomputer-explained/

Lol it's always a fucking (((Pumo))).

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48097712)

Tesla is spending $10B in 2024 on AI (mostly related to perfecting FSD) and for all we know they may need to continue spending this amount for years until FSD is robotaxi ready. We don't even know yet if their autonomy strategy (vision only, no lidar/radar or other sensors) is "correct" yet, so there is still no rush by any other company to aggressively even try to replicate it.

I think you are assuming because LLMs have been easy to commoditize, self-driving will be similar. I think the problem is much harder than you give it credit for. You need many billions of real world miles, develop and onboard an affordable hardware suite on millions of vehicles to collect all those miles, hire a capable team, know exactly how to optimally train the models (there is no off the shelf solution for the self-driving AI problem because it has never been solved yet), and then deploy them and hope very, very few people get hurt/killed before your science project is yanked off the market.

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48098578)

Driving is full of unpredictable edge cases that cannot be reliably handled without billions of miles of diverse driving data so even the rarest of scenarios can be handled appropriately. Human drivers may learn to drive in a few hours, but their experience comes with years of real-world interaction and constant feedback from a wide range of situations.

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48103394)

I think Tesla's competitors in the autonomous driving space have realized that you will never get to true autonomy unless you pick a couple cities with good weather year round and train your model to death on those cities. That's the approach that Cruise and Waymo have taken. They actually have functioning robotaxis in a couple places, and they are slowly expanding their service areas.

Tesla's FSD isn't anywhere close to being able to do this. And I'm not sure that FSD will ever be able to be more than a fancy driving assistance feature. Being able to drive safely 99.9% of the time in nearly all locations is impressive, but I'm not sure how much it helps you solve the problem of driving safely 99.9999999% of the time (which is what a truly autonomous vehicle needs to be able to do). It will be interesting to see what happens.

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48097700)

GIANT TWIST THAT NOBODY COULD HAVE EVER SEEN COMING: Putting aside the robotaxi prospect (which may or may not end up being solved by the AI-assisted supercomputing that Elon is currently throwing at the problem), consumer-oriented personal FSD systems don't have to be perfect in order to become the standard. They only have to be significantly better *on average* than the *average* human driver, and then we'll reach an inflection point where "pressure" (much of it artificially imposed in a top-down fashion by insurance companies who stand to save a ton of money) will begin mounting to fully normalize FSD even if it's only 99.99% instead of 99.9999% or whatever. Expect to see a massive (((media campaign))) seeking to accomplish precisely this task within the next five years, with legislators being brought to heel as well. At that point, Musk critics will be betting against Elon *and* (((The Auto Insurance Industry)))... I wish them the best of luck.

ENORMOUS OCEAN-SIZED DOUBLE TWIST (SIT DOWN FOR THIS IT WILL SHOCK YOU TO YOUR VERY CORE): Despite the fact that brilliant Jewish shitlib Pumos have been smugly assuring us for years now that Elon Musk is a midwit charlatan who will never deliver FSD on that level, Tesla is already almost there right now in 2024 (admittedly in no small part because a ton of drivers in this country are literally brown and drunk and retarded so the "average human driver" bar is actually pretty low at this point).

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48097718)

The issue is that I feel like the upside for this fancy driving assistance technology is much lower. The first company that produces a truly autonomous vehicle that works in all locations and conditions will be a multi-trillion dollar company that completely changes human civilization. A fancy driving assistance feature, on the other hand, can probably fill a profitable niche. But are the potential profits enough to justify Tesla's already sky high valuation? I'm not sure.

That is an interesting point, though, that if even assisted autonomous driving becomes significantly safer than human drivers that you might start to see pressure from insurance companies (and legislators) to make it a standard feature. It's going to be fascinating to see what happens.

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48102213)

Good point, Elon is probably fukt and just doesn't know it yet. What a dumb midwit retard!

*moves the goalposts another 10 yards*

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48097710)

Yep any day now, "Waymo" is going to surpass Tesla. Just wait -- everyone who REALLY knows FSD tech knows it's a MUCH better model, trust me I know these things!

*farts cum, adjusts Warby Parker glasses, straightens yarmulke*

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48097713)

CR, Betamax is OBJECTIVELY the SUPERIOR home video technology, all of the REAL experts agree. And FYI that's why Betamax will ultimately triumph in the marketplace -- it's simply BETTER, period. Trust me, I KNOW these things and THAT'S how the REAL WORLD works!

*adjusts buttplug*

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48102013)

It is, however, an amazing ADAS system. I honestly find TBF's bull case for Tesla to be more compelling. Things like seat belts, airbags, and backup cameras used to expensive add-ons for luxury cars. Now it's not even legal to sell cars that don't have these things. Once enough data comes out that FSD is safer than human drivers, I think you'll start to see pressure from insurance companies and regulators to mandate its use. And I can easily envision a scenario where Tesla starts licensing its FSD to other automakers because they don't want to build their own system from scratch when Tesla is years ahead of them.

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48104601)

Waymo is too expensive--more expensive than Uber--so they are not a real threat to Uber. Costs $200K to outfit their cars with all the sensors and dildos. Yes, those will get cheaper over time, but their self-driving solution is not very scalable. They need to pre-map every city, which is difficult because 10% of roads change every year, there's construction projects obscuring roads, etc.

The company that can build a car for $20K and outfit it with only $1K-$2K worth of self-driving cameras and chips will win the robocab market because they will be able to offer rides for $1/mile (cost of an Uber is $2-$3/mile mainly due to driver-related costs).

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48098633)

So weird how liberal autistic Jews continue to stamp their feet and FURIOUSLY INSIST that "Waymo" is waaaaaaaaaaaaaaay more credited than Tesla in 2024, etc.

BTW this is what a "Waymo" looks like for anyone who has never heard of this obscure little company that will never become a household name: https://www.npr.org/2024/03/14/1238489046/waymo-robotaxi-los-angeles

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48099553)

Your concerns about Waymo's business model are valid. It has a long way to go before it becomes profitable, and it is unclear if it is scalable. My response would be that even if it only operates in major cities initially, that would still be a game changer. Tesla has a very long way to go before they will even be close to being truly autonomous. (The issue about roads changing and construction problems is an even bigger problem for Tesla given that they are trying to build self-driving cars that will work anywhere.) Maybe they can just fine tune their FSD algorithm to get true autonomy in certain cities/regions, but Cruise and Waymo had the equivalents of Tesla's FSD on the road years ago. Getting from 99.9% reliability to 99.99999% reliability is an extremely difficult problem.

So it will be interesting to see what happens. My point is that Tesla has some competition in this arena, and it's far from clear that Tesla is going to win.

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48100563)

It doesn't matter if Waymo operates in major cities or not - the bottom line is their bloated cost structure forces them to be more expensive than Ubers, which means they will struggle for market share. The one advantage they have over Ubers is some people are willing to pay a premium and wait longer (Waymos fleet is less than 1K, Uber has 1.5 million drivers in the U.S.) for a ride without a human. That's a very small market.

The point of Tesla's vision-only approach re road changes and construction is that because their AI models are trained on enough video data of what to do around construction areas they won't need to be specifically programmed for how to act around those types of areas. They will intuitively behave as humans would - follow detour signs, respond to hand motions from construction workers, etc. Meanwhile, a Waymo will stop dead in its tracks like a train following a route would.

Tesla is by no means guaranteed to win, but my point is IF their approach to self-driving works, their superior cost structure to competitors will flip the entire rideshare/car ownership industry on its head.

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48101037)

And I'm not optimistic that Tesla will ever get to the point where they will be able to follow detour signs or hand signals from construction workers. Some of my biggest issues with current FSD is that it doesn't seem to know how to handle construction zones at all other than to avoid orange cones. There is a street by my house that was temporarily converted to one way almost two years ago while they build a new building. My FSD still tries to drive the wrong way down that street every single time (despite the huge "Wrong Way" sign). I think the only way that you are going to get true autonomy in the foreseeable future is to follow the Waymo/Cruise approach and get really familiar with one city at a time. Current machine learning technology simply can't handle all the weird edge cases that can happen when driving.

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48102268)

"I'm pretty bearish on Tesla FSD, but I'm actually quite bullish on Waymo/Cruise!"

Yes, we know you're a Jewish retard. Odd case that you're also a Pumo.

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48102290)

If you were just like “I don’t think Tesla FSD is there yet” or “I doubt FSD will ever get to 99.99999% perfection,” I would be happy to respectfully agree to disagree. Those are reasonable positions.”

“Ackkkkshually, Waymo is the REAL market leader in FSD!” is full retard, and you get treated accordingly with no apologies.

Also LJL at thinking for even a second that I give (or should give) a single shit what my “credibility” is like with Pumos. Go fuck yourself and hop into an over yesterday, kike.

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48105877)

Tesla FSD already follows hand signals sometimes, but is still awkward around construction/emergency sites and isn't reading road signs yet. But 99% of the work is basically done.

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48103425)

And 99% of the work has been done for autonomous driving for at least 5-10 years now. Unfortunately the final 1% is thousands of times more difficult than the first 99%.

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48104651)

Elon literally has like the 5th most powerful computer on the planet cranking away at this very problem using the latest cutting-edge AI technology (armed with billions of hours of real world driving data with more being added at an ever-increasing rate every single day) to help iterate next-generation computer-written code aimed at getting this tech from 99.99% to 99.9999%. He also has no shortage of money or willpower to continue to throw at this problem, and he has every possible incentive in the world to continue to do so -- because whoever fully automates cars* (see below for *) will literally print endless money for years to come. And his approach is already paying massive dividends; it's not a coincidence that this latest software update is a quantum leap forward from where this tech was just a year ago.

What sort of comparative financial and computing resources does Waymo have to throw at this problem, I wonder?

*As discussed above ITT, the idea that FSD needs to be 99.9999999% accurate or whatever in order to become socially acceptable is simply bullshit. This tech doesn't exist in a vacuum, we don't live in some hypothetical world filled with perfect human drivers who never kill people or cause accidents in cars. FSD only needs to become demonstrably better than the *average* driver on the road before the entire (((car insurance industry))) will suddenly make a huge coordinated push to normalize it for purely financial reasons... and once we hit that tipping point, you can bet that (((the mainstream media))) will fall right in line. You really think (((they))) won't happily save money pushing this tech?

You're a fucking idiot if you can't see that turn of events coming down the pike within the next few years, and I say that as a MFE in this field. You can consider this to be a firm prediction from the same guy who advised XO to buy BTC in 2013 and ETH/TSLA in 2017. HTH.

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48102029)

And Waymo is an Alphabet company, so they don't exactly have a shortage of cash to burn on this technology. Cruise is backed by GM, and Amazon is trying to get into the autonomous driving game as well. Tesla may very well end up on top, but they have some tough competition, and I'm not sure that they are the horse I would bet on right now.

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48102287)

Waymo's annual R&D budget is a small fraction of Tesla's annual R&D budget, Waymo's computing power is a little fraction of a little fraction of Tesla's computer power, and the IRL data pool that Waymo has at its disposal is a tiny fraction of a tiny fraction of a tiny fraction of the data pool that Tesla gets to work with -- and the delta on all of those fronts is only going to increase exponentially from here on out. (And real talk: lots of people think that Teslas are cool, while Waymos are objectively gay as fuck and look like they have dildos sticking up out of their roofs.) This race is over on every level.

It's always amazing talking to people like you about this. "Look, I know what I'm talking about here... and frankly, I think Tesla might very well be the underdog in FSD when you REALLY stop and think about it!"

Haha yeah haha, sure thing bud have a good night.

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48102301)

Look, I own a Tesla, I use the FSD, and I think they both are amazing. If Tesla's stock price drops to a more reasonable level, I will be dumping my entire net worth into it. But you seem to have your mouth so firmly around Elon's cock that you're completely oblivious to Tesla's potential weaknesses and threats.

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48104672)

"Waymo is absolutely kicking Tesla's ass in the autonomous driving arena"

Oh wow well I mean if a Pumo on the internet authoritatively asserts that this is true!

[CITATION NEEDED]

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48105175)

Again, as a driver assistance feature, FSD is amazing and seems to be far ahead of any competitor that I'm familiar with. But for truly autonomous driving, right now Waymo and Cruise are years ahead, and if you don't see that, you're simply not being objective.

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48105750)

Wow this sounds very scientific, can’t argue with this dispassionate and well-informed statement of accurate facts!

https://www.usatoday.com/story/money/cars/2024/05/16/waymo-investigation-crashes-violations/73712498007/

https://www.theverge.com/2024/5/14/24156238/waymo-nhtsa-investigation-crash-wrong-side-road

https://mashable.com/article/waymo-second-recall-2024

https://thefrisc.com/cruise-robotaxis-were-banned-from-sf-streets-now-self-driving-cars-have-miles-to-go-before-winning-public-trust/

Why do (((you people))) just shamelessly lie like this? You do realize that it doesn’t work very well now that goys have figured out how to use the Internet and it’s becoming increasingly impossible for (((you))) to control all information sources and shape all narratives, right?

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48105902)

IDK I heard a bunch of Pumos talking about how "Waymo" is pretty awesome (lmao)

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48097717)

Also, I rather enjoy driving on road trips, especially when the route is scenic or novel for me. I don't think FSD is for me and if I can avoid it I will.

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48097728)

Bro - advanced driver assist technology exists to make commutes, traffic, and long drives way more chill that people 20 years ago would have paid tens of thousands for and we get to have it for just $100/month.

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48099255)

They can raise subscription price whenever they want as well, but there is no reason to since higher use rates for now will help them gather more data from the fleet to keep improving the tech. Once it gets good enough to be unsupervised, they can easily raise the subscription price significantly.

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48099859)

Whoa, sounds like Tesla is acting like it's actually a software company or something!

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48099969)

It's almost like even here on XO (which is admittedly a self-selected group of relatively smart and educated people), a huge % of our population are fucking retards who have very little actual foresight.

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48102075)

I wouldn't want to be in a FSD vehicle unless I knew it (whoever "it" is) was fully liable in all cases at all times... in which case, why would I even need insurance?

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48100692)

CR.

TWIST: Tesla insurance is offered at a very attractive rate, but it requires drivers to agree to allow Tesla to use its cameras to spy on them and confirm that they are operating the vehicle in accordance with Tesla's dictates (which gives Tesla the ability to parse out cases where drivers fucked up and refuse to accept liability in those instances). It also gives Tesla even more visibility into cases where FSD actually does fuck up, which is very valuable information that Tesla uses to further refine the tech.

It's almost like Elon has carefully created a giant feedback loop that gives him ever-more control over the product and simultaneously also helps him to gain an ever-greater data-mining advantage over all of his competitors.

But that sounds pretty smart, and Jewish Pumos on XO have assured me that Elon is actually just a lucky midwit charlatan. This can't be right!!!

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48102091)

He can't keep getting away with it!!!

(Twist: He can, and he will.)

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48102344)

I'm sure there is some jiggering with the statistics there, but let's assume its at least even now - and improving rapidly each year. And it's going to improve incredibly quickly once there are no humans on the road and all the robots can talk to each other in real time.

I'm sure there'll still be accidents, but it could easily be 1/1000th the number of accidents we have now.

CSLG's "doctors" will finally be out of business.

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48101486)

Let's assume your FSD car is less likely to get into an accident than you are. Let's also assume that the likeliness of an FSD car getting into an accident will never be zero. In the (unlikely but not impossible) event your FSD car kills someone, are you responsible? At what point would you be willing to say "yes?" And if your answer is "no," then who is responsible? Is that party going to be willing and able to take on such responsibility en masse?

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48101766)

If you're asking if Tesla FSD tech can brake / turn a car faster than a human can, the answer is "yes of course it can." Will that necessarily save you from a drunk spic flying through a red light? No, but then again, you also aren't going to be able to save yourself from that situation in every case either. Contrary to what Tesla Haters like to pretend, really bad shit didn't just suddenly start happening on our roads the minute Tesla rolled out its FSD system.

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48102098)

This is the pink elephant in the room and explains why Tesla keeps adding more cameras inside the car.

My prediction is that FSD is never fully adopted until it's safe enough for 0 attention span. Otherwise it's dangerous as fuck because it induces proles to sleep.

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48102019)

FSD will be normalized and adopted just as soon as (((car insurance companies))) decide it's in their financial interest to collectively lobby and push for that outcome. Which means it only has to be demonstrably safer than the average human driver. You can bank on that happening within the next few years (frankly we're already there, but let's give it a few more years because more people need to experience it like the OP in order to start spreading the word and helping to pave the way towards broader social acceptance).

People screaming about how FSD needs to be 99.999999999% safe are the 2024 equivalent of the old fogies standing around their mainstreet horse watering troughs back in 1904 ranting about how "these danged gasoline-powered auto-mobiles shouldn't even be allowed on the roads because they just go too gott-darn fast!!!!" LJL yeah, Ford will never sell a million Model-T's and usher in the automobile age now!

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48102072)

Accidents will of course continue to happen for the foreseeable future, and as a result drivers will continue to be legally required to carry car insurance policies for the foreseeable future. It will be generations before any of that changes. Insurance companies know that -- they just want to find a way to pay out less and less to increase their profitability as much as possible, and FSD is their golden ticket to achieving those cost savings in the short and medium-term future.

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48102115)

If risk decreases while insurance companies try to keep rates high and price gouge others will see the opportunity to make money by undercutting them.

(http://www.autoadmit.com/thread.php?thread_id=5595873&forum_id=2#48102214)