Apple Co Founder: "FSD is a frightening, horrible experience".

[bnsfengineer]

I realize the engineering of WayMo and FSD isn't a fair comparison , but it seems they are miles apart with technology! Does anyone think we just keep chasing the carrot with Tesla's FSD? I remember back in July 2017 I test drove and then purchased MS and was told FSD was almost complete and waiting for NTSB approval... I think we just keep getting fooled by the fools!! I now have FSD on Model Y and to me its horrible and shouldn't had wasted the 10K. The 2017 MS had enhanced AP and seemed smoother and less jerky in AP and the summon always worked with the push of the key fob or the stalk... if some of you remember they got rid of that feature. Model Y constantly failing to connect when I try to summon even beside the car also with the key fob I purchased.... FRUSTRATING and for the first time of purchasing a new car I have buyers remorse and the best part is I cant sell it since my 9 month old car has massively depreciated(probably 40%) since the drop and tax incentive. Apologize for the rant but wondered how other people felt about being screwed over?

 

[beatle]

I mean, it kinda does in this case? If you were unhappy with a thing you bought for $3000 in 2016 because in 2022 it still doesn't do what was promised, why would you pay $15,000 for that same thing on another vehicle? And if you do, that's kinda on you, not Elon at that point.

This wasn't a sales pitch that was given once, it was perpetuated over years, with huge progress coming in "two weeks" along with telegraphed price cuts to "get in early before you're priced out!" People have been strung along for quite a while. Being misled or swindled isn't a crime.

 

[ELETRO]

this is why paid FSD should get upgrades no matter how hard is the retrofit and for free obviously it can be done,
its just harness and hardware.
Elon saying not feasible economically but paying more than 5k and up+15k still can't make it up? i know its software too but when hardware is not capable to perform what then?
lawsuit?

 

[beatle]

That's my guess. Tesla figures the cost of engineering and performing a retrofit to all those FSD incapable cars will exceed any payout they might be hit with as a result of a lawsuit. If the HW3 cars were truly capable of FSD as-is, they wouldn't be engineering a more capable system, they'd only be looking to make the newer systems cheaper.

 

[SO16]

That's my guess. Tesla figures the cost of engineering and performing a retrofit to all those FSD incapable cars will exceed any payout they might be hit with as a result of a lawsuit. If the HW3 cars were truly capable of FSD as-is, they wouldn't be engineering a more capable system, they'd only be looking to make the newer systems cheaper.

It seems cheaper is a goal.

Tesla’s New Autopilot HW4: The Benefits Of Cost Reduction And Enhanced Processing Power | Torque News

Tesla is about to launch a new set of sensors for its electric cars; the HW4 hardware is a major change in the way your Autopilot 'sees the world'. It is also a major cost reduction measure. The HW4 will be available across the Tesla fleet of cars, not just the more affordable EV models. The...

www.torquenews.com

 

[ucmndd]

And starting in 2016 it was $3000.... and 2017, and 2018, and some of 2019.

This is disingenuous, as the only way to buy “FSD” was to first buy the prerequisite Autopilot package for ~$5k.

The true cost of FSD at introduction in 2016 was $8,000. (Edit: I see that you acknowledged this in a subsequent post).

At any rate, an odd way to object to the categorization of spending “tens of thousands” of dollars on FSD as somehow being egregious or out of touch. It’s basically two cars.

 

[Electroman]

'Robbed my family" is the kind of hyperbole a billionaire would say to attract attention and create faux outrage.

 

[Electroman]

The true cost of FSD at introduction in 2016 was $8,000. (Edit: I see that you acknowledged this in a subsequent post).

But then you need to put value on AP functionality and say you lost $3000 and not $8000.

By this silly logic, one could say they bought the car only for FSD and the whole purchase of the car is a waste of money without that.

 

[ucmndd]

By this silly logic, one could say they bought the car only for FSD and the whole purchase of the car is a waste of money without that.

By the same “silly” logic, they could sell FSD for a penny and say it’s “free”. The point stands.

The car is prerequisite to buying EAP.
EAP is prerequisite to buying FSD.

But the most honest apples to apples comparison between the way things are bundled now and the way they were before is to include the mandatory EAP costs.

 

[Knightshade]

This is disingenuous

I do not think that word means what you think it means.

, as the only way to buy “FSD” was to first buy the prerequisite Autopilot package for ~$5k.

The true cost of FSD at introduction in 2016 was $8,000. (Edit: I see that you acknowledged this in a subsequent post).

Not just acknowledged, but explicitly called that out as part of the problem with Wozs math. So the fact you took the time to edit your post and admit that, but leave the disingenous comment there, is.... not a great look.



And I also pointed out the difference was relevant because EAP actually delivered a bunch of, really awesome, features.

That's not wasting $ like the 3k he'd have paid for FSD may have been.

But it does point out that the price for the whole package including FSD didn't really start jumping a ton in price until 2022. If you paid $8000 for EAP plus FSD in 2016....or you paid $8000 for FSD (which then INCLUDED all the EAP stuff) in 2020... you still got the absolutely worth 5k EAP features either way and were only "cheated" out of whatever you thought the remaining FSD promises undelivered amounted to.

At any rate, an odd way to object to the categorization of spending “tens of thousands” of dollars on FSD as somehow being egregious or out of touch. It’s basically two cars.

Except, it's not.

If an option for a car is only available by buying the car, and the car is great, but the OPTION sucks, would you claim the buyer lost the entire cost of the car because the option wasn't good? Same principle here.

So only a few thousand dollars from 2016 to early 2019.... and really not significantly more than that until this year

You'd need to slap it on a lot more than 2 cars to get to tens of thousands going back to 2016.

I mean, I guess if he means "One in 2016, and 2 in 2023" you get there.

But then we are back to "How dumb do you have to be to buy something twice, for much more money, in 2023 when you were still upset at not getting value out of buying the same thing in 2016?"

 

[Midpack]

It doesn't matter how many cars he bought. The principle applies to anyone who bought into the promise that has not been delivered on, whether that cost them $3k, $15k, $30k, or more. It's also frustrating to experience the regression of performance from AP1, but that's more of another topic.

Maybe I’m an outlier, but I’ve been following full self driving in general, and it’s been clear all along it was going to be extremely tough to reach level 4 much less level 5. Lots of people (self included), not just Musk, were overly optimistic because the various programs working on it had self driving 80-90% figured out in a matter of years. It’s the remaining 10-20% that is proving to be elusive or worse. There are so many non-standard circumstances e.g. construction sites, missing road markings/signs, sudden obstructed views, etc. There are unlimited unpredictable behaviors by drivers, pedestrian, bikes, kids - irrational moves. And as far as I know, no one has really dealt with weather - e.g. what happens when a road is iced over (obscuring all markings), it’s snowing and it’s night? Kinda makes one wonder how a vision based system alone will ever work? But I’m buying a Tesla with no expectation FSD will actually be perfected in the next ten years at least. We’ll see…

 

[beatle]

Maybe I’m an outlier, but I’ve been following full self driving in general, and it’s been clear all along it was going to be extremely tough to reach level 4 much less level 5. Lots of people (self included), not just Musk, were overly optimistic because the various programs working on it had self driving 80-90% figured out in a matter of years. It’s the remaining 10-20% that is proving to be elusive or worse. There are so many non-standard circumstances e.g. construction sites, missing road markings/signs, sudden obstructed views, etc. There are unlimited unpredictable behaviors by drivers, pedestrian, bikes, kids - irrational moves. And as far as I know, no one has really dealt with weather - e.g. what happens when a road is iced over (obscuring all markings), it’s snowing and it’s night? Kinda makes one wonder how a vision based system alone will ever work? But I’m buying a Tesla with no expectation FSD will actually be perfected in the next ten years at least. We’ll see…

The way vision-only ADAS handles just the highway right now makes me think they're not even close to 50%, much less 80-90%. It's downright nauseating in traffic, and the driving behavior is annoying to cars following you. I quickly gained faith in AP1 and its ability to take care of things on the highway, but the current system is a big step back here.

IMO Tesla should have focused their efforts on increasing the autonomy level for highway driving where ADAS is most valuable. Imagine how amazing it would be if you could browse the internet on your phone or read a book in your car while it is moving, even if you could only do it below 30 mph on a divided highway. Anyone who commutes in heavy traffic would be over the moon at this advance. These little victories would allow them better purchase on the climb to higher levels in more situations and at higher speeds. It seems like Mercedes is taking this route in Nevada. The seemingly all or nothing approach with FSD seems doomed to fail as Tesla seems to think you can just jump over the mountain instead of climbing it.

Also to your point, there are a lot of non-standard circumstances in driving, and handling those in a more predictable environment such as a highway would let them hopefully make more progress.

As for inclement weather, there are times where even a person should not be on the road. I will say that I've tried AP out in heavy rain and I was surprised at how well it handled a poorly marked highway whose lane markers were almost completely obscured by rain, but in snowy conditions there are plenty of decent human drivers that make mistakes due to poor visibility or unpredictable losses of traction.

 

[2101Guy]

In the April 2019 Tesla Autonomy Day at 3:31:44:

(emphasis mine)

I have no idea why the Woz would be confused by what Elon said the car would do... /s

And yet, you have the phanboyz who continue to swear that Elon never ever ever, agreed that FSD would be level 5. and yet its right there..clear as day.

 

[AndrewZ]

I'm pretty sure Woz is suggesting that he upgraded to a car specifically with more cameras/sensors and promises, so it wasn't so much just additional FSD cost but the idea that he bought a new car that was supposed to be able to do more.

Also not sure I agree that his net worth has any bearing on whether $3k, $5k, $10k, $50k is a lot of money. Those are large sums of money, and I think people who tend to become billionaires are less inclined to lose sight of that. Warren Buffett has talked about hitting McDonalds for his breakfast every morning and knows how much it costs down to the penny.

 

[kabin]

I mean, it kinda does in this case? If you were unhappy with a thing you bought for $3000 in 2016 because in 2022 it still doesn't do what was promised, why would you pay $15,000 for that same thing on another vehicle? And if you do, that's kinda on you, not Elon at that point.

Everything approaches 20-20 vision in hindsight. Almost any early repeat FSDb purchaser can be argued to be a dolt but that's ad hominem stuff and doesn't relieve Elon or Tesla of misdeeds.

 

[Knightshade]

I'm pretty sure Woz is suggesting that he upgraded to a car specifically with more cameras/sensors and promises, so it wasn't so much just additional FSD cost but the idea that he bought a new car that was supposed to be able to do more.

But it'd have to be multiple new cars, buying FSD repeatedly after the 2016 promises not panning out, to reach the amounts he's discussing.

Like I said earlier- dude who traded his 2012 Model S in for a 2016 Model S and bought FSD for $3000 and he thinks he got jacked out of $3000? Totally reasonable argument.

Guy who did that... then after getting jacked bought several more of the same thing? Harder to be sympathetic to the idea he was "robbed"

Also not sure I agree that his net worth has any bearing on whether $3k, $5k, $10k, $50k is a lot of money.

I think when he's throwing around terms like his family being "robbed of so much money" it's kinda relevant.

A guy with a net worth of 100k not getting anything for a $30,000 purchase is going to be in a really bad spot. A guy worth $100,000,000+ paying 30k and not getting value isn't missing any meals.

And yet, you have the phanboyz who continue to swear that Elon never ever ever, agreed that FSD would be level 5. and yet its right there..clear as day.

Can you cite anybody in this thread doing that, or are you just jumping into another thread to troll with strawmen again?

 

[DarkForest]

And yet, you have the phanboyz who continue to swear that Elon never ever ever, agreed that FSD would be level 5. and yet its right there..clear as day.

I’ve heard so many people say that in different threads, so I assumed it was true. I recently watched that Tesla Autonomy Day video for fun and was floored to see that he straight up promised L5 by end of 2019.

Lots of wild promises were made, not goals or objectives. I can see why the Woz would call Elon out like that

 

[aronth5]

I’ve heard so many people say that in different threads, so I assumed it was true. I recently watched that Tesla Autonomy Day video for fun and was floored to see that he straight up promised L5 by end of 2019.

Lots of wild promises were made, not goals or objectives. I can see why the Woz would call Elon out like that

This is always an eye opener for many when it comes to what Elon "promised". Covers up thru early 2021.
I'm just glad I never believed in any of the dates.

FSD Timeline Promises (summary)

 

[Knightshade]

I’ve heard so many people say that in different threads, so I assumed it was true. I recently watched that Tesla Autonomy Day video for fun and was floored to see that he straight up promised L5 by end of 2019.

Lots of wild promises were made, not goals or objectives. I can see why the Woz would call Elon out like that

I mean if someone hasn't been listening that's kinda on them.

By the same token, Elon also told us he has no idea when things he's never done before will actually be done-- at least all the way back in 2018 in the 60 minutes interview.

People should not ascribe to malice that which can easily be explained by stupidity." (LAUGHTER) So-- so it's, like, just because I'm, like, dumb at-- at predicting dates does not mean I am untruthful. I don't know-- I-- we've-- I never made a mass-produced car. How am I supposed to know with precision when it's gonna get done?

Sub in "FSD" or "reusable orbital booster" or any other "never been done before so I'm just guessing on dates" item and the same sentiment remains true.

One would be wise to read all his predictions with this understanding.

 

[AndrewZ]

I'm not sure how you defend this Knightshade, there are people on this forum today who still believe HW3 vehicles will become robotaxis and you surely know that won't be the case nor was that what they bought.

This whole thing was a debacle and it continues being a debacle. Woz knows exactly what angle was being played, hence his focusing on speaking in dishonest marketing terms.