Tesla's Vaporware strikes again

[NismoNVT]

Tesla's FSD was promised as Full Self Driving/Level 3/4 on the SAE scale.
Tesla was kicking the legacy BS to the curb, by supporting the tech through multiple iterations of future software AND hardware upgrades as the tech evolved.

The $5k-15k pricing amount was to finance the rapid progress of a world safer then ever.

Well Elon released the following: Hardware upgrade pathways will not longer be available. If you have happen to have older hardware (The official term is HW3 and its still being installed in vehicles in current production until they launch HW4), best of luck. I have calculated some back of the napkin stats, and old hardware will be 200% safer then a human, and new hardware will be 500% safer then a human. Oh and btw thanks for the FSD donations, licenses are non transferable.

 

[97prizm]

I've been saying this for years. Tesla treats their vehicles like disposable flip phones. Musk himself said two years ago that Tesla won't support their vehicles over 10 years old. At the end of last year original roadster owners were told Tesla will no longer repair or support their vehicles. There is one guy in California who does repair Tesla roadsters. Honestly for level 4/5 autonomous driving to really work the roads need to communicate back to the vehicle on conditions.

 

[NismoNVT]

I've been saying this for years. Tesla treats their vehicles like disposable flip phones. Musk himself said two years ago that Tesla won't support their vehicles over 10 years old. At the end of last year original roadster owners were told Tesla will no longer repair or support their vehicles. There is one guy in California who does repair Tesla roadsters. Honestly for level 4/5 autonomous driving to really work the roads need to communicate back to the vehicle on conditions.

Level 3 and 4 are available starting this year in piecemeal form and the mfg will take responsibility I believe for all computer caused issues. Allow me to clarify:

Mercedes is offering this on select highways in Level 3 form (specific highways not everywhere), Nevada i believe is an everywhere state.
Acura/Honda has a Japan only product that is Level 3 SAE
The robotaxis from Waymo and Cruise are geographically fenced in very specific locales are doing Level 4.

 

[anderson]

in before the lock.

i may be old fashioned but if you own a car , drive it yourself!

 

[NismoNVT]

in before the lock.

i may be old fashioned but if you own a car , drive it yourself!

True, and Teslas are probably the best car to be driving yourself without sacrificing comfort and basic amenties. But this isn't a knock against that, its simply offering a product that they basically failed to deliver that would revolutionize the burden of long distance travel or the monotony of office/home commute.

 

[oggb]

I'd bet this doesn't matter because the take rate on FSD was/is much lower than most people think and that the vast majority of people buying or considering a Tesla know what FSD is but are happy enough with adaptive cruise control.

Burning a bridge with low traffic might be worth it in order to build a better bridge.

 

[UncleDave]

Tesla's FSD was promised as Full Self Driving/Level 3/4 on the SAE scale.
Tesla was kicking the legacy BS to the curb, by supporting the tech through multiple iterations of future software AND hardware upgrades as the tech evolved.

The $5k-15k pricing amount was to finance the rapid progress of a world safer then ever.

Well Elon released the following: Hardware upgrade pathways will not longer be available. If you have happen to have older hardware (The official term is HW3 and its still being installed in vehicles in current production until they launch HW4), best of luck. I have calculated some back of the napkin stats, and old hardware will be 200% safer then a human, and new hardware will be 500% safer then a human. Oh and btw thanks for the FSD donations, licenses are non transferable.

the downs

It's never delivered on it s promise
Elons missed every target and should just shut up.
It's misnamed (but its capabilities are accurately described)

the ups -

If you dont want it you arent forced to buy it - complaining is just whining
Few to no other cars on the market have hardware retrofits for upgrades - any upgradability is a plus
You can demo the software for a limited time- if you don't want it dont buy it.
Tesla hasn't recognized a penny of revenue on it and corporately its accounted for correctly.
The license agreement you sign says it's not transferrable- if you sign this an think otherwise its on you.

With all its faults it's the best offering.

It's already safer than a human by an extremely large percent and data is sent in real time to the NHSA.

 

[DwightFrye]

Wasn't it the Model T that had to be backed up a hill because the gasoline flow was gravity driven and it would stall if driven forward up a steep hill ? FSD of whatever you call it is still an emerging technology. Kind of like the Model T was at the time.

However, Tesla has accrued an absolutely insane amount of data that has and is being used to develop and improve their FSD software and hardware. No other company comes remotely close.

When, and it is just a matter of time, that FSD becomes quite commonplace Tesla will be the leader. So much so that the likelihood of other manufacturers having to license FSD software from Tesla if they want a superior system in their own vehicles is very high.
More $$$$ for Tesla.

 

[Slow Car Sport Mode]

Wasn't it the Model T that had to be backed up a hill because the gasoline flow was gravity driven and it would stall if driven forward up a steep hill ? FSD of whatever you call it is still an emerging technology. Kind of like the Model T was at the time.

However, Tesla has accrued an absolutely insane amount of data that has and is being used to develop and improve their FSD software and hardware. No other company comes remotely close.

When, and it is just a matter of time, that FSD becomes quite commonplace Tesla will be the leader. So much so that the likelihood of other manufacturers having to license FSD software from Tesla if they want a superior system in their own vehicles is very high.
More $$$$ for Tesla.

I don't quite agree, I don't think there is any way that FSD or Full Self Driving can EVER be adapted on a mass scale, EXCEPT for maybe some engineered environments like a few blocks in NYC, LA, and that may be a stretch.. with specially made lanes, sensors built in to guide the car much like a narrow aisle forklift uses sensors, an array of things set up specifically to assist the FSD.

You are correct, comparison has been made to Model T and Tesla. Getting to the point of "there is nothing new under the sun" here.

Wide-open highways with truckers in the real world??? I see a pipe dream. With limited practical application.. same way, for some an electric vehicle is all they need.. good for them.. I think the vision of the world as The Jetsons, is the dream of flying cars being foregone for electric/FSD. Putting charging stations at every gas station could be closer to the mass adaptation of EV. Charging times still an issue.

 

[97prizm]

Forbes wrote a really great article a couple years back. They've been looking at automated driving, and their data suggests--In other words, about 30% longer without an “accident” in manual (with forward collision avoidance on) or TACC than in Autopilot. Instead of being safer with Autopilot, it looks like a Tesla is slightly less safe.

But not a lot less safe. And if the predicted 3:1 ratio of accidents freeeway to non-freeway is too high, it might even be about the same. But almost certainly not 1.5 times better as Tesla’s numbers imply.

 

[UncleDave]

Forbes wrote a really great article a couple years back. They've been looking at automated driving, and their data suggests--In other words, about 30% longer without an “accident” in manual (with forward collision avoidance on) or TACC than in Autopilot. Instead of being safer with Autopilot, it looks like a Tesla is slightly less safe.

But not a lot less safe. And if the predicted 3:1 ratio of accidents freeeway to non-freeway is too high, it might even be about the same. But almost certainly not 1.5 times better as Tesla’s numbers imply.

Love to see actual data.

The NHSA gets this all regularly.

Every manufacturer should be forced to answer to their driver assistance packages.

 

[PandaBear]

There are things that can be proven by statistics, like whether traffic light is safe enough for up to 60mph road, then there are things that can only be disproved instead, like you should never assume stop sign is a good idea in the middle of a 75mph highway / freeway.

When cars replace horses, the world changed the way roads were designed and build, and it didn't happen overnight because of any one particular innovation locked in by one company. Roads got paved, traffic laws changed, traffic signs were installed, driving becomes a privilege instead of a right, insurance mandated. Imagine what happen back then when you had a horse and were told to buy insurance and take tests to see if you are qualified to go anywhere.

I don't think we are close to full self driving because we haven't gotten enough people dead yet, we will need to see how people die from these self driving failures enough to see why they fail and how our roads and traffic signals, traffic laws need to be changed to make them work better. This is how wars innovate technologies rapidly, you cannot just sit around waiting for things to happen slowly, and the rapid improvements will eventually trickle down to the civilian world. Test pilots and soldiers died so they have a lot of test data, we haven't gotten enough people dead with self driving yet.

Self driving now is just a gadget, not something that saves anyone a lot of money. Without saving money and lowering cost, nobody would fund it and take the risk on it. Is self driving feature saving enough insurance cost to justify the R&D and component cost? I am not sure yet, we will need a lot of people dead and a lot of lawsuits to see how it works out in the end. I think it will start with commercial trucking on certified roads before we will see them on family vehicles.

When I see the roads and traffic laws changed for self driving, that's when I know it is ready for prime time.

 

[Slow Car Sport Mode]

There are things that can be proven by statistics, like whether traffic light is safe enough for up to 60mph road, then there are things that can only be disproved instead, like you should never assume stop sign is a good idea in the middle of a 75mph highway / freeway.

When cars replace horses, the world changed the way roads were designed and build, and it didn't happen overnight because of any one particular innovation locked in by one company. Roads got paved, traffic laws changed, traffic signs were installed, driving becomes a privilege instead of a right, insurance mandated. Imagine what happen back then when you had a horse and were told to buy insurance and take tests to see if you are qualified to go anywhere.

I don't think we are close to full self driving because we haven't gotten enough people dead yet, we will need to see how people die from these self driving failures enough to see why they fail and how our roads and traffic signals, traffic laws need to be changed to make them work better. This is how wars innovate technologies rapidly, you cannot just sit around waiting for things to happen slowly, and the rapid improvements will eventually trickle down to the civilian world. Test pilots and soldiers died so they have a lot of test data, we haven't gotten enough people dead with self driving yet.

Self driving now is just a gadget, not something that saves anyone a lot of money. Without saving money and lowering cost, nobody would fund it and take the risk on it. Is self driving feature saving enough insurance cost to justify the R&D and component cost? I am not sure yet, we will need a lot of people dead and a lot of lawsuits to see how it works out in the end. I think it will start with commercial trucking on certified roads before we will see them on family vehicles.

When I see the roads and traffic laws changed for self driving, that's when I know it is ready for prime time.

Just happy you included in your excellent post here what I stated in mine.. that the roads need to change for this to be anything more than an exercise in limited, controlled scenarios - like on roads specifically made to make FSD seamless and smooth (as in that instance, it maybe can be.)

I don't think there is any way that FSD or Full Self Driving can EVER be adapted on a mass scale, EXCEPT for maybe some engineered environments like a few blocks in NYC, LA, and that may be a stretch.. with specially made lanes, sensors built in to guide the car much like a narrow aisle forklift uses sensors, an array of things set up specifically to assist the FSD.

The rest of my post was similar.

There will never be replacement for the humans.

 

[supton]

It's already safer than a human by an extremely large percent and data is sent in real time to the NHSA.

Stuff like that kind bugs me, I like to think I'm better than average, and don't need no stinkin' driver aids.

It's a shame that one can't own several cars. One with power nothing, and you have to drive it on a regular basis, so as to keep your skills fully honed. Then the automated or nearly so one for during the week, for when you have to drive with everyone else--and when you're most likely to not be paying full rapt attention.

IMO we're in a dead man's zone here, where the automation isn't good enough just yet. Yet people are are already becoming "fully" reliant on these aids. I don't remember what it is, Shannow described it pretty well, something about how a task at hand has to command the right level of attention, not too little but not too much. The too little region is where people stop paying attention--and bad things happen because it's not fully automated.

 

[supton]

Wasn't it the Model T that had to be backed up a hill because the gasoline flow was gravity driven and it would stall if driven forward up a steep hill ?

IIRC it was only when below half full or similar. Maybe near empty, not sure where the "pickup" was (actually a drain hole I'd think).

 

[PandaBear]

Just happy you included in your excellent post here what I stated in mine.. that the roads need to change for this to be anything more than an exercise in limited, controlled scenarios - like on roads specifically made to make FSD seamless and smooth (as in that instance, it maybe can be.)



The rest of my post was similar.

There will never be replacement for the humans.

I don't think our posts were similar. What I do think will happen is the roads will change, over time, like 20-30 years. It will replace humans, in 20-30 years in commercial operations to save money.

For commuters, it will probably not happen that fast, until it is very mature and affordable.

 

[97prizm]

I'd bet this doesn't matter because the take rate on FSD was/is much lower than most people think and that the vast majority of people buying or considering a Tesla know what FSD is but are happy enough with adaptive cruise control.

Burning a bridge with low traffic might be worth it in order to build a better bridge.

I looked it up and was shocked how low it was. Initially the take rate was 53% it's now according to people that track it 14% or only 1 in 7 paying for the "FSD".

 

[oggb]

I looked it up and was shocked how low it was. Initially the take rate was 53% it's now according to people that track it 14% or only 1 in 7 paying for the "FSD".

Turns out most people don't want/trust self driving vehicles yet. It doesn't help that most people only ever see negative headlines re: FSD

 

[NismoNVT]

Turns out most people don't want/trust self driving vehicles yet. It doesn't help that most people only ever see negative headlines re: FSD

I think people want FSD, but want a fully delivered product rather then beta launches and becoming beta testers for 15K a pop.