It's weird because the video they released for that announcement seemed really good, but they haven't released that firmware yet for some reason.
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Elon: "Feature complete for full self driving this year"
- Thread starter dckiwi
- Start date
it only seemed really good until the CA DMV disengagement report came out
StealthP3D
Well-Known Member
Yes, it definitely sounds like a "work in progress".
Er...well it would have if you had published the entire announcement. Because then you would have read this:
"Before activating the features enabled by the new hardware, we will further calibrate the system using millions of miles of real-world driving to ensure significant improvements to safety and convenience. While this is occurring, Teslas with new hardware will temporarily lack certain features currently available on Teslas with first-generation Autopilot hardware, including some standard safety features such as automatic emergency braking, collision warning, lane holding and active cruise control. As these features are robustly validated we will enable them over the air, together with a rapidly expanding set of entirely new features. As always, our over-the-air software updates will keep customers at the forefront of technology and continue to make every Tesla, including those equipped with first-generation Autopilot and earlier cars, more capable over time."
Never has Tesla outright stated (or even implied) that full autonomy was ready today. They have ALWAYS been clear it was a work in progress.
And, on the next video they decided not to bother with the whole CA DMV disengagement reporting.
You're mixing two features together.
EAP was what needed validation, but even that was a lie. You first have to have SW to say you need to validate it, but the SW wasn't done. It certainly wasn't during the Dec time frame that was estimated for EAP.
FSD mostly hid behind the validation, and regulatory approval where they didn't say much about it.
FSD had no specified goals until very recently.
It really comes to the fact that EAP was a necessary lie, and FSD was an unnecessary lie.
EAP was a necessary lie because Tesla basically got themselves fired as a MobileEye customer. For all sense and purpose.
FSD was a totally unnecessary lie because they simply didn't need to make that promise. Especially since they had barely even started working on EAP.
StealthP3D
Well-Known Member
No, neither were a "lie". They are both works in progress. That's the beauty of Over-the-Air updates.
And for those who forgot or didn't see it, the videos are the Nov 2016 videos at Videos | Tesla and the report is at Wayback Machine (from Autonomous Vehicle Disengagement Reports 2016).
CA DMV Report Sheds New Light On Misleading Tesla Autonomous Drive Video was written on the above.
One can still see 2015 to 2018 reports at Testing of Autonomous Vehicles.
That's what you get when you sell something that does not yet exist. You find out that you can't get there from here. Or, in this case, they found out that the cars did not have all the needed hardware. So now they're stuck trying (still, apparently, with only limited success) to upgrade the computers in HW2.0 and HW2.5 cars. And there's no certainty that HW3.0 will be sufficient. I think it will not be, and in the end a bunch of people who paid for FSD will find that their cars simply cannot accommodate the needed hardware.
True. But they told people "If you pay for FSD today, your car will be robotaxi-capable" at some unspecified time in the future. That is an explicit promise that the car in question would (eventually) become FSD, and if that generation of cars is junked due to old age before FSD is available, the promise was broken. And if the eventual FSD system requires hardware that cannot be retrofitted into those cars (as I expect will be the case) then the promise was broken.
Selling something that does not yet exist, or selling a promise based on "work in progress" is a very poor idea if you want satisfied customers. Most works in progress are dead ends. It's entirely possible that in the end Tesla will have to lease technology from Waymo and build cars with entirely different sensors. And buying something that does not yet exist is gambling, pure and simple.
Betting on Roulette in a casino, is gambling on a pure chance outcome, with the odds slightly adjusted against you. Betting on the stock of a start-up company is gambling on the abilities of the people running the company. Buying and holding the stock of an established company with a proven product is gambling with the odds in your favor.
Buying FSD is donating money to a worthwhile project, with the odds strongly against getting what you paid for in time to make use of it on a car built today. FSD will come. I have no doubt of that. But it's very unlikely to come to my 2018 Model 3 because it's going to need very different hardware than my car could be upgraded to. Tesla has committed itself to a very narrow path to FSD, and that path might turn out to be the wrong one. FSD right now is a big gamble for both Tesla and for the people who pay for FSD.
I just hope that Tesla will be willing to adapt if their path turns out not to be the one that leads to self-driving cars.
Is the 2019 report out yet? I thought it was usually released in February.
Barklikeadog
Active Member
When do you think you'll get the email for the quick n easy Lidar install?
StealthP3D
Well-Known Member
That's false that a purchaser's actions could break a "promise" by Tesla. Tesla never promised your car would not be totalled before their development work was complete. That would be an insane promise to make. If you total your car before the car is FSD capable, that's not on Tesla, it's on you. Tesla might make the safest cars in the world but they cannot prevent you from totaling your car before their development is ready and approved. That much should be obvious.
BlueRocket
S90D HW3.0 MCU1 "No FSD 7 years" 2022.8.10.16
That video is such BS and it is already over 3 years old! I went for it... Hook, line and sinker. I wonder how many pedestrians they ran over before they got the final take?
I won't, because I didn't buy FSD. I bought what was actually available and I love it. I think you're agreeing with me, though.
I didn't say "totaled." I said "junked due to old age."
And I didn't say anything about a purchaser's actions breaking a promise by Tesla. I talked about Tesla's failure to deliver on a promise during the time that the promise still has meaning. Cars wear out. When they promise "Your car will do X" there is an implied time frame because cars only last for a limited time. If your car wears out due to normal wear and tear before FSD is available, they've broken their promise.
And they've tacitly admitted that they cannot fulfill the promise because they're no longer making it. The deal offered to me (which I declined, but many others paid for) was "Your car [that is, my specific car] will be capable of operating as a robo-taxi at some unspecified time [but by implication within the normal expected lifetime of the car] and with the hardware the car contains at the time of purchase." Now the promise is a list of features that will still (maybe for decades) require constant driver attention. And they're scrambling to try to figure out how to upgrade the hardware after admitting that their claim ("Your car contains all the needed hardware") was wrong.
All the inevitable pitfalls of selling a technology that does not yet exist.
What they should have said was, "We don't know if we can do this. We think we can. If you give us X dollars now, and if we succeed, you'll get the technology. Depending on market conditions, people who wait might have to pay more than you pay now." They'd probably have gotten almost as many orders.
Rumor has it that it was Nvidia.
It was running on Nvidia hardware as AP2/AP2.5 is Nvidia hardware.
I dunno.
My post was factual without actually relying on any facts.
But, that is a good point that we haven't seen the 2019 report yet as far as I know. So we're only about 99.99% sure that Tesla isn't bothering the disengagement reports.
Daniel in SD
(supervised)
I bet they will have disengagement reports for 2019. Uber tried not reporting testing to the DMV back in 2016 and it didn't go over well. I doubt they're doing any testing other than the premapped demo rides and video. The "sneak preview" shows they've still got a ways to go on perception which of course doesn't require any autonomous testing.I dunno.
My post was factual without actually relying on any facts.
But, that is a good point that we haven't seen the 2019 report yet as far as I know. So we're only about 99.99% sure that Tesla isn't bothering the disengagement reports.
I agree with the Cruise CTO that disengagement data is mostly useless for evaluating progress. Though I suppose that's only when you start to get into the thousands of miles between disengagements. I doubt Tesla is there yet...
BTW, during Autonomy Day 2019 Musk said they were going to take investors for FSD drives through the day. Later in internal email he wrote that they got a very positive feedback. But has anybody seen a single first-hand response from somebody who took a ride?
I’ve tried to search, and all I found was official Tesla’s video. It’s hard to imagine nobody has shared at least something anonymously on Reddit etc. if rides actually took place.
I’ve tried to search, and all I found was official Tesla’s video. It’s hard to imagine nobody has shared at least something anonymously on Reddit etc. if rides actually took place.
I am not 100 but I do believe that Gali, hyperchange, did comment on the rides. Something like "The bounding boxes were so tight", I am probably misquoting. I dont remember if he said anything else about the ride though.
Similar threads
