Thoughts about Autopilot 2.0

[donauker]

@calisnow, I congratulate you on all your effort in fighting all the AP2 FUD which a few posters seem to have made it their full time jobs in spreading! Based on my actual experience I have long ago decided to ignore them since most of their statements are so far removed from the facts of my experience that I can only assume that they either aren't driving the AP2 that I am or else they have an agenda.

At this time I have about 15,000 miles of AP2 enabled time and I know that for much of my use it was better in some areas then AP1 was a year ago when my 5,000 miles of AP1 use ended. I consider the latest versions of AP2 to be well worth the price that I paid for EAP & FSD even though they don't yet do near everything that I expect they will in the future. Having given my first deposit to Tesla in May 2nd, 2007 I have now spent over 10 1/2 years waiting for Tesla to deliver something or other and based on their latest reveal I expect I will always be waiting for them to deliver something!

I look forward to adding about another 2,000 miles of AP2 enabled time in this amazing vehicle before the end of this week as we once again do something that no other production car on that planet can do! I get to travel cross country while marveling at how great a job AP2 does at steering will I sit there and monitor it all while anticipating the next stop where we will fuel up for free while we get to play with our dog and take a welcome restroom break or grab a bite to eat.

 

[oktane]

@calisnow, I congratulate you on all your effort in fighting all the AP2 FUD which a few posters seem to have made it their full time jobs in spreading! Based on my actual experience I have long ago decided to ignore them since most of their statements are so far removed from the facts of my experience that I can only assume that they either aren't driving the AP2 that I am or else they have an agenda.

At this time I have about 15,000 miles of AP2 enabled time and I know that for much of my use it was better in some areas then AP1 was a year ago when my 5,000 miles of AP1 use ended. I consider the latest versions of AP2 to be well worth the price that I paid for EAP & FSD even though they don't yet do near everything that I expect they will in the future. Having given my first deposit to Tesla in May 2nd, 2007 I have now spent over 10 1/2 years waiting for Tesla to deliver something or other and based on their latest reveal I expect I will always be waiting for them to deliver something!

I look forward to adding about another 2,000 miles of AP2 enabled time in this amazing vehicle before the end of this week as we once again do something that no other production car on that planet can do! I get to travel cross country while marveling at how great a job AP2 does at steering will I sit there and monitor it all while anticipating the next stop where we will fuel up for free while we get to play with our dog and take a welcome restroom break or grab a bite to eat.

Not sure how your experience can be so vastly different. Do you feel AP2 drives better than you, worse than you, or at the same level? For me I'd say AP2 drives much worse than I do and practically anyone else I associate with. Depending on their experience, other readers may feel your post is the one "far removed from the facts" and perhaps you are the one with the "agenda"?

Don't get me wrong, I feel bad for the engineers and I know they are doing the best they can reproducing MobilEye's work in 1 year. I have respect for them. It's the marketing sleeze that pisses me off.

 

[donauker]

Not sure how your experience can be so vastly different. Do you feel AP2 drives better than you, worse than you, or at the same level? For me I'd say AP2 drives much worse than I do and practically anyone else I associate with. Depending on their experience, other readers may feel your post is the one "far removed from the facts" and perhaps you are the one with the "agenda"?

Don't get me wrong, I feel bad for the engineers and I know they are doing the best they can reproducing MobilEye's work in 1 year. I have respect for them. It's the marketing sleeze that pisses me off.

I am not the one that posted hundred of posts bad mouthing AP2. I posted one post stating the amount of my experience on both AP2 and AP1 and how I feel about it. Hardly an agenda.

Yes, my AP2 car can certainly do a much better job of staying consistently positioned in the lane over a 12 to 14 hour day while I get to be much more aware of the big picture when traveling.

 

[SomeJoe7777]

It's possible we'll never get there - and Tesla made a mistake in assuming they could replace MobilEye with the AP2 sensors and Tesla Vision. Though I don't expect that to happen.

Let's keep in mind what happened and why an unexpected turn of events caused a lot of difficulty for Tesla.

Tesla's original plan with AP2/TeslaVision was to have cars equipped with BOTH the TeslaVision modules and the MobileEye EyeQ3. Those vehicles would be the key to TeslaVIsion development. All of them would run on the EyeQ3 chip for functionality while TeslaVision was being developed and running in shadow mode on the vehicle. All data would be sent back to Tesla so that they could refine TeslaVision until it was at the point that it was as accurate or more accurate than EyeQ3. Then the vehicles would be shifted over to use TeslaVision as the primary AP software.

When Tesla tried to enter into a contract to have MobileEye supply the EyeQ3 for those new refresh vehicles, MobileEye demanded that Tesla give them their hyper-accurate mapping and vision data as part of the deal. Tesla refused, and MobileEye pulled out of the contract, leaving Tesla with no EyeQ3 chips.

That left Tesla with the unenviable and highly difficult task of not only having to develop TeslaVision far more rapidly than expected, but to do so after already having advertised AP capability to those new car buyers, AND having to do the development without having a reference software running side-by-side to compare and emulate.


What Tesla should have done when this sequence of events occurred is put out an offer to existing AP1 owners that they could volunteer to become a TeslaVision development car. Those cars would be retrofitted with TeslaVision to run side-by-side with EyeQ3. Those owners would be compensated with a discount on a future AP2 vehicle. If they had done this with say, 1000 AP1 cars, they would have had enough of a development platform to seriously speed up TeslaVision development, and it wouldn't have cost too much. The AP2 retrofitted hardware could be pulled out of the car and used for spare parts after the development cycle was over, and those cars again resold as AP1 CPOs.

 

[oktane]

I am not the one that posted hundred of posts bad mouthing AP2. I posted one post stating the amount of my experience on both AP2 and AP1 and how I feel about it. Hardly an agenda.

Yes, my AP2 car can certainly do a much better job of staying consistently positioned in the lane over a 12 to 14 hour day while I get to be much more aware of the big picture when traveling.

Agenda is the wrong word, perhaps "bias"? You clearly are a loyal Tesla owner as evidenvced by your long history of Tesla ownership (nothing wrong with that) but I have no longer have loyalties to Tesla even though I used to be a die-hard fan.

 

[NerdUno]

@SomeJoe7777: Another nice touch when Tesla knew that they could no longer deliver AP2 vehicles with MobilEye would have been to alert the purchasers and give them the option of canceling their orders. Instead, Tesla chose the stonewall path and, to this day, has failed to acknowledge that they misled Q4 2016 purchasers. But, what goes around comes around, and Tesla's day of reckoning is fast approaching.

Tesla’s Burning Through $4 Billion a Year

 

[iffatall]

I don't have AP HW on my car, and I have zero experience using AP, except for a couple of days of driving AP1 on a loaner well over a year ago. However I avidly read almost all AP related threads here, and follow both sides of the argument very closely.

I gotta say, based on reading everything, it was very obvious to me AP2 was far behind AP1 until a couple of months ago, even though there had been incremental improvements over time. But now, I get the feeling its driving quality is arguably at par (ignoring not reading speed limit signs and not displaying vehicles in adjacent lanes the whole time).

Just thought I should post this, to give an outsider's $0.02 in the middle of this whole argument.

 

[d21mike]

I don't have AP HW on my car, and I have zero experience using AP, except for a couple of days of driving AP1 on a loaner well over a year ago. However I avidly read almost all AP related threads here, and follow both sides of the argument very closely.

I gotta say, based on reading everything, it was very obvious to me AP2 was far behind AP1 until a couple of months ago, even though there had been incremental improvements over time. But now, I get the feeling its driving quality is arguably at par (ignoring not reading speed limit signs and not displaying vehicles in adjacent lanes the whole time).

Just thought I should post this, to give an outsider's $0.02 in the middle of this whole argument.

I tend to agree with you. Seems to me there is a lot of pent up anger from the early AP2 customers. But this poll should be about the current state of AP2. And looking at the poll results it is clear the current state of AP2 is being well received. I personally love it and use it all the time. But since I got my car on 9/28/17 I do not have the history that early AP2 customers have so I can only respond to what it does now. And I have zero experience with AP1.

 

[THXULTRA]

So far I have been really impressed with AP2. I can tell that it continues to get better and I'm looking forward to future software updates. So far I have noticed a few things with the current version.
1.) when crossing intersections sometimes it goes to the left more then I'm comfortable with and I end up taking control.
2.) Seems to break harder then I do in traffic. Seems to brake later then I would.
3.) some roads wants to get in turn lanes instead of going straight.

For the most part though works great. Really like that Tesla is improving the software constantly. Looking forward to my next update to see how things improve even more.

 

[donauker]

Agenda is the wrong word, perhaps "bias"? You clearly are a loyal Tesla owner as evidenvced by your long history of Tesla ownership (nothing wrong with that) but I have no longer have loyalties to Tesla even though I used to be a die-hard fan.

So since you no longer have loyalties to Tesla you have a need to post hundreds of posts on this forum slamming AP2 with over the top statements? Based on my personal experience your statements are so exaggerated they have always appeared to me to be obvious FUD.

By the way what other car could I drive today that does over 99% of my cross country driving for me? Say I want just this one thing and I forget about the fact that I get free fuel and the fastest accelerating production car to boot!

But, what goes around comes around, and Tesla's day of reckoning is fast approaching.

Tesla’s Burning Through $4 Billion a Year

So what does this additional obvious FUD have to do with the current topic? As a business owner for 38 years I have long known that investing in a business is not at all equivalent to burning money. It is investing in future profits and with what I have witnessed Tesla and Elon pulling off in the last 10 years I expect profits won't be that big of a problem.

 

[NerdUno]

There's a difference in investing in your business and investing $4 billion a year of other folks' money in your business. At $10K profit (or less) on most Model 3's, that's a lot of cars to make up for money that's already been spent. It doesn't require a math genius to figure out that, at a current burn rate of $8,000 per minute, Tesla better pick up the pace in the assembly line...soon.

And, by the way, everything you happen to disagree with isn't FUD or over the top. So kindly leave the insults out of the discussion. I think we all appreciate where you're coming from.

 

[Sawyer8888]

“However, we still expect to achieve full self-driving capability with safety more than twice as good as the average human driver without making any hardware changes to HW 2.0. If this does not turn out to be the case, which we think is highly unlikely, we will upgrade customers to the 2.5 computer at no cost.”

“....If this does not turn out to be the case, which we think is highly unlikely, we will upgrade customers to the 2.5 [or 3.0 or 3.5 or 4.0...] computer at no cost.”

The hype of self-driving has faded for me. AP2 is decent, AP1 is still a bit better and Elon keeps smiling but he is probably just as frustrated. His pride and his advisors won't allow him to keep us posted on actual attainable timeframes - never have, and that's the problem for me. I'm pretty much over it now. Would probably take another year or two to get over it if I had bought FSD.

 

[donauker]

There's a difference in investing in your business and investing $4 billion a year of other folks' money in your business. At $10K profit (or less) on most Model 3's, that's a lot of cars to make up for money that's already been spent. At a current burn rate of $8,000 per minute, they better pick up the pace in the assembly line.

And, by the way, everything you happen to disagree with isn't FUD or over the top. So kindly leave the insults out of the discussion. Name calling isn't helpful. I think we all appreciate where you're coming from.

I did no name calling. FUD is a very clearly defined acronym for Fear, Uncertainty and Doubt. This seems like a perfect fit to the type of posts I am referring to. Calling out posted FUD is not name calling.

We all know that Tesla is working hard at ramping up the Model 3 assembly line and there is no reason to believe that it is not going to continue to accelerate. History shows that they accelerated the Model S line far beyond what FUD posts a few years ago stated was an impossible number to reach.

 

[Leopold Stotch]

Not sure how your experience can be so vastly different. Do you feel AP2 drives better than you, worse than you, or at the same level? For me I'd say AP2 drives much worse than I do and practically anyone else I associate with. Depending on their experience, other readers may feel your post is the one "far removed from the facts" and perhaps you are the one with the "agenda"?

Don't get me wrong, I feel bad for the engineers and I know they are doing the best they can reproducing MobilEye's work in 1 year. I have respect for them. It's the marketing sleeze that pisses me off.

It's not a question of AP2 verses YOU.
It's a question of AP2 + YOU verses just YOU

I'd go with AP2 + YOU anyday. I'd be much happier and feel much safer if everyone was driving Tesla's using AP2 while they pay attention (as Tesla states you to do).

 

[NerdUno]

Why is it that all the cheerleaders have to make things personal? It's just a damn car, Tesla fans. Nobody violated your sister.

 

[NerdUno]

I did no name calling. FUD is a very clearly defined acronym for Fear, Uncertainty and Doubt. This seems like a perfect fit to the type of posts I am referring to. Calling out posted FUD is not name calling.

We all know that Tesla is working hard at ramping up the Model 3 assembly line and there is no reason to believe that it is not going to continue to accelerate. History shows that they accelerated the Model S line far beyond what FUD posts a few years ago stated was an impossible number to reach.

Burning through $8.000 a minute isn't FUD. It's simple math. Divide $4 billion by 365x24x60. Tesla can ramp up production 'til the cows come home, and they're never going to be able to produce a new vehicle in less than 90 seconds. Bloomberg doesn't play the FUD game. They deal in cold, hard facts that investors use to determine where to invest. Given the enormity of the cash burn, the likelihood that investors are about to dry up for Elon's "vision" is quite real. Not sure I'd want to plunk down $250,000 in advance for a 2020 model Roadster given the current financial situation. But, here's hoping all of the Tesla Worshippers go for it. Better yet, help Elon out and buy two.

 

[oneday]

Why is it that all the cheerleaders have to make things personal? It's just a damn car, Tesla fans. Nobody violated your sister.

So you can’t accuse me of thumbs downing without saying why.

You comments are off topic and better suited elsewhere.

 

[Electroman]

Based on my personal experience your statements are so exaggerated they have always appeared to me to be obvious FUD.

I have been saying this for a long time now. oktane may have a gripe about the absence of FSD, but his AP2 'sky is falling' persistent comments are overly exaggerated. Nit picking on the edge cases, and repeating it over and over again.

Appreciate @calisnow for his valiant fight against this AP2 FUD.

 

[calisnow]

Your other thread not so long about AP2.0 improvements (on latest firmware) has given me hope. If more AP2.0 owners come back and report the same/agree with you, that would be wonderful.

I believe there is a thread with quite a few other owners who have similar impressions of AP2 .40-and-later. Too lazy to dig for it but it's less than a month old thread. I believe it may have been the thread about .40 or .42

10yrs of MobileEye vision processing technology Tesla is trying to do in 1yr - that is a marvelous undertaking and commendable (if they can achieve close to it). It is the false promises on timelines (likely due to underestimating the complexity) that have been most painful to digest.

I do acknowledge the realness and validity of your feelings. You're not expressing them in a manner which suggests you have an agenda nor do you come across like an unstable crank with a personality disorder stinking up the forums.

I will also finally admit that my perspective biases me - I love tech and I love cars. I also am the type who sets impossible goals for myself as a tool to achieve things in my own life. So having read Elon's bio and also being a huge fan of the "aim for the impossible as a motivational tool" technique - Elon's missed promises just don't phase me because he still delivers the best driving experience and autonomous experience in the world available for purchase. And I get immense pleasure observing the car learn and improve from update to update. I think we are the lucky few living a grand moment in history and I'm savoring every bit of it.

I know Elon has a world class all-star AI team working 24/7 on autopilot and I'm far more in awe of what they are achieving before our eyes than I am disappointed that they haven't yet made autopilot as reliable as my washing machine.

If however I was a non tech head who just wanted a self driving car and thought I was getting one last year I can understand the frustration.

 

[calisnow]

I have been saying this for a long time now. oktane may have a gripe about the absence of FSD, but his AP2 'sky is falling' persistent comments are overly exaggerated. Nit picking on the edge cases, and repeating it over and over again.

Appreciate @calisnow for his valiant fight against this AP2 FUD.

Thanks for the word of encouragement - much appreciated, I mean it.