Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

AP2 - Definitely heading in the right direction...

This site may earn commission on affiliate links.
I can do @oktane one better. I had a demo P90DL AP1 car ordered for exactly one day when Elon made his grand announcement. The sales guy called me the next morning and asked if I wanted to swap for a state-of-the-art AP2 vehicle (for a few cents more) that was going to run circles around my P90DL. I said sure. He said, "Oh, sorry, we can't refund your $1,000 deposit even though the car never left the lot in New Jersey, but I guarantee you it'll be worth it." Added $40K to the price tag, and look where we are.

I'm delighted @Driver Dave loves his adventure. So do I. But it would be so much better if Elon had told the truth. I've been in the software development business for decades. I don't lie to my customers. Nor have I ever felt there was a need to do so just to improve my bottom line. The sad part (for Elon) is I've told this story on our blog to tens of thousands of followers and to literally dozens of friends, many of whom probably would have bought a Tesla by now if we had had a half way decent experience.
 
I seriously don't know what you are driving, but driving AP2 is definitely as usable as much as AP1 is.

Maybe you have forgotten how AP1 was? Do you have the latest update?

I drive both, back and forth, daily. There is little to no difference on basic highway driving, well lined local roads, stop and go traffic.

If I had to pick again, I would definitely still choose AP2.

AP2 will continue to get better. It has a future.

AP1 is stuck.


AP1 was extremely impressive. It never felt unsafe. AP2 is lackluster and feels unsafe. That's the best I can describe it. Yes AP2 will continue to improve, through beta testing by users like you and I.
 
I watched the video several times, can you point me to the time at which the various small videos are spliced together? I couldn't see it with the naked eye.

You can't tell by watching the video, which is the point. The video is made to appear as its one cut (but it's not). Eye witnesses have corroborated multiple takes on every segment of road to get the resultant fake video.
 
  • Disagree
  • Informative
Reactions: cdub and NerdUno
AP1 was extremely impressive. It never felt unsafe. AP2 is lackluster and feels unsafe. That's the best I can describe it. Yes AP2 will continue to improve, through beta testing by users like you and I.

For me, AP1 was solid on highways, but definitely hit or miss on local roads. Basically I wouldn't use it.

AP2 was lackluster definitely when it came out, but latest update allowed me to use it regularly. And definitely use it on local roads more than I do AP1.

Your experience may be different, this is just what I'm seeing.
 
You can't tell by watching the video, which is the point. The video is made to appear as its one cut (but it's not). Eye witnesses have corroborated multiple takes on every segment of road to get the resultant fake video.

I'm not sure I agree that it was multiple takes. However, the evidence is that it took them 179 tries to make the video over 3 days of trying.

Up to you if you think a car counts as being able to drive itself if it can only make it 0.5% of the time. Some people would call that fake, some wouldn't.
 
  • Disagree
Reactions: 3Victoria
Still can't fathom why people would be upset at getting the latest hardware, and watching it develop as the year goes on. I mean, how fun is that?

If I was sold a ride, and I wanted a ride, that would be great. If I was given a ride for free, or someone paid me to take the ride, that would be great.

What I was sold was a working product within a specific time. Tesla's development speed gives me ZERO faith that they will get there in any reasonable time. This is not slipping 2X. They are now 25 weeks in to a 1 week estimate, and nowhere near done. In what world is that OK? In what way does that give you any faith they will be able to do anything they promised?

I was sold a Concorde ride, and given a trip on a bus and a cargo ship. But I should be happy because it's the journey that matters and I got there without dying, right? What is this, United airlines? ;)
 
Last edited:
You mean the fleet learning that Tesla crowed about a year ago, but somehow didn't help them deliver AP2 quickly?

Tesla explains how its entire fleet is learning to be better self-driving cars together

AP2 uses the same radar as AP1 yet out of the gate it braked for every flat surface around.

Seriously, us "naysayers" actually own Teslas today, and were promised functionality 6 months ago. In what logical world do you trust anything Tesla says, much less some article on Electrek based on a few images uploaded through someone's router? In what way should I be happy about my situation?

I see you're already promising to buy AP on your Model 3. No matter how much it costs and how little it does? Or when it comes time to put your money down, are you actually going to look at Tesla's history of development and where it is before you pay? Because I didn't get to do that.
 
How was it fake? I took as nvidia hardware running the nvidia software showing where they where headed. I never ever thought the production cars were going to have this working out of the box.

I watched the video several times, can you point me to the time at which the various small videos are spliced together? I couldn't see it with the naked eye.

You can't tell by watching the video, which is the point. The video is made to appear as its one cut (but it's not). Eye witnesses have corroborated multiple takes on every segment of road to get the resultant fake video.

I'm not sure I agree that it was multiple takes. However, the evidence is that it took them 179 tries to make the video over 3 days of trying.

Up to you if you think a car counts as being able to drive itself if it can only make it 0.5% of the time. Some people would call that fake, some wouldn't.

Sigh... Let me set the record straight (in fact, I corrected you before @oktane).
Wait on adding AP2 and FSD or you will be sorry...maybe

There were 2 videos.

This first video from October 20, 2016 had cuts (although it is still possible they used footage from one continuous take; they just didn't think to show all views together as they did in the second video). There were 8 disengagements during the dry days (10/15, 10/17) that they could have possibly used footage. The other disengagements in the reports were either from wet days or after the video.
Full Self-Driving Hardware on All Teslas

A second video was released November 18, 2016 that had no cuts. There were no reported disengagements for the 20 miles they drove in November.
Tesla Self-Driving Demonstration

Disengagement report:
https://www.dmv.ca.gov/portal/wcm/c...a/Tesla_disengage_report_2016.pdf?MOD=AJPERES
 
Last edited:
If I was sold a ride, and I wanted a ride, that would be great. If I was given a ride for free, or someone paid me to take the ride, that would be great.

What I was sold was a working product within a specific time. Tesla's development speed gives me ZERO faith that they will get there in any reasonable time. This is not slipping 2X. They are now 25 weeks in to a 1 week estimate, and nowhere near done. In what world is that OK? In what way does that give you any faith they will be able to do anything they promised?

I was sold a Concorde ride, and given a trip on a bus and a cargo ship. But I should be happy because it's the journey that matters and I got there without dying, right? What is this, United airlines? ;)

You got fooled by the demo video ? How sweet !
 
Last edited by a moderator:
  • Like
  • Disagree
Reactions: NerdUno and bhzmark
Seriously, us "naysayers" actually own Teslas today, and were promised functionality 6 months ago. In what logical world do you trust anything Tesla says, ...In what way should I be happy about my situation?

What ya belly achin' about? If you've been traumatized by Elon's hyperoptimistic timeframes - I dunno, get some counseling or get a lawyer. Me - I'm lovin' it. When there's an alternative vehicle available as advanced as this one I'll look at it. Meanwhile - you're just screamin' at the sky makin' yourself upset. AP2 gets better every month - it's pretty rock solid now (despite what Oktane says - I still think maybe he got a dud or something - one of us should go driving with him and take some video to see what he's talking about) - it'll keep getting better. Nothing I can buy is as advanced. Try smilin' - might make you feel better.
 
I'm not sure I agree that it was multiple takes. However, the evidence is that it took them 179 tries to make the video over 3 days of trying.
This is a misinterpretation of the data, as pointed out about. They drove 20 miles, without any disengagements, for the Nov video.
AP2 uses the same radar as AP1 yet out of the gate it braked for every flat surface around.
Stationary objects have always been a problem for radar systems, mainly because there are so many of them: signs, fence posts, litter, potholes, etc. The previous method was to ignore them, and this led to the Mobileye-solution not 'seeing' the tractor-trailer in the fatal accident. AP2 is using the radar is a new way, and not just detecting objects directly in front of the car, but also in front of the car ahead. Creating a white-list of stationary objects to ignore is a better solution than ignoring them. (Note the white-list learning is spread to all cars)
 
  • Like
Reactions: teddytoons
Stationary objects have always been a problem for radar systems, mainly because there are so many of them: signs, fence posts, litter, potholes, etc. The previous method was to ignore them, and this led to the Mobileye-solution not 'seeing' the tractor-trailer in the fatal accident.

My recollection of the accident was that Tesla wasn't using radar at all and the camera didn't "see" the 18-wheeler because of the sun. The tractor trailer obviously wasn't a stationary object so the radar limitation wouldn't have applied anyway.
 
Sigh... Let me set the record straight (in fact, I corrected you before @oktane).
Wait on adding AP2 and FSD or you will be sorry...maybe

There were 2 videos.

This first video from October 20, 2016 had cuts (although it is still possible they used footage from one continuous take; they just didn't think to show all views together as they did in the second video). There were 8 disengagements during the dry days (10/15, 10/17) that they could have possibly used footage. The other disengagements in the reports were either from wet days or after the video.
Full Self-Driving Hardware on All Teslas

A second video was released November 18, 2016 that had no cuts. There were no reported disengagements for the 20 miles they drove in November.
Tesla Self-Driving Demonstration

Disengagement report:
https://www.dmv.ca.gov/portal/wcm/c...a/Tesla_disengage_report_2016.pdf?MOD=AJPERES


Which collaborates @oktane statement that the video were fake. Gave the illusion that it was doing something it isn't capable of, not even as a demo. Which is on par for tesla
 
  • Disagree
  • Funny
Reactions: cdub and NerdUno
Which collaborates @oktane statement that the video were fake. Gave the illusion that it was doing something it isn't capable of, not even as a demo. Which is on par for tesla
What? That doesn't match his characterization at all. People were confused by the fact there were two videos.

@oktane was referring to the 1st video in October, where there is a possibility of the video being cut together from multiple takes on the same road (given they switched views). He talked about eyewitnesses seeing filming over multiple days, which did happen for the 1st video (you can see from the report 10/15, 10/17 had dry disengagements).

However, others were talking about the second video in November, where there were no disengagements and no cuts (all views were shown at once, rather than cutting to different views). I don't believe there were any reports of eyewitnesses for this second video.
 
  • Informative
Reactions: dhanson865
What? That doesn't match his characterization at all. People were confused by the fact there were two videos.

@oktane was referring to the 1st video in October, where there is a possibility of the video being cut together from multiple takes on the same road (given they switched views). He talked about eyewitnesses seeing filming over multiple days, which did happen for the 1st video (you can see from the report 10/15, 10/17 had dry disengagements).

However, others were talking about the second video in November, where there were no disengagements and no cuts (all views were shown at once, rather than cutting to different views). I don't believe there were any reports of eyewitnesses for this second video.

I will defer to you on this. Is it possible the other video was also spliced together, despite the multiple camera angles?
 
Can someone please direct me to the thread, "AP2 - Definitely heading the right direction"? I had thought I "clicked" on that link, but I was wrong. The link I clicked on took me to this thread that looks like every other thread that starts off discussing AP2 and then somehow (quickly) ends up turning into the same complaints about how people got ripped off and duped some 6 to 9 months ago. To the best of my knowledge, no matter how hard we try, we cannot change the past. The fact is that the time machine has not been invented yet. You can't go back in time and not purchase your Tesla or the AEP/FSD option or purchase an inventory AP1 vehicle despite the announcement of new hardware that is supposed to be awesome. Wait, wouldn't it be ironic if Elon Musk actually invented a time machine and you could go back in time and not purchase your Tesla. Of course, if enough people did that, maybe he would not have had enough money to put the resources into inventing the time machine in the first place. Whoa. o_O

The truth is that I feel bad that people are not happy with their purchase. I truly do. I get it. I think that we all do. Everyone has that right. I was just hoping that maybe we could actually have a thread about AP2 that actually sticks to the topic/question raised and not whether a video was spliced 9000 times.

Be safe.
 
My recollection of the accident was that Tesla wasn't using radar at all and the camera didn't "see" the 18-wheeler because of the sun. The tractor trailer obviously wasn't a stationary object so the radar limitation wouldn't have applied anyway.
The truck was crossing perpendicular to the line of travel and therefore appeared to be stationary to the radar. I suspect the camera 'saw' the lane markings continuing under the trailer (oops).
 
I will defer to you on this. Is it possible the other video was also spliced together, despite the multiple camera angles?
Almost no possibility. They had no disengagements, what reason is there to splice?

Even the first video, I'm going to put it out there that they very likely didn't splice either. Otherwise there is no reason to do as many takes as they did.
 
  • Like
  • Disagree
Reactions: cdub and NerdUno