.48 feels like AP2 finally passed AP1

[Joe F]

@Carl

But we're not even at feature parity, that much is simply factual: speed sign recognition is missing, non-highway lane change is missing, non-lane change multi-lane vehicle ID is missing, vehicle ID type is missing for AP2 (and if you want to count it, auto-wipers are missing from AP2).

AP2 has no feature AP1 does not. AP1 has several features AP2 does not. That's really the point I'm trying to make here. It is very hard to argue AP2 is at parity, let alone beyond, until it has the same or at least something different extra than AP1.

I'm sure you can't possibly believe IC vehicle display it more important that what your own eyes see through your windshield. I think you're more intelligent than that.

I'll give you the speed limit advantage goes to AP1. That needs to be addressed.

But your bold statement quoted above is extremely misleading to anyone finding this thread while researching a Tesla purchase, and it's flat out false. Here, I'll correct it for you in the interest of being factual:

AP1 does not have the additional cameras, enhanced ultrasonic sensors and vastly increased processing power that defines AP2 H/W.

Does that nullify your contention that AP2 does not have the same or something different than AP1?

Please don't fixate on non essential pretty pictures on the IC. Ignoring the enhanced H/W features of AP2 while continuing to draw attention away from the potential the suite provides is disingenuous and appears more self serving.

 

[⚡️ELECTROMAN⚡️]

Not this crap again.

Come on.

Have other people said that they had the same thought?

 

[Carl]

Why is it AR or us that don't enjoy parity that should document it? I think they who claim parity should document it, like in pharma.
Tesla made a video once, why can't they make another? They earn money on this product? Or invite journalists to a test with the competition, like BMW does?

Fully agree. Anyone on the forum claiming anything in respect of parity or lack of parity should base their claim on experience. And I would even more love to see a real, December 2017, comparison of BMW/Audi/Volvo/MB/Tesla semi-autonomous driving (same road, same obstacles etc.). That would be a great Xmas present!

 

[AnxietyRanger]

Fully agree. Anyone on the forum claiming anything in respect of parity or lack of parity should base their claim on experience. And I would even more love to see a real, December 2017, comparison of BMW/Audi/Volvo/MB/Tesla semi-autonomous driving (same road, same obstacles etc.). That would be a great Xmas present!

We don't need personal experience to assess facts, though. One would argue personal experience is too limited anyway, because AP uses mapping and thus can highly depend on region. For a bigger picture we need to assess lots of inputs (lots of member insights). I think this is what we've been trying to do here.

 

[AnxietyRanger]

Have other people said that they had the same thought?

A lot of people in recent times have expressed doubt over me, who apparently haven't followed my Model X years on TMC, as some form of ad hominem - and it gets tiresome, even though some senior members have corrected these misconceptions, which I appreciate.

Just trust me, I'm one person, I have this single account on TMC only, I have one current Tesla and I have had one past Tesla. It really is that simple.

 

[AnxietyRanger]

I'm sure you can't possibly believe IC vehicle display it more important that what your own eyes see through your windshield. I think you're more intelligent than that.

It isn't more important, but it is something AP1 gives that AP2 does not give. When assessing the objective reality, what I feel is important does not matter. What are the facts matters.

I'll give you the speed limit advantage goes to AP1. That needs to be addressed.

Agreed.

But your bold statement quoted above is extremely misleading to anyone finding this thread while researching a Tesla purchase, and it's flat out false. Here, I'll correct it for you in the interest of being factual:

I don't intend to, though. I have repeated time and time again - indeed the whole premise of this thread is - this is about the current software reality. The title of this thread, ".48 feels like AP2 finally passed AP1", does not reflect hardware reality either.

AP1 does not have the additional cameras, enhanced ultrasonic sensors and vastly increased processing power that defines AP2 H/W.

Obviously true. Anyone making even the most basic research knows this, of course.

Does that nullify your contention that AP2 does not have the same or something different than AP1?

Not in the case of current software reality. Future potential for AP2 obviously is immensely bigger on an unknown timescale.

Please don't fixate on non essential pretty pictures on the IC. Ignoring the enhanced H/W features of AP2 while continuing to draw attention away from the potential the suite provides is disingenuous and appears more self serving.

Facts are pesky little things, because we can't wish them away. If we're saying AP2 is at parity with AP1, either we need to quantify that suitably (e.g. AP2 steering is at parity with AP1), or we need to wait until it really is at parity.

If we're just talking about hardware, why talk of parity at all? Obviously AP2 obliterates AP1. But that's not the topic of this thread, nor is it the topic of the AP1 parity conversation.

 

[Carl]

We don't need personal experience to assess facts, though. One would argue personal experience is too limited anyway, because AP uses mapping and thus can highly depend on region. For a bigger picture we need to assess lots of inputs (lots of member insights). I think this is what we've been trying to do here.

Fully agree. Lots of inputs and insights of members, from all regions, who can experience and comment on the factual AP1/AP2 parity, is needed. I, for sure, am not one of them

 

[Joe F]

It isn't more important, but it is something AP1 gives that AP2 does not give. When assessing the objective reality, what I feel is important does not matter. What are the facts matters.

The fact is, AP is comprised of both H/W and S/W. They're intertwined. You can't have one without the other. Your boldly stating AP2 has no feature AP1 does not, is flat out false, and IMHO intentionally misleading. You're not entitled to your own facts.

 

[brendon1970]

I do wish that Tesla would provide a feature roadmap for AP. Then a lot of the discussions in this thread would potentially go away. I suspect they will not, as there is undoubtedly technical and commercial risk in doing so.

For example, the vehicle “classification” and adjacent lane display - maybe it was a design dead end? Maybe its never coming back? Maybe its presence gave confidence where such confidence was not warranted - after all nobody can quantify what if anything AP was doing in relation to that information.

I think there is plenty of room for UI improvement with AP, however I am also of the opinion that we are not currently seeing the full picture - there are plenty of questions about what UI will be present around the more automated portions of enhanced AP. I have to believe that Tesla have at least preliminary design concepts for all anticipated functionality - would love to see some of that!

Cheers
Brendon

 

[AnxietyRanger]

The fact is, AP is comprised of both H/W and S/W. They're intertwined. You can't have one without the other. Your boldly stating AP2 has no feature AP1 does not, is flat out false, and IMHO intentionally misleading. You're not entitled to your own facts.

I am talking about the current software reality. I have made that clear time and again - and that is the context of this whole thread. In fact, that is the context of the whole "AP1 parity" talk.

Why would anyone talk of AP1 parity if hardware was the comparison? Obviously this thread should be titled ".48 feels like AP2 is 8-40 times better than AP1", if so... but again, that's pretty useless in reality.

In reality, .48 does not show the increased number of AP connected cameras (8x) or the performance (40x) compared to AP1.

 

[AnxietyRanger]

I do wish that Tesla would provide a feature roadmap for AP. Then a lot of the discussions in this thread would potentially go away. I suspect they will not, as there is undoubtedly technical and commercial risk in doing so.

For example, the vehicle “classification” and adjacent lane display - maybe it was a design dead end? Maybe its never coming back? Maybe its presence gave confidence where such confidence was not warranted - after all nobody can quantify what if anything AP was doing in relation to that information.

I think there is plenty of room for UI improvement with AP, however I am also of the opinion that we are not currently seeing the full picture - there are plenty of questions about what UI will be present around the more automated portions of enhanced AP. I have to believe that Tesla have at least preliminary design concepts for all anticipated functionality - would love to see some of that!

Cheers
Brendon

Nice ideas.

I would welcome even basic transparency about the current status, why deadlines were missed etc., without any forward-looking statements.

Have some candid engineer write that on the blog once a quarter. Would have made the EAP/FSD journey much more transparent and IMO reasonable.

 

[Joe F]

I do wish that Tesla would provide a feature roadmap for AP. Then a lot of the discussions in this thread would potentially go away. I suspect they will not, as there is undoubtedly technical and commercial risk in doing so.

For example, the vehicle “classification” and adjacent lane display - maybe it was a design dead end? Maybe its never coming back? Maybe its presence gave confidence where such confidence was not warranted - after all nobody can quantify what if anything AP was doing in relation to that information.

I agree, I'd love to see a roadmap.

WRT adjacent lane and even general vehicle classification, I feel the Tesla minimalist design, at least to me, is more appealing. The less clutter the better. The only thing that would matter to me might be better blind spot display on the IC, but then again I use my mirrors for that.

Ideally, blind spot detection would delay a lane change, and when the additional camera's are activated that could further simplify the IC. I want to spend my time taking in the entire picture ahead, and can't reliably do that while staring at the display. Let the car do what it does best. It's not going to do it any better if I stare at the IC.

 

[AnxietyRanger]

Ideally the vehicle ID display would show all the cars around me on adjacent and my lane, front, sides and back. That would be IMO helpful.

 

[Joe F]

I am talking about the current software reality. I have made that clear time and again - and that is the context of this whole thread. In fact, that is the context of the whole "AP1 parity" talk.

Agree to disagree. You still don't get to have your own facts, and make false statements based on same. My last on this with you. My apologies to the OP and any still following this thread.

 

[AnxietyRanger]

Agree to disagree. You still don't get to have your own facts, and make false statements based on same. My last on this with you. My apologies to the OP and any still following this thread.

I'm not trying to have my own facts. I am discussing, as factually as possible, the AP2 software reality of today.

I totally agree the potential the hardware offers in AP2 is massively better than AP1. And one day it will surpass AP1 greatly. I thought that was plain and obvious. I just didn't think it was relevant in a thread about AP1 parity.

 

[Carl]

Come on AR, do the test. You wrote quite some posts on this forum about how bad AP2 is compared to AP1, so you should be able to take the same amount of time to find someone near you who has an AP1 car, to do the test, together, no? I think you owe us that, after all your quite definite opinions on AP2 vs. AP1 without having tested the difference yourself I, for sure, would really look forward to it!

 

[verygreen]

I am talking about the current software reality. I have made that clear time and again - and that is the context of this whole thread. In fact, that is the context of the whole "AP1 parity" talk.
In reality, .48 does not show the increased number of AP connected cameras (8x) or the performance (40x) compared to AP1.

But have you seen the crash snapshots? AP2-snapshots are lightyears ahead of the AP1!

(for those who have not seen, here's what you get out of AP1 in case of a crash: First Look: How Tesla Autopilot Camera Captures Accident Video )

 

[AnxietyRanger]

@verygreen AP2 having better crash logging could actually be a fair point. I would not yet put it past AP1 based on that, but it is a start in all seriousness.

 

[AnxietyRanger]

Come on AR, do the test. You wrote quite some posts on this forum about how bad AP2 is compared to AP1, so you should be able to take the same amount of time to find someone near you who has an AP1 car, to do the test, together, no? I think you owe us that, after all your quite definite opinions on AP2 vs. AP1 without having tested the difference yourself I, for sure, would really look forward to it!

If I can sync schedules with one of the local AP1 owners, sure I will try. Or get a suitable loaner in the future. As always, I will report my experiences as I get them.

One thing though, Tesla's execs have been talking big of an upcoming Q4 update to AP2. Maybe this will all be moot in days once those FSD differentiating features etc. hit?

After all, real feature differences are much more significant and interesting IMO than comparing minute details and differences in lane-keeping ability. That's IMO not very exciting.

Make no mistake (well, barring massive and unlikely company failure) AP2 will exceed AP1 in software ability (it already does in hardware ability, of course). It is a question of when and how, not if. We're just not yet there today.

 

[WannabeOwner]

or when the "55mph if you're a truck" sign

I've got one of those near here. A perfect replica of a speed limit sign (right colours, shape, size, etc) embedded in a big yellow rectangular sign with the caption "In 1 mile". Grrrr ... but I don't think that's Tesla's problem, the signage needs to change to be consistent, consistent, CONSISTENT

In Europe I presume by the time I get home my mailbox would be stuffed full of photo tickets.

Nope, you don't need enough to stuff your letterbox full before they take your license away But it isn't actually that easy to get caught over here. In order to prove that the camera is working properly it takes two photos, one second apart, and there are markings on the roads (so two sources of data to compare). In order for local authority to keep the revenue they must not be hidden, so they are painted bright yellow, and have warning signs on the road nearby (which are only allowed to be within X-miles of an actual camera). Or something "not terribly unreasonable" like that. Thus easy to know where they are. Waze knows where they are. Then there are mobile camera crews, not huge numbers of those, but mostly in known places, and unless you are first on the scene Waze knows about them too (but not with enough warning-range for Laser). So in practice its not a huge problem, unless you want to be doing 100+ MPH - then you only need one ticket to lose your licence ... so not a problem with your letterbox when you get home

I do wish that Tesla would provide a feature roadmap for AP

I agree, but I would like that in all aspects. More detailed documentation about upgrades, rather than me discovering the change when the car does something that it didn't do yesterday. I'm struggling to understand why Tesla does not do this, at least to some greater degree than at present (i.e. not enough to alert competitors)