Gotta admit I think your burn rate comment is kind of FUDish. Happy to discuss why I disagree with your analysis but probably this thread is the wrong place.
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Your analysis of Tesla's business and the danger (or not) that its huge negative cash flow poses is very different from mine. I'll post my thoughts later tonight.
Voted, but there wasn’t a good answer for me.
I use it often, so I picked the first choice. No, I’m not super confident in it at all times and it isn’t smoother than a fresh jar of skippy. Keep my hand gently on the wheel. Makes my time on the roads, esp freeways much better.
I use it often, so I picked the first choice. No, I’m not super confident in it at all times and it isn’t smoother than a fresh jar of skippy. Keep my hand gently on the wheel. Makes my time on the roads, esp freeways much better.
NerdUno
Member
@calisnow: Hate to say it but our technology experience and enthusiasm is quite similar. I've worked in the tech arena for many, many decades. So how do we reach such different conclusions when it comes to Elon's methodology? I think the answer lies in your being willing to give him a pass on the misrepresentations believing that they were just an expression of his dream. I, on the other hand, went to law school. If Elon had actually expressed your interpretation rather than representing in writing that the EAP technology would be available in December, 2016, I would still be all in on the car. Instead, we're 10 days away from December, 2017, and "the dream" still looks to be another year or two away even in the best case scenario. Having been a lawyer for many, many decades, I can tell you that exaggerations of this magnitude get folks in legal trouble, and Elon probably is not going to escape very substantial financial penalties for his actions. And that's too bad because I, too, view him as one of America's best hopes for moving electric vehicles and autonomous driving forward. He already has.
Posting in the monthly AP2 disappointment thread
Seriously though, Tesla's AP2 is screwed IMO. I don't think they'll ever reach any level of confidence for the EAP features they described, let alone FSD. I had high high hopes for FSD after Tesla put out those FSD YouTube videos. But now I realize it was a deception.
Also, what is with Elon saying they've figured out the neural network problem? They still can't deal with slopes, and I can't see them EVER solving the slope problem with the AP2 hardware:
I think what may be happening with the slopes is: The road slope causes the lanes of the road to curve on the camera image. But this is exactly how a turn in the road looks like. Therefore the AP system is confused and thinks that the slope is a curve. Humans know better because of knowledge of the context (e.g. angles of cars) and because we have two eyes for 3d vision. Maybe once Tesla turns on all front facing cameras, they can fix this problem.
As for the overall topic of the thread, I do think we can exceed AP1 performance within a reasonable timeframe. However, to get to FSD would be exponentially harder. And we might not see it with this generation of hardware or in the years that I will own my current Tesla. Remember that this is something that no one in the world has really perfected so far, especially with the hardware Tesla is using (no LIDAR). There should be no guarantee that more development time would solve this problem. That's how hard this is.
I am not nearly as optimistic but also not as pessimistic as some of the people on this forum. I work in the tech industry, with experience in similar types of problems. The difficulty curve with this problem, like most problems in tech, is not linear. It gets exponentially harder as you try to do better. And at some point you reach the limits of the hardware, in which case further improvement is not possible. You must recognize this possibility.
A simple thought experiment would be to think about the gap between AP1 and AP2 at this point. And then think about how much time you think it'd take to reach parity (for example, 6 months? 3?) Now think about further improving on this parity performance by about that much. Where would that be? (e.g. automatically change lanes, exit highways, local roads with red lights?) That marginal improvement would be much harder than the margin it took to reach parity with AP1. So if you think it takes 6 months to reach parity with AP2, then further improvement by that much would take a lot longer than 6 months. You can choose your own numbers here, but my intuition says that it's not going to be quick at all.
powertoold
Active Member
I’ve followed Tesla and AP2 closely for a long time. In the last year, they’ve improved AP2 at a linear snail pace. I don’t care about that though. What’s disappointing is Elon’s promises. Even during the last call, he said they solved the neural net problem, which has been very difficult. What the heck is that supposed to mean? Nothing’s been solved in .46. We’re all supposed to expect some amazing future in 3 months that never comes.
I think everyone hopes Tesla is holding back some miracle cure for EAP and FSD, but why would they do that when they can send out the update now and at least auto steer would be better.
I also follow a lot of advancements in ML, and you can’t just fix a NN by tweaking certain things. You have to rethink the whole dataset and algorithm, and sometimes you get gains in one area and losses in others. It’s not like programming features in a traditional sense.
I agree with most of your post, except for the "linear snail pace" part.
It's hard for us as users to quantify quality improvements in autosteer (I'm sure Telsa has better data and tools to evaluate this), but what we can objectively quantify are actual features that are released. And from my notes (which should be mostly accurate), I think the last time we had a new autopilot feature release was 17.17.4 which released AEB, side collision enhancements, and auto high beams. That update came out in May, more than half a year ago! Therefore I do think AP progress has slowed considerably since the rapid pace in the first couple months after AP2 release.
(It's possible, maybe even likely, that Tesla has been doing lots of improvements in shadow mode and have been waiting for a good time or quality threshold to release them at once. But what we can be sure about is that no major features have been publicly turned on in the past half year.)
While true, all this happened months before Tesla actually announced AP2. They knew MobilEye would not be on board when they announced EAP and FSD and the timelines.
Yet they announced December 2016 EAP anyway. And again FSD differentiating features in April-July, 2017. It is November 2017 and we don't even have AP1 parity yet, let alone EAP, let alone any FSD differentiating features.
Silence from Tesla. No updates, no mea culpas or transparency worth mentioning, no new revised timelines, just pretty much the same texts still sold today. And Autopilot marketed like this - read and compare to actual fact:
Autopilot
Do you have nothing better to do in life than come on here to write aggravating posts? I guess that was a rhetoric question because I know the answer guaging from your meaningless responses. But to alleviate your anxiety regarding whether I own a Tesla or not, the answer is I do. Its a 75D Model S with air suspension, PuP, AP 2.0. But hey if it makes you feel better and superior assuming that i'm a non-owner just coming on these forums to spew hate on Elon, sure go ahead.
Last edited:
Again, by the volume of drivel you have written it reinforces my belief that you have nothing better to do in life and that you may be dealing with issues that need to be addressed by a professional. You accuse me of "hyperbole" and yet write stuff like the above which is laughable. "Give you chills"..."clearly you don't know this..." "clearly you don't know that". Get off your high horse dude. I get it that you own a Tesla and it might be the only good/luxury car that you've owned, the world around you has seen and experienced a ton more. I see you're a Tesla fan and so are most of us who are posting on this forum. What is fascinating is to see this blind faith and obsession by immature posters like you who buy into one narrative and then will never ever challenge it. This very type of mindset of harboring extreme, obsessive, unwavering views or lack of openness to seeing all sides is what has gotten our country in the mess that it currently is.
You write nonsense such as "blinding speeds"..."crazy fast". Really not sure what world you live in because when a company tells me that X feature costs $4,000 and I buy it, I expect it to run smoothly right out the door as soon as the purchase is made. Simple as that. Or be truthful enough to tell a customer that "your Tesla comes with AP 2.0 hardware, once it is fully functioning we will open up the feature for sale and you will receive it as over the air update". Why talk about something as "available" and then not have it fully functioning even a year down the road? Having something half baked a year later is not "crazy fast" by any means, it is borderline fraudulent.
Anyway, I do not expect any meaningful response out of you but wanted to clarify my position a bit more. Lastly, I did vote but not the bottom option. Please go somewhere else and write your novels on some other forum.
I feel the same way. I look forward to updates for minor things such as entry/exit profile or perhaps improving their streaming radio but i have given up any hopes on AP 2.0. I use it sparingly and have lowered my expectations to compare it to a TACC on my other car at best.
Amazing that you had the time and energy to read his epic thread. Regardless this dude seems to have had the best, picture postcard experience with AP 2.0 and is in full confidence of his capabilities. Good for him but sadly that is not the case with most of us. And "false promise on timelines" is the best way to put how Tesla overpromises and under-delivers.
THIS. Right here is my biggest irritation with Elon/Tesla. They are still selling these EAP and FSD features yet not once have they owned up to the fact that they've dropped the ball on roll out and execution to paying AP 2.0 customers. At this point I have given up believing AP 2.0 is getting any better. Just give me the auto-wipers and better functioning auto high beam and that will be fine by me. I'll just consider it a few thousand dollars wasted on a con feature and next time I'll be 5 times more careful when purchasing my next Tesla. Even the auto-high beam feature is so useless that at times it will work and at others it'll irritate the person on the road on the other side because it refuses to lower the lights (I live in the country so long stretches of 2 lane highways with no median and no street lighting).
It is what it is. I will continue to root for Tesla success. I believe in the company and feel proud of driving a zero emissions Tesla each day. I LOVE this car. But I have started to be very skeptical of Elon's hyperbole nonsense over each keynote and on twitter.
Yes some of us early AP 2.0 purchasers from Q4 2016 feel a bit burnt. Also, please note that the AP 2.0 crowd still doesn't have things like auto-wipers and while we have the auto high beams its performance is pretty inconsistent. But yeah if I had spent money on AP 2.0 in Sep 2017 vs. last year, perhaps my perspective on the roll out and execution would have been different too.
But there's some truth to his post calling out the "cheerleading". I am a huge cheerleader for Tesla and I LOVE my car and this company but why is fandom turning into this obsession? Why can't we call out our idols and the things, brands we look up to for when they make a mistake? Serious question.
For one we are on an enthusiast forum. And everyone is entitled to their opinion. A lot of us possibly over defend Tesla because we love the company, the mission, and the product.
I however see plenty of criticism and objective analysis here.
And often times “cheerleading” is brought only when people disagree. It is not constructive and kind of ridiculous to even say on an enthusiast forum.
The point of my post was merely to bring up Tesla financials and sensationalist articles on cash burn has nothing to do with autopilot and belongs in other threads.
Hmmmmm, Sounds like a job for buttershrimp.
I've started a youtube channel dedicated to testing Autopilot 2.0 in difficult scenarios, all whilst listening to deep soulful melodies by contemporary artists. If you don't like that, please take a look but perhaps with mute on, if you really don't like the deep soulful rhythms and rhyming. Tips and tricks are welcome.
As for the trend of this thread in the past few pages. Here are my thoughts. Elon makes rational folks wince from time to time and it is probably most evident to lawyers. On the other hand, many folks are willing to give him a pass because the guy is likable, works extremely hard, and has truly flipped the auto industry on it's head. He is a force of nature, and luckily, a seemingly likable one....
I wouldn't worry too much about Tesla until Elon is no longer in the CEO position, or they start making cars that require an advertising budget. The moment Tesla becomes run by large corporate boards, people like Bob Lutz, or worse, a super careful non-visionary focused on ROI, then I think we can all expect the same sort of stagnation in the auto industry that we've seen prior to Tesla. Just my opinion...
I've started a youtube channel dedicated to testing Autopilot 2.0 in difficult scenarios, all whilst listening to deep soulful melodies by contemporary artists. If you don't like that, please take a look but perhaps with mute on, if you really don't like the deep soulful rhythms and rhyming. Tips and tricks are welcome.
As for the trend of this thread in the past few pages. Here are my thoughts. Elon makes rational folks wince from time to time and it is probably most evident to lawyers. On the other hand, many folks are willing to give him a pass because the guy is likable, works extremely hard, and has truly flipped the auto industry on it's head. He is a force of nature, and luckily, a seemingly likable one....
I wouldn't worry too much about Tesla until Elon is no longer in the CEO position, or they start making cars that require an advertising budget. The moment Tesla becomes run by large corporate boards, people like Bob Lutz, or worse, a super careful non-visionary focused on ROI, then I think we can all expect the same sort of stagnation in the auto industry that we've seen prior to Tesla. Just my opinion...
I guess that really is the difference.
The people who worry (or not) about Tesla are identifying with the company and its mission.
The people who worry (or not) about their own purchase experience are identifying with, well, themselves.
These people are bound to see very differently being sold 9k worth of self-driving tech and then not delivered as scheduled.
Similar threads
