Human vision is far superior to current neural nets.
As for 3d maps by cameras, yes it looked nice, but it has to always look nice. That isn't assured.
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Tesla, TSLA & the Investment World: the Perpetual Investors' Roundtable
- Thread starter AudubonB
- Start date
Human vision is far superior to current neural nets.
As for 3d maps by cameras, yes it looked nice, but it has to always look nice. That isn't assured.
One of guys who attended the Rivian party said that the CEO carefully avoided mentioning climate change in order not to offend any climate denying customers attending. So I am not so sure. It could be just big auto green washing.
bdy0627
Active Member
Chicago inventory is the lowest since I have been following it 6 weeks ago. There is just 1 pre-raven S listed. No Xs. No 3s.
I went and looked up on the Waybackmachine the subscriber count on the TeslaMotors subreddit. Basically it took an elbow up at the launch of the model 3 and has been growing with extremely low variance +500/day. That's a big number I think, nearly as much as model 3 production, but it's strange to me how consistent it is and also that it is a straight line rather than curved. Not exactly sure what to make of it but obviously more impressive than any other brand can offer.
I like these both, logical. Good stuff man.
How do delivery costs shake out in the financials, Tesla charges roughly 1200 for a delivery fee correct? If it goes over is that cost put into COGS or is it already apart of COGS?
EVNow
Well-Known Member
The battery pack and motors are the costliest items in the car. So, if the cells cost $125/kWh and the motor $2500 - that would be $9k + battery packaging - lets say $12k. The rest of the car can't cost $30k ! If it does - then, there are gross inefficiencies that they can fix.
I agree with this and it is perplexing. Mazda sells its entire 3 starting at 22k$.
I wonder if Tesla simply is unaware that this needs explaining. If they can give a strong believable reason why the reported numbers are high by 4k$+ they would basically eradicate this downswing overnight imo.
Carl Raymond
Active Member
They are not even analogous. You might compare vision to cameras. You might compare recognition to neural nets. The machines will be superior and lidar will be dropped by all, of that I have no doubt. The only question is when.
EVNow
Well-Known Member
But, do you actually know how good it is ? Can you quantify it ? Not a rhetorical question.
BTW, when you say "current NN" - lets be clear you are not talking about what Tesla has with HW3. We actually don't know how good their stuff is.
The 3D maps, again, have to be as good or better than human vision. For eg., we know in certain areas, our 3D vision fails badly (like figuring out how far an approaching train is). Then, there are a lot of people with one eye, bad vision etc who drive.
Hummm.... I wonder how many of the abandoned in place robots in the 3 line can be used on the Y line. New end effectors, of course.
EVNow
Well-Known Member
Yes, also, mostly softball questions. Nothing on FSD, cell manufacturing situation or any topic of current interest.
bdy0627
Active Member
I thought that choice of words very odd. What other CEO uses the word cash burn about his own company? I don't know why, but it struck me as odd and clearly intentional. He certainly knew the email would leak. It's easy to deduce what would follow. Why would he intentionally do this? I really don't know, but there is that damn wink just a week or two later...
EVNow
Well-Known Member
Depreciation for 3 was calculated by @Doggydogworld roughly as 500 (variable) + 1500 (fixed, allocated @ 90k/quarter). It is not that big an amount.
If some robots are not being used - they should actually be taken out of 3 depreciation, though. And as a poster suggested above, could be used for Y (or even Shanghai).
Me thinks, Musk isn't that bright in choosing different words and phrases depending on the audience. It is a gift and curse.
It is a gift because that is what endears him with his millions of followers, as he comes out pretty honest. In fact I haven't seen someone at that level so brutally honest.
Its a curse, because at his position as a CEO of a public company and given that there very very powerful vested interests rooting for his failure with MSM very eager to twist his words, he has to be very careful in what he says. This is not only about avoiding words like
'burning cash', 'weeks away from bankruptcy', 'if we make base Model 3 now we will die' , 'LIDAR is a fools errand' etc..
but also about giving aggressive timelines, production & profit estimates.
Last edited:
EVNow
Well-Known Member
Occam's razor says he wrote that to put some fear into the heart of business managers (who need to negotiate contracts and cut costs). Well, at least he hoped it would. I think he is quite poor at communicating with non-SV VC/tech crowd.
Tesla releasing a wireless charger for phones in the console gap at 125$ a pop. This is the right idea. Humorously Fred spends a good amount of time advocating the cheaper non OEM version on amazon. Apparently he won’t be bought easily. He fights for the people!
I wonder if Tesla will sell an upgrade that allows people other than Zen masters the ability to close that console lid on the first try.
I wonder if Tesla will sell an upgrade that allows people other than Zen masters the ability to close that console lid on the first try.
Last edited:
Dan Detweiler
Active Member
Thank you!Is Wall Street Clueless On Tesla [TSLA]? | CleanTechnica
The most material tweet of 2018 was not the supposedly infamous “funding secured” tweet. It was the April 13, 2018, tweet that Tesla would be profitable in Q3 and Q4, that Tesla would not need to raise any money in 2018.
But it was not in line with the analysts’ views, so it was ignored, ridiculed, and considered a poor stock-pumping attempt. What all the analysts missed was that Tesla CEO Elon Musk’s Twitter account is an official Tesla publication channel, giving that tweet the status of an official press release to the financial community.
The Economist
✔@TheEconomist
· Apr 12, 2018
Tesla will need to raise $2.5bn to $3bn this year, according to Jefferies, a bank https://econ.st/2GKub8z
Tesla is heading for a cash crunch
The road ahead for Elon Musk’s car company is looking more perilous
economist.com
Elon Musk
✔@elonmusk
The Economist used to be boring, but smart with a wicked dry wit. Now it’s just boring (sigh). Tesla will be profitable & cash flow+ in Q3 & Q4, so obv no need to raise money.
18.5K
10:11 PM - Apr 12, 2018
Twitter Ads info and privacy
2,549 people are talking about this
At the conference call about the Q1 results on May 2nd, 2018, a number of analysts kept asking about the money Tesla needed to raise — in their view. It was implied repeatedly that Tesla absolutely had to raise more money, despite Musk repeatedly saying they didn’t. This cumulated in an analyst insisting that Elon disclose how much money Tesla was going to raise, despite those repeated previous denials for the need to raise any money at all. It was very embarrassing, an analyst essentially telling the CEO to stop lying and tell everyone what the analyst wanted to hear.
A little lesson for analysts: CEOs and CFOs try not to lie to the financial press in official statements, press releases, quarterly financial reports, and press conferences. They can go to jail for doing that. And if you, great journalist or analyst, have a strong suspicion backed up by some evidence that one of those two is lying, call the SEC. The SEC loves to know about it, as long as it is about Tesla.
Worth noting here is that there was no slip in wording. Elon repeated many times that Tesla wouldn’t raise money in the rest of 2018, and would show profits in Q3 and Q4.
A year later, it is not much better. The current confusion on material info is about demand this time, instead of production capacity and profits. While demand for fully electric vehicles outside the USA is rising at a speed that carmakers can’t match, the American analysts show again that they don’t know the difference between sales and deliveries, or how the car market functions outside the USA. Or they pretend not to know.
They tell baseless stories about Tesla, embarrassing themselves by showing their lack of knowledge and ruining their credibility among those who can put 2 + 2 together.
Last week, a Tesla store in Vancouver, BC, reported selling 800 cars in a week. They will likely be delivered in Q3. Those boneheaded analysts will most likely report a surge in demand in Q3 and an “alarming drop” in Q4, while all that is likely to happening is a rush in Q2 that turns latent demand into actual sales because of a local incentive at the risk of running out. That is going to create what looks like a wave in demand, or roller coaster with a hard fall at some point, instead of steady-state demand. Is demand really going to fall off a cliff? Or is there simply going to be an unbalanced expression of consistent demand?
It is not really difficult when you have your facts straight.
So, here is a short tutorial for analysts:
There are two types of sales.
And here are some examples of issues affecting the two types of sales:
- The commercial sale — the customer deciding to buy and signing the sales contract. This is important for judging demand, and for a salesperson’s bonuses.
- The delivery and final payment. This is important for the accounting department and for allocating profit to the right period.
In the USA, type 1 sales and type 2 sales mostly occur on the same day. Knowing the number of type 2 sales is a good indicator of demand and the success of an ad campaign. In the rest of the world, Europe especially, there are usually a few weeks to a few months between the type 1 sale and the type 2 sale. That is also true for domestically produced vehicles. Most dealers do not have unsold new cars on their premises. They are considered a waste of money and a sign of bad salesmanship.
- Factories that are being upgraded have temporarily less production, less production implies fewer deliveries. (type 2 sales)
- Ships (not) arriving in overseas export markets dictate the logistics flow and spikes/drops in deliveries. (type 2 sales)
- Incentives being changed can cause spikes and drops in customer decisions to (not) buy a certain car. (type 1 sales)
- Incentives that are tied to date of delivery can spur before-deadline deliveries. (type 2 sales)
- Ad campaigns and media attention can increase the number of customers deciding to buy. (type 1 sales)
- Seasonality because of license plate series. (type 2 sales)
- Spikes after bonuses are being paid. (type 1 sales)
For various reasons, only type 2 sales are communicated to the financial press and the public at large. Type 1 sales are often shared with trade associations and government statistical bureaus. They are not shared with the competition and outsiders. Car journalists and financial analysts in Europe don’t pay attention to the numbers of a single month. Only when a pattern emerges over several months is it time for conclusions. Even when there is a likely cause and effect, as with the arrest of Ghosn in Japan, conclusions are very carefully made (in this case, mostly after confirmation by Nissan).
Tesla’s Vehicle Delivery Expectations (FactSet) Are Being Artificially Inflated. Image by Michael Grinshpun
People Repeatedly Claim Tesla Demand Is Falling When Data Shows Tesla’s Demand Is Rising. Image by Michael Grinshpun
In data sciences, you learn that pure data is meaningless. The process of giving meaning to data transforms it into information. To give meaning to data, one has to know the rules that apply to the data. When not knowing the rules, and therefore applying a wrong set of rules, the result is gibberish. Not recognizing the gibberish is as big an error as applying the wrong rules in the first place.
The information of Tesla having a demand problem is actual gibberish, but not recognized as such.
Recently, we have had two other examples of Wall Street not turning data into factual, correct information. Both were “leaked” internal emails from Elon to Tesla personnel. The first contained hyperbole about the burn rate the newly acquired capital could be exposed to. The second was about receiving many orders for cars.
Many articles are written about the question of how those highly paid Wall Street analysts lack basic reading comprehension to interpret the emails so wrongly. The best is “Some Thoughts On Elon Musk’s Emails To Employees — & The Media + Wall Street’s Response.” There is no reason for me to repeat what is in this article — just read it if you haven’t. Elon’s response to the article was, “Yup.”
The SEC is very critical about releasing material information that might not be 100% accurate. Elon Musk uses sometimes a big brush to paint the future he expects. Elon is not “scripted.” Other CEOs can give verbatim, years later, the same generic answer to a question they haven’t received in years. Elon will give two different answers when the same question is asked in the same conversation. He is responding in a human way to the question in the moment and going off of a memorized script checked obsessively by lawyers.
The SEC is right in asking for more accuracy in the communications, but the SEC shares the same handicap as much of Wall Street — it is apparently not able to recognize material information.
Last year’s April 13 tweet was not recognized for what it was, a very important public release about the Model 3 production problems being solved, and implications for the productivity and profitability of Tesla in the rest of the year.
The first letter this month was thought to contain the prediction that Tesla would run out of money in 10 months. It did no such thing, and if thinking about it for a few seconds, the numbers would not add up. When you have a cash position of $4B and you burn $2B in 10 months, you are not out of money — you still have $2B in cash. Also, Elon was not making a prediction or forecast — he was simply illustrating that $2B didn’t mean they had no cares in the world. There was no material information in that letter. It was a normal internal communication to employees like every company makes regularly — be frugal, don’t waste money.
The second letter did contain important information — about order intake, production volume, and production goals — exactly the kind of information Wall Street is always asking about and not getting, because they routinely interpret it incorrectly, just as happened this time.
And last but not least, there’s the complete misunderstanding of the difference between orders, logistics, deliveries, and demand. Wall Street analysts we are supposed to trust produce wild fairytales many conspiracy websites would be jealous of, joining a cult of “demand truthers” that until recently was only really present on #TSLAQ.
Let’s make a note about exceptions in the financial industry. Bank of America did state that the current Tesla stock price was the result of short seller manipulations, not the fundamentals. The expected messiness of the first quarter was, for Oppenheimer and ARK Invest, not a reason to revise their analyses and their long-term projections. Kudos to those three, which used their brains and facts instead of reproducing the reverberating sound in the echo chamber of Wall Street analysts, pundits, and $TSLAQ.
The effect of the other’s failures are $billions of lost value by real investors, the people the SEC was created to protect. And then there’s possibly $billions in profits for illegally manipulative, destructive short sellers, one of the market-disturbing groups the SEC was created to prosecute. The success of these short sellers is made possible by the Wall Street crowd of analysts, pundits, and mainstream financial journalists aiding and abetting them. It is their shortsightedness and echo chamber culture that mislead investors who are reallyinterested in how the company is doing.
To recapitulate, here are four examples of Wall Street completely missing the meaning of the data it receives:
What does it matter how accurate Tesla’s communications are when Wall Street is so incapable of understanding what is happening, what is important, and what is not important.
- FUD-induced rumors about falling demand combined with misunderstood European sales numbers.
- Ignoring the April 13 tweet about profitability and the following shareholder letter and verbal explanations.
- Seeing alarming financial forecasts where there is only motivational hyperbole in a normal letter to employees, a letter in which there isn’t any forecast whatsoever.
- Concluding lack of demand when there is actually a message of high order intake and possible record deliveries.
Okay, accurate information is important for us, specialized electric vehicle and renewable energy news outlets, compulsive Tesla followers, and Tesla investors. We need the information. Thus, I have a request.
At the end of the quarter, two numbers are published, production numbers and deliveries. Production tells us how well the factories are performing. Deliveries are about financial results, but a bit meaningless without the average sales price and gross margin.
I have a request for Tesla: Can you replace the delivery number with the net order intake number. Then the first quarterly report covers both production and sales. That would then be followed 3–4 weeks later by the normal report about the financial results that are achieved thanks to those performances.
Wall Street won’t understand the numbers, or will pretend not to, but hey, that is nothing new.
Probably the single best, most insightful summary of the current state of affairs as it relates to Tesla and TSLA that I have seen.
Dan
It's empirical. No one knows including Elon. A specific kind of Brute force algorithm worked for chess but the algorithm was insufficient for Go or poker or etc. State of the art neural nets are not capable of implementing human level abstract intelligence, which comes into play when contemplating how various objects are about to interact with each other in a novel scenario.it is unknown whether current narrow AI logic is sufficient. The worst mistake is to think the hardware chip is a big differentiation. It just makes the system behave poorly in corner cases faster.
Krugerrand
Meow
I would add a butler to my private island and two yachts. Butlers need to make money too.
Similar threads
- Locked
- Locked
