Tesla, TSLA & the Investment World: the Perpetual Investors' Roundtable

[Ogre]

Just wait until HW69!!!

Seriously though, everyone keeps saying better. While HW4 is smaller, quicker, and cheaper than HW3, if Tesla is to be believed, it shouldn't affect FSD, EAP or any of the other programs that utilize it, otherwise they would figure out a way to upgrade it.

I doubt the people clamoring for it could tell the difference in a blind A/B test. But time will tell.

I’m sure a human driver wouldn’t be able to tell the difference.

If you take Musk’s comments at face value, FSD with HW3 should be 10x better than a human driver when finished. But perhaps HW4 is 11x and HW5 is 12x… would anyone but an expert be able to tell them apart?

 

[Nocturnal]

Would you mind typing the following?

"$1000+ by late 2024?"

 

[15Peter20]

I was within half a glass of wine of selling all my TSLA at close yesterday because the stock always tanks after earnings regardless of how good or bad (Obvs I was planning to re-buy at the dip).

A combination of relative sobriety and some sage words on this thread held me back from pressing the button.

HODL!

 

[FSDtester#1]

And the chasing begins... cut p/t's on the way down and raise on the way up...

Tesla Surges On Earnings; Analysts Hike Price Targets Despite Revenue Miss

Tesla stock has soared from its Jan. 6 bear market low.

www.investors.com

 

[Rotte]

I was within half a glass of wine of selling all my TSLA at close yesterday because the stock always tanks after earnings regardless of how good or bad (Obvs I was planning to re-buy at the dip).

A combination of relative sobriety and some sage words on this thread held me back from pressing the button.

HODL!

I was about to do the same thing. When was the last time the stock actually went up the day after earnings? Then, I remembered my mantra, "the best investor is a dead investor", and did nothing.

 

[MC3OZ]

Weird way of saying, "you will have a safer robotaxi with HW4 or HW5 than your HW3 vehicle."

The main requirement is HW3 significantly safer than you doing the driving.

 

[FSDtester#1]

This shows you Elon's state of mind right now...

From Yahoo Fiance:

Analysts were mixed on the results, but investors reacted positively to Musk's statement during the earnings call that the firm is seeing its "strongest orders year to date" in the company's history. It expects to deliver about 1.8 million vehicles this year.

It seems the firm's positive results had put the firm's flamboyant CEO in a good mood as he had a joke at the expense of a legal opponent.

"Changed my name to Mr. Tweet, now Twitter won't let me change it back," Musk tweeted.

He made the faux pas in reference to lawyer Nicholas Porritt, who is representing shareholders suing him over 2018 tweets about taking Tesla private. He accidentally called Musk "Mr. Tweet" during the trial on Monday.

Absolutely Musk at his finest !

 

[Ogre]

For the record. At current rate, we are 10 days from the ATH.

160*1.1*1.1*1.1*1.1*1.1*1.1*1.1*1.1*1.1*1.1 = ~415 which is roughly the ATH.

 

[Prunesquallor]

Toyoda-san does not seem to be answering the phone today.

https://twitter.com/x/status/1617801825406832641

 

[MC3OZ]

More like HW3 will remain a Level 2 ADAS, and who knows how many more iterations will be required to produce a generalized Level 4-5 robotaxi. Nobody has any idea what that looks like and won't know until it exists in the real world, which it doesn't.

Also not sure if training on new hardware iterations will be applicable to hardware on previous gens. I'm pretty sure any change in camera location/geometry/number, any change in sensor fusion, will be applicable only to that specific configuration.

They might need 2 seperate lots of training, but they can train from the same video.

If this is true, I see it as an advantage, becuase HW4 will get somethings right that HW3 gets wrong and that data can be used to improve HW3 training.

The accuracy of just about anything can be improved by having and independent data set to validate against, especially if that inpependent data set is more accurate.

There are only 2 scenarios where HW3 can't do the job.

1. Sensor suite is inadequate- or doesn't perform in some conditions.
2. Hardware is too slow even after heavy optimisation.

It is likely that Tesla can do a lot to squeeze more performance out of the HW3 hardware, mostly via a more optimized set of neural networks.

Why are we even worrying about this? Elon and the Tesla staff are more highly qualified to worry about it and are being paid to worry.

 

[ZenRockGarden]

And the chasing begins... cut p/t's on the way down and raise on the way up...

Tesla Surges On Earnings; Analysts Hike Price Targets Despite Revenue Miss

Tesla stock has soared from its Jan. 6 bear market low.

www.investors.com

I could totally be a Wall Street "expert". Set a price, and then move it up or down, belatedly following the movement of the stock itself.

In fact I could automate it.

 

[CorneliusXX]

I find the sneering attitude on here to concerns about parking sensors shortsighted and extremely patronizing. I am sure a lot of people here think that they are gods gift to driving because they don't find a need for parking sensors. Well done. Now drive a model Y through the side streets in Rome and London for a day, and tell me what your bodywork repair bill amounts to.

Like it or not, countries exist where the roads are not a hundred feet wide like they seem to be in Los Angeles. Also, believe it or not, EVERY other car company on earth making premium vehicles includes parking sensors *because people use them*.
I have a brand new model Y performance, delivered without parking sensors. It is absolutely definitely by a huge margin, the worst thing about the car.
And the replacement tech in a few months? Yup sure, just like FSD will be in a few months?

Its incredibly silly for 'investors' in a company to be willfully blind to a genuine criticism of the product. Its also absolutely TERRIBLE customer service by tesla to remove parking sensors from $90,000 cars with no fixed timeline for feature equivalence. It totally soured my excitement about my new car. For the sake of a few cheap sensors, they harm the image of a $90k car. Bad decision, bad execution.

Agree. Coming from the wide open roads of Australia to the goat tracks that make up most of London I've had to significantly increase the precision of my driving. The parking sensors are a lifesaver.

 

[DarkandStormy]

I have a bit of bad news for you then. (Sorry @wipster another bad news post from Ogre)

HW6 is going to be safer than HW4. You might want to wait a bit longer.

PS: Shouldn’t the bar be “Is it safer than a human”?

Musk's bar is 1,000% safer than a human.

More like HW3 will remain a Level 2 ADAS, and who knows how many more iterations will be required to produce a generalized Level 4-5 robotaxi. Nobody has any idea what that looks like and won't know until it exists in the real world, which it doesn't.

Odd, I was told that from 2016 on every Tesla had the hardware needed for level 4-5 robotaxis / FSD. And for a couple iterations at least, Tesla was willing to make those hardware changes for customer who bought the package. But now the line is drawn at HW3?

 

[CorneliusXX]

View attachment 900001

Project Highland theory

Maybe it's not the Model 3 structure, but HOW the structure is built. If you look at the above photo from the Earnings Report, it looks very different from the ones we've seen in the past. It looks like everything is being welded in one place, vs. running down a line with welds added at each step/location. Thoughts?

I was caught by this picture too. Elon used to go on about density of manufacturing all the time during the mid 2010s - this looks like a significant step in that direction

 

[KBF]

I don't like this definition of sandbagging.

IMO, this would go along like 'state a number that will make your audience not unhappy while there is a very good possibility to actually perform great'

I had no idea it might involve "unfair advantage" piece. YMMV

EDIT:
This is what I found - "to conceal or misrepresent one's true position, potential, or intent especially in order to gain an advantage over"
this definition is MISSING the word unfair. I like this expanation better.

source

I don't like slang in general. Too much connotation that varies from region to region. I'm certainly not tied to the google definition I used - I should have used ChatGPT and asked it to use an American definition or something. I'm okay with @Ogre's "gun-shy" I suppose, but I'm not sure that matches their mood/intentions, either. I shouldn't have opened that can of worms, though.

No it is not an assumption. It is literally what they guided. Numbers on PDF

I'm open to being wrong about Tesla's underlying assumptions - but I went back to the PDF and couldn't find any quarterly projections for 2023, just annual. I downloaded the shareholder deck on Tesla's Investor Relations page. I could only see annual guidance (holding to the 50% CAGR, "around 1.8M cars for the year") on p. 12 - do they have 2023 quarterly estimates on a different page or different document?

 

[H Mak]

I was within half a glass of wine of selling all my TSLA at close yesterday because the stock always tanks after earnings regardless of how good or bad (Obvs I was planning to re-buy at the dip).

A combination of relative sobriety and some sage words on this thread held me back from pressing the button.

HODL!

Very interesting. I can guess your thinking. Stock is up 40% from its January low, margins will be hammered by price cuts, stock will plunge. Take profits and buy back in at a lower price.

A few years ago I sold a portion of my holdings into earnings after a strong rising price. The stock continued its rise after earnings and I bought back the shares at a higher price. Lesson was thoroughly learned. It's still tempting though, I must admit. But I'll never do it again.

 

[H Mak]

I usually try not to counter most posts.. but ...


Actually..."Tesla stock has soared 55% since a Jan. 6 low of 101.81, coming up to their 50-day and 10-week lines."

He was talking about yesterday when it opened at $141 and closed at $144

 

[betstarship]

So, for those that invested at $400...~20% has been recouped so far.

 

[FSDtester#1]

He was talking about yesterday when it opened at $141 and closed at $144

Deleted, thanks... still good news today up 55% from January lows..

 

[UkNorthampton]

No-one mentioned the new stock split.. /s

.

It's corrected itself now but was messed up for a while.