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

[spacerfc]

The next thing someone is going to tell me is that I’m old-fashioned for believing a person’s word is his bond; that words mean anything.

Well, I suppose that is one way to go through life. Not any life that I’d like to live, though.

May I extrapolate your views to include that you likewise would treat with such equanimity if the BoD also were to say “Things changed: we’re not going to honor that Incentive Package.”?

It should be evaluated on a case by case basis, if the company is in distress and BoD thinks the only way to save the company is cancelling the incentive package, then absolutely it should be cancelled.

Similarly if Elon thinks an asteroid is going to wiping out the country and he needs to sell enough shares to raise money to deflect the asteroid, then yes he should break his words.

 

[elasalle]

Will ignoring a moderator break the board? Asking for a friend.

Tried it once, you will not get access to the whole thread

 

[larmor]

The optimus video is very concerning. It can mimic / track movements and copy-- yay great, so can any robot. But then it can also review 1000s of videos of t shirt folding as well. And then use an AI engine to parse out variations and then do it on its own. In of itself not a problem, but lets say someone sees the optimus as a weapon, and trains it to shoot, and then besides the video training of above, merges current algo that select targets. How long before a government or group represented by a government buys enough shares to get a controlling interest, and then wa-la, optimus becomes the terminator.

 

[Drax7]

Elon has frequently said that we should take him at face value. At face
value his statement is a veil threat .

He has achieved beyond his dreams, and been compensated beyond his dreams, so
there is nothing to argue about that.

He should be in a position to control Tesla, we
the market can agree on that.

Buying X at the expense of diminishing his control over Tesla
was purely his doing. Huge error.

So far FSD is incomplete, and of little commercial value.
So future compensation should be based on FSD achieving
At least level 4 and beyond. Giving away (25-18) % of the
company in exchange for level 5 FSD would seem incentive
enough.

 

[SmokyPeat]

The optimus video is very concerning. It can mimic / track movements and copy-- yay great, so can any robot. But then it can also review 1000s of videos of t shirt folding as well. And then use an AI engine to parse out variations and then do it on its own. In of itself not a problem, but lets say someone sees the optimus as a weapon, and trains it to shoot, and then besides the video training of above, merges current algo that select targets. How long before a government or group represented by a government buys enough shares to get a controlling interest, and then wa-la, optimus becomes the terminator.

No problem, we'll just have Robocop take out the Terminator bots. Did this potential outcome really just occur to you?

 

[AudubonB]

But I don't think Elon ever promised/said he wouldn't sell any TSLA shares. I think he said something more along the lines of "I'll be the last one out." Which to me means he won't ever sell most or all of his shares.

But in reality, did you really expect him to never sell any TSLA shares? Even when they are his only compensation while he obviously has living expenses.

Paragraph 1: This argument has been brought up before. As with any such statement, the appropriate way to parse it is to take it to its logical extreme: By your logic, all he would have to do would be to retain one single share to remain faithful to your belief. I also do remember Mr Musk using that "last one out" line in at least one other occasion. Taken all such comments in hand, my considered understanding was, as I wrote above, that such a sale would be far in the future - on the order of a decade or more - as it was earmarked solely for the colonization of Mars, and therefore it did, indeed, mean any shares.

Paragraph 2, Q1: Yes. As he was addressing a sobersided group of a few dozen investors, ranging in portfolio size from my holdings to the very largest, he was speaking to a group who take very seriously what a company spokesman and most especially, a CEO who is the single largest shareholder, says.

Paragraph 2, Q2: This argument also has been brought up multiple times. It is 100 percent not true. A responsible large investor optimistic of his holdings' future will NOT sell his shares; rather, he will make use of the line of credit such a holding allows.

 

[2daMoon]

The optimus video is very concerning. It can mimic / track movements and copy-- yay great, so can any robot. But then it can also review 1000s of videos of t shirt folding as well. And then use an AI engine to parse out variations and then do it on its own. In of itself not a problem, but lets say someone sees the optimus as a weapon, and trains it to shoot, and then besides the video training of above, merges current algo that select targets. How long before a government or group represented by a government buys enough shares to get a controlling interest, and then wa-la, optimus becomes the terminator.

All that would be needed is a satellite based internet system, a network in the sky, so we could hack into those pesky killer bots and re-code them, right?



Right?

 

[AudubonB]

It should be evaluated on a case by case basis, if the company is in distress and BoD thinks the only way to save the company is cancelling the incentive package, then absolutely it should be cancelled.

Similarly if Elon thinks an asteroid is going to wiping out the country and he needs to sell enough shares to raise money to deflect the asteroid, then yes he should break his words.

You never have been faced with an investor lawsuit against a BoD, have you? I have - thank goodness, only vicariously. Very few things uglier.

 

[mongo]

Elon has frequently said that we should take him at face value. At face
value his statement is a veil threat .

Aren't "at face value" and "a veil[ed] threat" the exact opposite of each other?

 

[Artful Dodger]

The optimus video is very concerning. It can mimic / track movements and copy-- yay great, so can any robot. But then it can also review 1000s of videos of t shirt folding as well. And then use an AI engine to parse out variations and then do it on its own. In of itself not a problem, but lets say someone sees the optimus as a weapon, and trains it to shoot, and then besides the video training of above, merges current algo that select targets. How long before a government or group represented by a government buys enough shares to get a controlling interest, and then wa-la, optimus becomes the terminator.

Oh no, you want 25% of the Company too?! /S

 

[ItsNotAboutTheMoney]

The optimus video is very concerning. It can mimic / track movements and copy-- yay great, so can any robot. But then it can also review 1000s of videos of t shirt folding as well. And then use an AI engine to parse out variations and then do it on its own. In of itself not a problem, but lets say someone sees the optimus as a weapon, and trains it to shoot, and then besides the video training of above, merges current algo that select targets. How long before a government or group represented by a government buys enough shares to get a controlling interest, and then wa-la, optimus becomes the terminator.

Randall Munroe was way ahead of you.

Robot Future

xkcd.com

 

[2daMoon]

Looking at Teslarati I noticed an article on SIXT and Stellantis implying a deal that will replace Tesla rentals.

In the article's photo I noticed this...

Has Stellantis created a mechanized Pushmi-Pullyu?

Will it be electric?

What could be the purpose of such a vehicle?

 

[AudubonB]

For couples on their way to divorce court.

 

[Drax7]

Aren't "at face value" and "a veil[ed] threat" the exact opposite of each other?

It’s a threat, no equivocating

 

[Rarity]

Add to this how the "remote control" aspect for Optimus is merely part of the learning phase, where the robot is shown the human activity via touch, movement, and vision, then, can interpret how to do that activity on it's own.

There would be little gained by having to have an operator for each robot all the time. As pointed out, there are already robots that do that and applications for the Tesla Bot isn't targeting use for highly specialized tasks, such as surgery.

Boring, repetitive, dangerous, are the job qualities the army of Optimi will target.

After the first "wow," I started thinking about how clumsy the teleoperator must be. The teleoperator is no Marie Kondo, probably just some rando engineer with AR gloves. Solving the clothes folding task requires hiring a few Marie Kondos who are experts and dexterous at folding clothes and who are willing to part with their knowledge on reasonable terms.

The whole issue with LLMs is that once we start exploring their output in our area of expertise, we can see that at this point it is a mess of nonsense. Similar to a general journalist's output. Domain experts generally do not expose their knowledge to the outside world. Rather, we can only see the end product. Like with all learning, we have to find good teachers.

 

[Krugerrand]

The next thing someone is going to tell me is that I’m old-fashioned for believing a person’s word is his bond; that words mean anything.

Well, I suppose that is one way to go through life. Not any life that I’d like to live, though.

May I extrapolate your views to include that you likewise would treat with such equanimity if the BoD also were to say “Things changed: we’re not going to honor that Incentive Package.”?

I appreciate a man of his words. I have a handshake deal with my GC. Tell me who else does that in this day and age?

However, I’m not so immovable nor ignorant to not realize sometimes things change beyond one’s control. I don’t expect a man to keep to his word if it was detrimental or a self-sacrifice.

There is a place of balance. You know this, but sure be obtuse. It’s a pandemic of obtuseness lately around here.

 

[UncaNed]

Whomever they call partners... Doubt they´d go as far as calling Tesla a partner for also building EVs


Good that they have been thinking about CO2 efficiency, at least diesel electric for the following ships. Does anyone know how much of a difference that makes for long distance shipping? Would guess under those circumstances it is mostly constant power anyway except for starting to move, so not much of an advantage?

Anyway, it definitely proves they have BIG plans for exporting cars... Let´s see how this turns out.

BYD hat nun ein eigenes Schiff für den Autotransport

Der chinesische Autohersteller BYD will in Sachen Logistik offenbar unabhängiger werden und plant eine Flotte von acht eigenen Schiffen.

insideevs-de.translate.goog

Looks like plenty of room for a solar panel canopy on top. I bet that would provide significant cost effective power.

 

[SOULPEDL]

It is learning.

I'm still trying to figure out if Optimus is learning, mimicking, or copying - each have different levels of complexity and purpose. Spare me, I have limited knowledge in AI but extensive experience in human learning + basic common sense. I've concluded that it's possible it's doing all 3 combined for reasons given below.

Add to this how the "remote control" aspect for Optimus is merely part of the learning phase, where the robot is shown the human activity via touch, movement, and vision, then, can interpret how to do that activity on it's own.

Agree, with a bit of a twist.

Think about "seeing vs doing" and the "hands-on experience" gained from doing a task, even if just mimicking or copying (as in pure R/C).

What if the data is faster to acquire by inputting directing from Optimus Sensor data while it is simply copying the teacher's motions R/C with a bodysuit? It's like a transfer function to get the task in Optimus data form, while combined with actual touch data. Hey, isn't this like shadow mode on FSD where the driver inputs the data directly into the controls? But how would you do that with Optimus?

You can't 100% drive Optimus... unless you use a body suit, gloves etc. Although, I could also argue that it doesn't take much data analysis to translate pure vision into bone kinematics for Optimus, but perhaps an extra step is removed to optimize the iterative process - a step closer to real-time learning.

I have a funny feeling more will be revealed here. No way Optimus is just copying, it's the data capture, analysis, and speed of learning we have yet to hear about - in less than 2 weeks. And it may have to do with the compensation package - Elon knows what's coming more clearly than most.

Anyone think a DITM leap is worth the risk? Will 2 yrs out take us past the recession?

 

[Zero CO2]

https://twitter.com/x/status/1747275580381487305

 

[Todd Burch]

After the first "wow," I started thinking about how clumsy the teleoperator must be. The teleoperator is no Marie Kondo, probably just some rando engineer with AR gloves. Solving the clothes folding task requires hiring a few Marie Kondos who are experts and dexterous at folding clothes and who are willing to part with their knowledge on reasonable terms.

The whole issue with LLMs is that once we start exploring their output in our area of expertise, we can see that at this point it is a mess of nonsense. Similar to a general journalist's output. Domain experts generally do not expose their knowledge to the outside world. Rather, we can only see the end product. Like with all learning, we have to find good teachers.

If you want perfectly folded clothes, yeah you probably want to train using data from the winner of the ICF (International Clothes-Folding) Championship held in Saskatoon, Saskatchewan, every February. (Just kidding, don't look it up).

But all you really need is a decently-folded shirt. Most people can fold a shirt decently. (Maybe there should be an upcharge for retail-store style folded clothes?