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

[growler23]

The minimum fuel cost for a starship launch is gonna be about 1 million dollars. The actual full operating costs are higher.
...

What cargo, specifically, is worth spending an extra 200 million dollars-- per ship worth- to get there faster?

Based on my memory of the 1980's, the answer is "Air Jordans".

/s?

 

[Tony73]

Tl;dr: to investors: automotive is a crazy hard business to be in, especially when things go wrong. We talk of 5k vehicles per week. Imagine needing to rework all those cars while resuppling the parts pipeline due to a quality escape.

What are you even arguing against?

I never claimed I was talking about needing a guy to fix a thing. I said:
"You've obviously never been on the wrong end of a automotive manufacturing plant shutdown...
$1.3 million up to $3 million per hour in 2006"

The wrong end meaning the supplier that caused the shutdown (which I could have been more explicit about, but didn't realize it mattered).

The specifics of the article beyond cost of plant shutdown were irrelevant to me, I just needed a non-proprietary number. A shutdown plant costs $X million to the OEM per hour for a shutdown. The penalty to a supplier is a multiple of that. There is your hours = millions situation, that's all (see also Giga Berlin power loss).

The point remains that there is not enough sufficiently urgent and valuable cargo in the world to make it worth anyone’s while to build space launch infrastructure to transport cargo by space.
“Hours = millions” situations is not a business opportunity that scales enough to justify $250 million launchpads.

Air cargo reduces travel time from weeks to less than a day and over 90% of intercontinental cargo still moves by ship.

In ten years or so, after starship is proven out, reliable and human rated, it can make SpaceX billions as a space tourism platform.
Selling experiences is way more valuable than moving parcels. Only one starbase is needed. Finding 20,000 people a year willing to pay $75k for a couple orbits round the earth is relatively easy

 

[scaesare]

because it's going to be 1 million dollars just in fuel, and likely 5-10 million in total costs for ONE shipment this way.

BTW, the available cargo room is comparable to a 747. Yes that flight might take 10-12 hours instead of 1 hour-- but the cost will be massively less- like many millions of dollars less.

What cargo is worth spending many millions of dollars (PER shipment) to get it there in 1 hour instead of 10-12? Other than lifesaving medicine (which you don't generally need to ship 150 tons of at a time) I can't think of a single thing.

Plus the 747 can go anywhere in the world there's a decent sized airport. Starship will only be able to go places there's an actual rocket base (so it can refuel and leave again later).


Starship will be incredibly useful and valuable for things outside our atmosphere. The idea it'll "revolutionize cargo shipping on earth" is just economically and mathematically nonsensical.

Disaster response? Humanitarian crisis? Rapid military response?

 

[Usain]

Chuck's left turn is really not that "edgy": probably one encounter for every 20 to 30 miles urban driving. Chuck himself likely did his left turn hundreds of times alone. For such a not so edge case, Tesla has been collecting data for weeks, according to Chuck's X posts. We've been told that the FSD team is able to pull any video for any cases when needed, which is essential to the perceived advantage of billions of miles driven. The intensive collection of Chuck's left turn seems contradicting to that capability.

Hopefully I'm wrong on this.

I hope you are wrong as well, and I think you are. (But I could be wrong too, of course)

Concerning Tesla's data advantage, think of it this way. To do end-to-end drive-anywhere training with vision-only, it takes millions of video clips. Nobody else has that. Therefore, nobody else is even attempting what Tesla has done with FSD V12.

Nobody.

 

[replicant]

Remember.Remember.Remember.

https://twitter.com/x/status/1768616348282913222

Yet, the stock market is at an all time high. One could "Buy the dip" but the current dip could be a tiny one, before the actual drop Elon was talking about (quoting the expectation of Berkshire Hathaway's management about the economy, which has little to do with Tesla's current stock situation).

 

[Usain]

Chuck's left turn is really not that "edgy": probably one encounter for every 20 to 30 miles urban driving. Chuck himself likely did his left turn hundreds of times alone. For such a not so edge case, Tesla has been collecting data for weeks, according to Chuck's X posts. We've been told that the FSD team is able to pull any video for any cases when needed, which is essential to the perceived advantage of billions of miles driven. The intensive collection of Chuck's left turn seems contradicting to that capability.

Concerning Chuck's UPL, it's not unexpected that Tesla would need additional targeted clips in order to get some things right.

But I have another theory about Chuck's famous turn and it concerns the placement of the B-pillar camera.

As has been thoroughly discussed, humans attempting Chuck's turn can lean forward and get a better view of oncoming traffic. That fact alone could mess up the training data. Taken from the wild, you end up with videos of obstructed-view left turns where a human proceeds safely, but the B-pillar camera doesn't really know it's safe. So the training videos could give the system a false sense of security and the car wouldn't creep enough to see oncoming traffic from the B-pillar.

I suspect that Tesla's human drivers working on Chuck's turn are instructed to perform the maneuver with their head against the headrest in order to mimic what the B-pillar sees. In other words, the drivers are not allowed to lean forward and must instead creep a little more to do the maneuver safely. That way, the system learns that it too must creep a little more to do the unprotected left.

 

[Knightshade]

Disaster response? Humanitarian crisis? Rapid military response?

None of which this would be particularly useful or cost effective for.

A 747 could move the same amount of cargo per trip, for roughly 4-9 million dollars per trip less money.

The flight might be a few hours longer- but can land at any airport in the world with a decent sized runway--- versus Starship which besides costing many millions more per trip and only take off or land at a location there's an entire (still functional) starbase.


The only thing on your list I could see any real-world use of would be the military-- if they developed some rapidly deployable temporary starbase facility they could somehow drop somewhere--- but you'd still need to deliver THAT stuff via cargo plane and deploy the temp starbase (ditto the fuel to refuel the ships at the target site)... It would still be hideously cost ineffective compared to just cargo planes- but when has the military ever cared about that?

There'd be no economic case at all for doing any of that for disaster or humanitarian stuff vs just using cargo planes.

(nor do any of these weird niche cases make much of a dent, economically, in the volume of global shipping--which was the original thing Starship was suggested as impacting- see below on that)

What are you even arguing against?

If you don't know what the argument was about, why are you trying to participate in it without determining that first?


The argument was about if Starship would revolutionize "global cargo/shipping in the next half a decade". I was arguing it won't, and there's no economic case to be made otherwise- then cited the insanely higher costs compared to either sea or air shipping and asked what could possibly justify that cost.

Nothing you posted in reply actually addressed that since, again, the only source you gave was not citing a "slow transport of a fix" problem at all which is the only aspect Starship would be relevant to (and not even then unless both sides of the supply and demand HAPPENED to be really close to an existing Starbase).

There is your hours = millions situation, that's all (see also Giga Berlin power loss).

But "faster than an airplane at 10 million dollars more cost" would not have fixed that. Nobody was waiting on slow aircraft to bring in cargo there.

So you appear to have not read what the actual discussion was about, then made up your own argument where the actual "solution" under discussion doesn't fix the actual problem.

A transition quarter or 2 will not depress the stock permanently.

Nobody said it would?

Is this strawman day or something?

That said- the relatively flat growth isn't a quarter or two, it's looking like a year or two. There are no new products this year, only the start of a ramp of one introduced last year (CT) and which Elon already said won't ramp enough to be material to the financials this year... and data is increasingly showing the S3XY markets are basically at or near peak volume in terms of available buyers.

Next year we'll get the rest of the CT ramp, but that's still planning for only maybe a 10% total unit growth fleet-size-wide, and hopefully Semi will start to ramp.

It's 2026 before you can present much of a case for returning to high unit growth with new product-- it'll be a massive growth spurt though assuming next-gen goes well and on schedule.

(there'll be some nice stuff in energy this year, and more of it next as Shanghai megapack plant finishes and ramps, but vehicles- this isn't a 1-2 quarter lull)

 

[Usain]

Figure has that big back pack. Does not look as ready to go as Optimus.

I think it's far too early to know who has a lead.

I'm betting that in the end, several players will come out with good bots that do useful work.

Tesla's advantage will be in manufacturing.

 

[scaesare]

None of which this would be particularly useful or cost effective for.

A 747 could move the same amount of cargo per trip, for roughly 4-9 million dollars per trip less money.

The flight might be a few hours longer- but can land at any airport in the world with a decent sized runway--- versus Starship which besides costing many millions more per trip and only take off or land at a location there's an entire (still functional) starbase.

Is a $10 million dollar premium worthwhile to get a couple hundred tons of supplies somewhere in a couple hours when there could be hundreds/thousands of lives at stake?

I could see that for some scenarios.

The only thing on your list I could see any real-world use of would be the military-- if they developed some rapidly deployable temporary starbase facility they could somehow drop somewhere--- but you'd still need to deliver THAT stuff via cargo plane and deploy the temp starbase (ditto the fuel to refuel the ships at the target site)... It would still be hideously cost ineffective compared to just cargo planes- but when has the military ever cared about that?

There'd be no economic case at all for doing any of that for disaster or humanitarian stuff vs just using cargo planes.

(nor do any of these weird niche cases make much of a dent, economically, in the volume of global shipping--which was the original thing Starship was suggested as impacting- see below on that)


<snip>

Infrastructure for point-to-point terrestrial cargo would obviously need to be built.

 

[GhostSkater]

Now I understand why Tesla uses the LG packs here in Europe, at around 800 km on my rental Model 3 I can say the average speeds here ignoring paid highways are so low, all at 90 km/h or lower

Even the 72 kWh usable pack is giving 500 to 700 km of range, 130 Wh/km | 208 Wh/mi

Some drives I did were at 95 Wh/km | 152 Wh/mi, which also is 6.6 mi/kWh, mind blowing, my off road electric mountainboard does 14 Wh/km

And this on 19” non aero wheels and 10°C to 18°C weather

 

[Knightshade]

Is a $10 million dollar premium worthwhile to get a couple hundred tons of supplies somewhere in a couple hours when there could be hundreds/thousands of lives at stake?

I could see that for some scenarios.

Such as?

Humans can go 3 days without any water, much longer with no food. I can't think of too many situations where a couple hours difference in delivery would change the outcome for hundreds/thousands of lives.

Especially since you'd ALSO need it to be IN a place that just happened to already have a starbase-- and whatever the disaster is left it fully functional.

And ditto on the supply end of the trip too.

Infrastructure for point-to-point terrestrial cargo would obviously need to be built.

How likely it that in places thatALSO need super-rapid humanitarian aid?

How much of a financially successful business do you image that would produce?


Anyway, further discussion has moved here (thanks @mongo for that)

Starship “EP2P” (Earth Point-to-Point) Discussion

Based on what I have heard Elon and Gywnne say, all BFR EP2P flights will be sub-orbital. Of course. It does not make sense, from a physics perspective, to do anything other than a ballistic trajectory. Anything else will take a lot more energy and will cost much more, without significant benefit.

teslamotorsclub.com

 

[elasalle]

Or, more likely perhaps, set up TE factory plus CKD.

Don't think India will allow TE for car imports .. don't think that is they way they intended. We've had high profile visits/meetings and i think each party knows what they want and will get.

Mix - 1K S/X/CT for Bollywood and Ambanis ... 7K 3/Y for IT nerds in Delhi, Bangalore and Hydrabad ..

But 5 years to setup factory(does not have to be Giga size) for the MQ/2 would work really well for Tesla. Also include SuperChargers in the mix. India could be a good base to make moves into Africa as well ...

Solar seems to have been commoditized so not sure if it's worth it.

 

[Waiting4M3]

...

One thing most of us ignore is that Tesla already has done Complete Knocked Down(CKD) with Model S in Tilburg. The standard solution for market entry in smaller markets is and has been for generations CKD. From Ford Model T to commercial passenger airliners one version of that or another is very common, almost ubiquitous.

...

Interesting point about CKD. I wonder if unboxed process will allow Tesla to ramp Model 2 production quickly on other continents once GF Austin can mass produce unboxed parts, which would be much more cost effective to ship vs the S/X that went to Tilburg, new factories just need to start with a shell and general assembly, and can build out other steps while already producing cars

 

[Stormy24]

Green Day. Wow

 

[mongo]

Green Day. Wow

~7.6 million shares at the closing cross, triple option expiry action?

 

[DarkandStormy]

https://twitter.com/x/status/1768737948965462410

 

[2daMoon]

Green Day. Wow

It's St Paddy's Day weekend.
Duh!

 

[Todd Burch]

Green Day. Wow

They sound great, right? It's something unpredictable, but in the end it's right. I hope you had the time of your life.

 

[cliff harris]

For the love of god this is the Tesla investors forum, not the spacex logistics debate club.

 

[Quesder]

Tesla will have a hard time competing against Xiaomi new vehicle SU7 (available 3/28/2024), which supports 800V fast charging, uses 100 kwh battery, and has a range of 800 kilometers (CLTC). Comparing to that, TESLA models are kind of old.
The Chinese market is crucial to TESLA, and its market share is facing so many challenges.

xiaomiev-com (sorry, cannot find English version)

0-100 km 2.76 seconds
265 km/h max speed
33.3 meters 100-0 km/h brake distance
800V charging 15 min for 500 km, 800 km CLTC range

200K - 400K RMB (Yen)