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

[scaesare]

So it was Panasonic that was straight up lying? At least this puts down the benchmark to measure Q2. 6GWh of cells means 500MWh of Tesla Energy plus 73333 Long Range Model 3's. Substantially less and cell supply isn't the bottleneck anymore.

So, as somebody who was also trying to make sense of this earlier (and got it wrong), in light of Elon's recent tweet I went and looked at what Panasonic said:

“Panasonic established a battery production capacity of 35GWh in Tesla’s Gigafactory 1 by the end of March 2019 in line with growing demand. Watching the demand situation, Panasonic will study additional investments over 35GWh in collaboration with Tesla,”

(emphasis mine)

So it doesn't appear there's any lying going on. When taken at face value, what Panasonic, Tesla , and Elon are all saying jive (Occam's Razor, and all that).

A manufacturing line for physical objects (be they batteries or doorknobs), is still a largely mechanical logistics effort. For something that needs to produce millions of items a day at high speed, this isn't just a matter of installing the machines and turning them on.

My father worked in a plant that produced glassware.... 100's of thousands of bottles/jars a day. Getting a manufacturing line up and running was an immense tuning effort, and peak production took some time to reach the capacity limits of the machinery. Things like:

- Conveyor system alignment issues causing product to not flow correctly
- Speed matching between conveyors and machinery
- Annealing temps of ovens inconsistent
- Temperature/humidity variations changing the cooling rate of the glassware
- Transfer equipment alignment/tuning
- Mold imperfections/refinement
- Travel angles of the conveyors
- Guide and barrier adjustment
- Sensor/gate adjustment
- Air pressure variances
- etc...

And this process is somewhat iterative as speeds increase. What might work at 100 items a minute might not work at 125 items/min... all it takes is one conveyance mechanism to be a fraction of a percent slower than the one preceding it, and product begins bunching up or you have to throttle the entire line.

And there are dependencies: increasing the speeds requires greater air-pressure to operate the molds faster but which increases closing force which induces stress in the glass which requires a different annealing temp profile.

Now while certainly batteries and bottles are vastly different, and certainly the GF has highly automated lines, it still doesn't mean there isn't a big effort to tune this to reach theoretical capacity. It just may be that some of it is done in a terminal session rather than with shims and a wrench. But there are still also going to be a lot of physical adjustments too.

I can imagine some of the stuff that you might have to tune on a battery line:

- Conveyor rates
- Transfer alignments
- Casing clamping force
- Material insertion alignment
- Drying oven temp profile adjustment
- Cap attachment alignment/tuning
- Electrolyte filling/tuning
- Lead attachment
- etc...

There are a myriad of logistics involved... and a whole slew of interdependencies we probably can't imagine. So what works for 24 GWh/yr rate may not work for 28. And that may not work for 32... and that for 35. And this is all in addition to the materials issues that @sparcs mentioned.

So, if this is all going to go on for the next period of time, and they have no significant indicator there's a showstopper, it would seem to make sense that now that they've reached there initial goal (35GWh of capacity installed), they stop and evaluate next steps while the production ramps up.

 

[CorneliusXX]

Selling at current prices, while including individually-owned vehicles in the Tesla network (with Tesla sharing profits with the owner), is one option

If they sell at current prices, even if the vehicle is shared through the TN, they will still be giving up a substantial amount of economic value.

 

[CyberDutchie]

Uber is valued above $100B not because it's making money, it loses billions each year. Investors buy the stock because they think Uber will make money in the future.

If Tesla can demonstrate they are leading in FSD development, and to show they are on track to do ride services on Tesla Network using autonomous EVs, this could change Tesla's valuation dramatically.

There are more than 2 billion vehicles in the world, total yearly milage more than 10 trillion miles. If 10% of that switch to ride sharing services, at $1 per mile, that's one trillion dollars revenue. If it turns out Tesla is leading with autonomous EV, and becomes the only one that can offer the service at low cost (driverless EV), they could end up taking the whole market, with $1T revenue per year, the cost for that revenue is about $0.15T (the depreciation of 20 million EVs plus electricity).

I know there are lots of IFs in the above assumptions, you can see whoever gets true FSD out first is potentially a big deal. By the way, I agree with you that traders should be aware of the "buy the rumor, sell the news" phenomenon. I have seen it on many stocks.

A couple weeks ago Continental estimated the market for self-driving software to grow to $1.2T by 2030. If Tesla would license their FSD software, there’s another major recurring revenue stream.

Article here:

Autonomous cars to quadruple vehicle software revenues to $1.2 trln by 2030 Continental

 

[EVNow]

Beyond the regulatory hurdles to FSD, I agree with other posters here that there are substantial psychological hurdles. Yes, an FSD car may be safer than an average driver, but almost everyone considers themselves above average!

I bought AP & FSD during the sale - but yet to try out even auto-steering. I'm now trying to use TACC. Once I feel confident about TACC, I'll try others.

ps : I don't want bodily injury apart from financial

 

[Joe F]

And then there's a huge number of crashes caused by repeat offenders, people who should probably never have been given a license in the first place, and should have had their license revoked after the first reckless crash. This is not well documented but I've started to dig up the evidence of it.

On a two day drive going home after the SpaceX launch up I95, through bouts of extremely heavy rain and traffic, we had hours of delays due to accidents. Probably counted about 10 on both sides of the highway.

The most serious involved an SUV that lost the fight with a semi that required the jaws of life. Sad to see the stupidity of some drivers in heavy rain. On another accident scene which had emergency vehicles just arriving, running with Autopilot, my MS saw traffic ahead breaking, even before I could see brake lights and I started stopping, pumping the brake pedal to warm those flying up behind me. Watching for possibly being rear ended, I saw a car swerving around someone behind me, then almost hitting a first responder who was crossing the road before the car went skidding off the side of the road. Scary stuff out there.

If anyone ever doubts a Tesla with EAP does not drive more safely than many average drivers, try driving through Virginia on I95 and around Washington DC on any day of the week and get back to us with your thoughts then.

BTW, the launch was awesome!

 

[scaesare]

I agree, that the advancement of the neural net could potentially be reaching exponential increases at this point. While Elon is always optimistic, I think the frequency of his optimistic statements have been increasing (plus it's also really crazy how far Navigate on Autopilot has come since its first introduction in such a short time period, compared to other HW2+ features).

I think Google Deepmind's gaming demonstrations are a perfect example of the exponential nature of neural nets. Deepmind went from learning how to beat Go world champions to beating Starcraft 2 world champions in a manner of years. (For anyone unfamiliar with Starcraft 2, it is thousands of orders of magnitude more complex than Go yet also requires a good deal of "human intuition"). From a human perspective, it uses hundreds to thousands of years worth of game playing to reach this point. Tesla is the only self-driving developer that is capable of emulating hundreds to thousands of years' worth of human driving in the real world--the entire HW2.5+ fleet has increased dramatically in recent months and is increasing at a brisk pace today. If their NN development has indeed entered this "exponential" phase, we're in for something pretty earth-shattering.

TSLA-wise, though, I don't think anyone will notice or care about it until it's too late. They could announce a business worth $300 billion this month and investors will probably not care.

Indeed.

I was trying to figure out where in this thread to interject this, and your post is a great one.

Adding to it, everyone should read about how AlphaZero became the most powerful chess playing entity on the planet in just 4 hours. What's incredible about this is that the researchers didn't do this by teaching their AI how to play chess. They did it by developing a general-purpose AI advanced enough, and then simply teaching it the rules of chess, and just letting it play.

No loading initial strategies, no showing it 1000's of games to model itself after. They let it learn by playing (itself), and 4 hours later it crushed every comer, including the previous best chess programs/AIs that had been painstakingly tuned for countless hours. It took 8 hours to master Go (considered significantly harder than chess) and obliterate everyone and everything there as well.

Interestingly, in both cases the AI developed some winning strategies that were counter-intuitive to it's creators and experts, but quite effective.

So that's not to say that we can load AlphaCarZero in to Tesla's FSD computer and let it loose and call it a day. But it does underscore @eloder 's point that if " advancement of the neural net could potentially be reaching exponential increases at this point" and if so "we're in for something pretty earth-shattering" when that happens.

It may not be on the 22nd, but at some point I suspect we are going to turn a corner and the issue of "good enough" will rapidly vanish as a question. It won't be perfect, but as Elon says, a 10X reduction in auto fatalities will be hard to argue against.

 

[bdy0627]

On a two day drive going home after the SpaceX launch up I95, through bouts of extremely heavy rain and traffic, we had hours of delays due to accidents. Probably counted about 10 on both sides of the highway.

The most serious involved an SUV that lost the fight with a semi that required the jaws of life. Sad to see the stupidity of some drivers in heavy rain. On another accident scene which had emergency vehicles just arriving, running with Autopilot, my MS saw traffic ahead breaking, even before I could see brake lights and I started stopping, pumping the brake pedal to warm those flying up behind me. Watching for possibly being rear ended, I saw a car swerving around someone behind me, then almost hitting a first responder who was crossing the road before the car went skidding off the side of the road. Scary stuff out there.

If anyone ever doubts a Tesla with EAP does not drive more safely than many average drivers, try driving through Virginia on I95 and around Washington DC on any day of the week and get back to us with your thoughts then.

BTW, the launch was awesome!

Yeah, we just had a huge pileup of cars in February, due to blizzard conditions. It involved over 100 cars with 71 injured. Only problem is, autopilot likely would not have helped since visibility was so poor. It came on pretty suddenly. I don't think we have an answer for that. Fortunately, there was only 1 death (plus a cat - sorry @Krugerrand).

 

[Fact Checking]

Only problem is, autopilot likely would not have helped since visibility was so poor. It came on pretty suddenly. I don't think we have an answer for that.

Radar sees through snow and rain very well.

 

[bdy0627]

Radar sees through snow and rain very well.

My autopilot nearly always shuts off when some snow collects on the front of the car.

 

[EVMeister]

My autopilot nearly always shuts off when some snow collects on the front of the car.

Does that generally happen whilst driving? Or is this stuff you could've cleared before setting off? Not being snarky, just curious.

 

[VValleyEV]

OT: New and old technology:
Big Boy No. 4014 fired up for the first time; May 12 excursion announced | Trains Magazine

Saw this monster in 2013 at the Pomona fairgrounds while they were working on restoring it. Over 600 tons, 6300 horsepower, built in 1941.

As the article says, it will be heading to 150th anniversary celebration of the transcontinental railroad.

For comparison, the Space-X Falcon Heavy rocket, first commercial launch yesterday, weighs 1420 tons, generates 45 million horsepower, and can lift a payload equivalent to a fully loaded 747.

Here’s a photo I took of the locomotive, with a worker alongside for scale.
View attachment 396364

And here of course is what 45 million horsepower looks like:
View attachment 396365

As several of you have pointed out, I flubbed the stats on payload size.

The payload of Falcon Heavy is greater than the weight of a fully loaded 737, not 747.

From SpaceX website:
Falcon Heavy
“With the ability to lift into orbit nearly 64 metric tons (141,000 lb)---a mass greater than a 737 jetliner loaded with passengers, crew, luggage and fuel--”

And then further on, they mention comparing the thrust of Falcon Heavy to 747:
“Its first stage is composed of three Falcon 9 nine-engine cores whose 27 Merlin engines together generate more than 5 million pounds of thrust at liftoff, equal to approximately eighteen 747 aircraft”

Mea maxima culpa.

 

[schonelucht]

Credits don't affect margins or ASP.

Yes. They do.

 

[EVNow]

In AI, data is king. Nobody has more than Tesla.

I remember attending a Big Data conference some 5 years back. Almost all speakers were talking about how Tesla is going to get to FSD first because they have so much data. Apparently not the Silicon Valley CW anymore ...

 

[schonelucht]

There are a myriad of logistics involved... and a whole slew of interdependencies we probably can't imagine. So what works for 24 GWh/yr rate may not work for 28. And that may not work for 32... and that for 35. And this is all in addition to the materials issues that @sparcs mentioned.

I work in industrial automation. I am painfully aware of what you are saying. As long as you are not actually producing 35GWh because some parts of the chain aren't capable of sustaining that volume, you are not capable of producing 35GWh.

 

[EVNow]

Anyone using navigate on autopilot knows highway driving is already now at L3.

There are a lot of those on that subreddit with Tesla. So, they may be saying this from experience … could be the question of number of disengagements. Some may consider even one disengagement not L3.

 

[mongo]

In AI, data is king. Nobody has more than Tesla.

Both on the raw collection and validation (FSD with nags) sides of things.
400,000 AP capable cars now, soon to be 500k. If 40% buy FSD, that is 200k validation vehicles. With a 30 mile average commute, that is 30 million miles of data per work week. 1.5 billion miles per year.

 

[Cosmacelf]

Hello All, thank-you for the incredible insights over the years....first post here....

Is it possible that the Tesla Network could be spin-off from Tesla....thus the "Investor" wording of the April 22nd event ?

Some sort of corporate structuring to make visible the latent potential of the Tesla Network ?

Yes this is possible. It wouldn't be a Tesla capital raise, so Elon wouldn't be going back on his word there. A spinoff would allow them to raise money for the spinoff and value it closer to Waymo, Uber, etc. It would be a bargain in comparison.

The only negative is that some analyst's valuations of TSLA are predicated on making money from such a ride sharing system, and thus TSLA value would fall. It isn't true that the market values FSD at $0. However, presumably, it wouldn't be a 100% spinoff, so TSLA would retain some ownership.

If Elon does go this way, it had better be a lot more thought out than the $420 tweet!

 

[neroden]

Interesting comments from @sdmoores on GF1 output suggests there may have been raw material constraints. He is a battery materials expert and one of the most knowledgable sources on the EV transition generally. He had a private tour of GF1 at the start of March with both Tesla and Panasonic sections.
"Tesla Gigafactory is at 24GWh expanding to 35GWh by April. It’s our @benchmarkmin estimate that the max amount of lithium ion battery cells it can produce in 2019 will be 32GWh. 420-450,000 Model 3’s worth."
"It’s been a big challenge for the key battery raw materials of lithium, cobalt, graphite anode and nickel to scale at Tesla’s pace. So this could slow the Gigafactory ramp. 32GWh Max in 2019 and maybe a low of 28GWh. Our estimates from @benchmarkmin"
"This will be 400,000 to 457,000 Model 3’s worth of lithium ion battery cells in 2019. The 2170 cells that Tesla invented with Panasonic."

Oh-kay. So they aren't willing to say *which* materials they think there's a shortage of. Nickel's easy to get, there's lots of nickel. Lithium? Cobalt? Graphite?

 

[neroden]

Yes, but note the gist of my comment: my point is that FSD is evaluated based on actual real world probabilities and economics, not as @Zhelko Dimic's comment suggested:



In that sense IMO Elon was correct in suggesting that the threshold that matters is practical protection: if FSD is 3x safer, (or maybe 5x safer), objectively, based on real world data, including any tail risk, then it will be preferred to humans in all economic decisions.

Not some "10x-100x safer than humans" threshold.

Ask yourself why we don't have driverless trains everywhere. The tech was developed in the 1970s and perfected in the 1990s. It's at least 10x safer than human drivers and substantially more efficient economically.

I await your answer.

 

[Prime_Number]

Yes this is possible. It wouldn't be a Tesla capital raise, so Elon wouldn't be going back on his word there. A spinoff would allow them to raise money for the spinoff and value it closer to Waymo, Uber, etc. It would be a bargain in comparison.

The only negative is that some analyst's valuations of TSLA are predicated on making money from such a ride sharing system, and thus TSLA value would fall. It isn't true that the market values FSD at $0. However, presumably, it wouldn't be a 100% spinoff, so TSLA would retain some ownership.

If Elon does go this way, it had better be a lot more thought out than the $420 tweet!

You are ABSOLUTELY right that it must be well thought out.

A few points of inspiration can however redefine the challenge. The fundamental question is not how to price Tesla FSD technology vs other technology providers, it is solving the challenge of getting a person from point A to point B. In this context, L3 ability + Boring Company tunnels can beat the performance of any L5 without the Boring Company.

The moat that the Supercharger Network is vs other electric car makers could find its parallel in The Tesla Network with Boring Company tunnels.