2017 Investor Roundtable:General Discussion

[Ulmo]

Interesting new developments in LIDAR tech:

Luminar reveals sensors that could make self-driving cars safer than human

I feel like we had an all-fronts attack on innovation in the 1970s-1990s (possibly earlier as well), and that projects such as the above would have come out a generation (20 to 30 years) earlier if it were not for that. I feel like that was done in order to let the power brokers get their pawns in a row. Now, that they've gone through a few cycles of sweeping up all of the new world order stuff, they are ready, and that technologies like this will come out at a higher pace, still measured compared to natural innovation, but not choked like it was in decades past. This fits with the theory that car companies even let Tesla get away with making a successful EV at all.

This means a few things:

• Tesla will have access to comparatively amazing technology over the next three decades
• So will everyone else
• There will be fights for who gets this stuff

That article has an important quote in it:
"the CEO said car companies have already offered to buy up, at any price, all the LiDARs that Luminar plans to produce in its first run, starting this year."

That means that it could still be quashed, but these days, quashing may not be the only motivation of competitive car companies: they might actually want to compete with product improvement, not just quashing all innovation.

Of course, I've been highly critical of Tesla's recent innovation on Model S & X, considering they have many long standing user interface, usability, and integration issues that are easy software fixes; maybe Tesla is throttling its innovation to a degree much larger than most of us understand, or maybe they're just not doing as much as they could be.

A lot of possibilities in there. NVIDIA still is working on car driving AI, and not keeping exclusive with Tesla.

 

[tander]

Elon Tweet-storm summation:

• "Tesla Semi truck unveil set for September. Team has done an amazing job. Seriously next level."
• "Next gen Roadster will be convertible"
• "Pickup truck unveil in 18 to 24 months"
• "July [timeframe for Model 3 final unveil?]"

I've been looking at the pickup truck as a final coup de grace to the ICEs and competitors that aren't headed toward electric. Ford sells around a million F series a year with average price of about 50k IIRC. But the semi truck thing is surprising to me, anybody have a sense of how big it could be?

 

[WarpedOne]

anybody have a sense of how big it could be?

Only limited by battery supply.

 

[tander]

Interesting new developments in LIDAR tech:

Luminar reveals sensors that could make self-driving cars safer than human

Isn't what they are already putting into Tesla's expected to do this?

 

[Ulmo]

Remember these:
Tesla's battery storage announcement involves Oncor project
Whatever happened to Oncor's big energy storage plans? ?

Latest news:

"NextEra Energy Inc. (NEE) can't acquire one of the largest U.S. electricity transmission businesses, Texas regulators said,
...
The regulator's decision likely disrupts yet again the debt-restructuring plans of Energy Future Holdings Corp., Oncor's 80% shareholder, which has spent nearly three years under chapter 11 protection. Energy Future's stake in Oncor, a cash-producing, regulated business, was supposed to be the source of cash to fund a massive bankruptcy-exit package.
...
Under Energy Future's chapter 11 plan, which a bankruptcy judge has approved, NextEra pledged to pay $4.4 billion in cash and stock for the Oncor stake as well as to pay off $5.4 billion in debt. NextEra had also committed to pay $2.4 billion to acquire the 20% Oncor stake held by investors other than Energy Future."

Lots of projects fall through. Until they're built and paid for, the revenue can't be really counted. Looking at it from an environmental perspective, until they're plugged into the grid and delivering clean energy and shutting down fossil fuel use, they aren't helping the environment.

 

[schonelucht]

Q3 earnings was fantastic. IMO the muted reaction was due to the extremely pessimistic sentiment surrounding the SCTY acquisition which was still in process.

Also Q2 was bad and the stock didn't really react to that either.

 

[Ulmo]

Isn't what they are already putting into Tesla's expected to do this?

No. Either Elon has been lying about what they will enhance in their vehicles to gain competitive advantage, or they have been against a stream of improvements in their AI systems:

• Things they mostly ended up doing but said they wouldn't:
  • California development (at first, they only had Israeli development, then went local).
  • More CPU power (at first, they used lower capability chips from Israel, but now going NVIDIA with CPU upgrade option in AP2 HW cars).
  • Number of sensors (at first, they said they wouldn't be increasing number of sensors, but now they have).
  • Big Data (at first, they claimed big data was unnecessary, but now they fully embrace it).
  • Intelligence rather than procedural (at first they swore by Israeli's procedural methods, but lately have been converting to more thought-based systems).
  • Hire more amazing experts in the area (Tesla has pushed off a lot of capable AI engineers, but has also managed to amass a few more -- I can't see into their company far enough to know the kind of level of status of this that I'd like, but from their continual fan-club defenders, they're defending a rather poor position (claiming things like "it's hard and they're doing it as fast as humanly possible; they have a moat", etc. garbage & lies), but that may be fake as well).
• Things they said they wouldn't do, but who knows:
  • LIDAR (still claiming not to want to use LIDAR; too expensive)
  • More sensors (still claiming a central bitch-seat position is a sufficient sensor position, even though it's never been done in history of driving in civilized traffic)

In other words, Tesla is a moving target on AI that they will use, tend to either change their mind a lot or lie about it or both, and don't want to reveal a lot of information, probably because they don't want others to know what they're doing, haven't figured it out fully themselves, and/or don't want to be caught in lies, which makes sense. Innovation and development needs flexibility, and winning in a competitive environment requires not revealing all their successes before they come to market.

But, so far, Tesla has said they will not use LIDAR, is the current status, subject to revision.

 

[schonelucht]

Final unveil of Model 3 in July is the most significant tweet of the bunch to me : everything still on track.

 

[30seconds]

Only limited by battery supply.

Cue SeekingAlfa article on how Tesla is doomed because they don't have a big enough battery factory

 

[AcesDealt]

No. Either Elon has been lying about what they will enhance in their vehicles to gain competitive advantage, or they have been against a stream of improvements in their AI systems:

• Things they mostly ended up doing but said they wouldn't:
  • California development (at first, they only had Israeli development, then went local).
  • More CPU power (at first, they used lower capability chips from Israel, but now going NVIDIA with CPU upgrade option in AP2 HW cars).
  • Number of sensors (at first, they said they wouldn't be increasing number of sensors, but now they have).
  • Big Data (at first, they claimed big data was unnecessary, but now they fully embrace it).
  • Intelligence rather than procedural (at first they swore by Israeli's procedural methods, but lately have been converting to more thought-based systems).
  • Hire more amazing experts in the area (Tesla has pushed off a lot of capable AI engineers, but has also managed to amass a few more -- I can't see into their company far enough to know the kind of level of status of this that I'd like, but from their continual fan-club defenders, they're defending a rather poor position (claiming things like "it's hard and they're doing it as fast as humanly possible; they have a moat", etc. garbage & lies), but that may be fake as well).
• Things they said they wouldn't do, but who knows:
  • LIDAR (still claiming not to want to use LIDAR; too expensive)
  • More sensors (still claiming a central bitch-seat position is a sufficient sensor position, even though it's never been done in history of driving in civilized traffic)

In other words, Tesla is a moving target on AI that they will use, tend to either change their mind a lot or lie about it or both, and don't want to reveal a lot of information, probably because they don't want others to know what they're doing, haven't figured it out fully themselves, and/or don't want to be caught in lies, which makes sense. Innovation and development needs flexibility.

But, so far, Tesla has said they will not use LIDAR, is the current status, subject to revision.

I get the impression it's a bit of a mess over there. Their lead dog actually left and they sued him. Anyone who talks about level 4+ being close is bluffing or deluded. The game is to show a demo, get a lot of money, and ignore the fact that going from demo to reality is the hard part.

I am optimistic DESPITE all this. They just have to get there first. I don't care that they are full of *sugar* about their current status, so long as they are ahead of everyone else in probability of winning. This is a 2-3x bagger with EVs, but it's a 10-20x bagger with success at this.

Can I get a reference on denying big data? I suspect it's more nuanced than your statement. Er, I guess that's part of believing in procedural logic. If that's all, then I don't need a ref.

On edit, I didn't know *sugar* was automatically swapped in.

 

[Ulmo]

I've been looking at the pickup truck as a final coup de grace to the ICEs and competitors that aren't headed toward electric. Ford sells around a million F series a year with average price of about 50k IIRC. But the semi truck thing is surprising to me, anybody have a sense of how big it could be?

It's odd Tesla hasn't gone into the pickup (and work van) market. There is a huge sector of society that treats Tesla drivers like crap because they're too antagonistic against each other (much in the way most good drivers disdain most Prius slow pokes, but with even more urgency, because Tesla went all the way, and their owners are all that much more of a target). A pickup would erase most of that antagonism. Right now, owning a Tesla is something tantamount to a declaration of war against all the manly sport-watching beer-drinking hicks from the center of the country and anybody even halfway like them (almost like a Trump vs Tesla branding, even though that doesn't exist in real life). I am a Tesla Model S owner, and the sneers I get from all of the pro-strength pro-business people scare me; I have to hide my car constantly. It's so bad I miss the days of the sneers I used to get for being a Mercedes snob, because at least that was an established position, and it just slotted me into lifestyles where everyone treated me great, even if they were dismayed they had to, with the occasional car chase from some invader immigrant who wanted me dead just because Mercedes (and that 300hp 7-gear all the way up to 130MPH with GREAT SBC brakes helped a lot); with Tesla, about the best I get is that the various Social Justice Warriors give me a slightly wider berth, slightly less animosity, put me a few months later in the schedule of their planned genocide against us --- hardly a win, in my book. I'd rather Tesla be a brand that people adore, like, respect, or accept, rather than get so damned vicious about. To me, a pickup (and work van) would do that, because those on "the other side" would suddenly have access to an electric vehicle that works "on their side". 18 months feels like an eternity. 2-3 years to market is 2020-2021.

Meanwhile, in terms of pollution, heavy cargo trucks are a big target. It's almost as if Tesla wants to solve environmental issues at the expense of solving their business issues. At least they're doing both, so that's the good news, but it seems like they're dragging their feet about it all.

 

[drinkerofkoolaid]

I'm still very nervous about what will happen on April 29th, and am hopeful Trump won't do anything too crazy before then in an attempt to create a stand off with US debt holders.

 

[Reciprocity]

It's odd Tesla hasn't gone into the pickup (and work van) market.

The plan has been pretty clear. Very expensive short run roadster, then high end $100k sedan to fund the next step. Some how they slipped the X in as it wasnt part of the original plan. Then a mass market car.

One reason I can see why a truck wouldnt have come before the 3 is the battery requirement would be much larger then a small car. People want to be able to tow a boat or other large stuff with a truck. Batter is going to have to be 130kwh or larger to be able to tow a large trailer 400 miles.

 

[AcesDealt]

It's odd Tesla hasn't gone into the pickup (and work van) market. There is a huge sector of society that treats Tesla drivers like crap because they're too antagonistic against each other (much in the way most good drivers disdain most Prius slow pokes, but with even more urgency, because Tesla went all the way, and their owners are all that much more of a target). A pickup would erase most of that antagonism. Right now, owning a Tesla is something tantamount to a declaration of war against all the manly sport-watching beer-drinking hicks from the center of the country and anybody even halfway like them (almost like a Trump vs Tesla branding, even though that doesn't exist in real life). I am a Tesla Model S owner, and the sneers I get from all of the pro-strength pro-business people scare me; I have to hide my car constantly. It's so bad I miss the days of the sneers I used to get for being a Mercedes snob, because at least that was an established position, and it just slotted me into lifestyles where everyone treated me great, even if they were dismayed they had to, with the occasional car chase from some invader immigrant who wanted me dead just because Mercedes; with Tesla, about the best I get is that the various Social Justice Warriors give me a slightly wider berth, slightly less animosity, put me a few months later in the schedule of their planned genocide against us --- hardly a win, in my book. I'd rather Tesla be a brand that people adore, like, respect, or accept, rather than get so damned vicious about. To me, a pickup (and work van) would do that, because those on "the other side" would suddenly have access to an electric vehicle that works "on their side". 18 months feels like an eternity. 2-3 years to market is 2020-2021.

Meanwhile, in terms of pollution, heavy cargo trucks are a big target. It's almost as if Tesla wants to solve environmental issues at the expense of solving their business issues. At least they're doing both, so that's the good news, but it seems like they're dragging their feet about it all.

They should play the US manufacturing side. Put out a pickup, make it in West Virginia and make some commercials about deploying fence posts out in the range.

I have to think:
1) battery allows for remote powering of heavy-duty electrical tools
2) battery allows for indoor use
3) battery allows for powering of campers and so on without generator noise
4) battery allows for better weight distribution, possibly even the ability to *shift* weight on demand
5) motor more easily made waterproof (I don't know about this one, might be interesting).
6) Is it easier to create ground clearance if the batteries-on-the-floor thing is given up.. put the batteries in bed + frunk?
7) TORQUE
8) One motor per wheel?

 

[N5329K]

It's odd Tesla hasn't gone into the pickup (and work van) market. There is a huge sector of society that treats Tesla drivers like crap because they're too antagonistic against each other (much in the way most good drivers disdain most Prius slow pokes, but with even more urgency, because Tesla went all the way, and their owners are all that much more of a target). A pickup would erase most of that antagonism. Right now, owning a Tesla is something tantamount to a declaration of war against all the manly sport-watching beer-drinking hicks from the center of the country and anybody even halfway like them (almost like a Trump vs Tesla branding, even though that doesn't exist in real life). I am a Tesla Model S owner, and the sneers I get from all of the pro-strength pro-business people scare me; I have to hide my car constantly. It's so bad I miss the days of the sneers I used to get for being a Mercedes snob, because at least that was an established position, and it just slotted me into lifestyles where everyone treated me great, even if they were dismayed they had to, with the occasional car chase from some invader immigrant who wanted me dead just because Mercedes (and that 300hp helped a lot); with Tesla, about the best I get is that the various Social Justice Warriors give me a slightly wider berth, slightly less animosity, put me a few months later in the schedule of their planned genocide against us --- hardly a win, in my book. I'd rather Tesla be a brand that people adore, like, respect, or accept, rather than get so damned vicious about. To me, a pickup (and work van) would do that, because those on "the other side" would suddenly have access to an electric vehicle that works "on their side". 18 months feels like an eternity. 2-3 years to market is 2020-2021.

Meanwhile, in terms of pollution, heavy cargo trucks are a big target. It's almost as if Tesla wants to solve environmental issues at the expense of solving their business issues. At least they're doing both, so that's the good news, but it seems like they're dragging their feet about it all.

I drive the speed limit in my Priuses and yes, there are some people who want me to go faster. I see that as their problem, not mine. Anyway, around here you can't throw a rock and not hit a Prius, so it's become just another part of the natural automotive landscape.
Robin

 

[Reciprocity]

Autonomous also allows trucks to platoon as if they were one vehicle. If Tesla owned the freight company, they could platoon a lot of trucks and achieve even more efficiency. They could even rotate lead trucks to balance out the batter load throughout the platoon to skip recharging stops.

 

[racer26]

been thinking about the take rate on autopilot. if that number is really at 75%, it makes me wonder what happens on the model 3 margins? maybe on the model 3 a $5k option only is taken at a 50-60% rate, but still... that's a $2500-3000 per vehicle average.

that's almost all profit too because the development cost with the same sensor suite is sort of included in the other vehicles. $2500-3000 is probably going to be a solid 5-8% of the model 3 price. that's a big kick to whatever gross margin would be, maybe 1/3rd the overall profits if the thing runs at 24% gross margin. pretty stunning.

This is a key part of why I say there is no way EAP is a $5k option on Model 3.

Its $5k on MS/MX because it can be. A 5k option on a car that starts at 70k and quickly tops 100k is one thing.

5k on 35k makes it a significant percentage of the cost of the car, and many more users will decide they don't need it that bad. The take rate will go down dramatically. Tesla wants ideally at or near a 100% take rate, and the higher volume of Model 3 will mean that they can charge less for it without giving up margin on a percentage basis.

 

[AcesDealt]

This is a key part of why I say there is no way EAP is a $5k option on Model 3.

Its $5k on MS/MX because it can be. A 5k option on a car that starts at 70k and quickly tops 100k is one thing.

5k on 35k makes it a significant percentage of the cost of the car, and many more users will decide they don't need it that bad. The take rate will go down dramatically. Tesla wants ideally at or near a 100% take rate, and the higher volume of Model 3 will mean that they can charge less for it without giving up margin on a percentage basis.

I'll wager that if they lower the price on 3, they will on S/X.

 

[racer26]

I'll wager that if they lower the price on 3, they will on S/X.

Quite possibly, else you anti-sell the S/X.

 

[Ulmo]

Can I get a reference on denying big data?

It was many years ago. I'd have to look it up. I'm pretty sure Elon said it, and there's a good chance it's stuck on his Twitter feed somewhere, or in a very old quarterly conference call, or some random interview in some random outdated publication with bit rot. I hate to say it, but in terms of time to corroborate my statement (which I remember pretty clearly from what he said), it might be quicker to just ask him how his ideas about Big Data have evolved with respect to AI, and whether he tried to mislead others in the past about this (most likely to protect Intellectual competitiveness). He has since started OpenAI, so I have a feeling his position on this has changed very dramatically as he has had time to fully immerse himself in the issues and come up with better overall approaches.