2017 Investor Roundtable:General Discussion

[StapleGun]

FYI - I'm getting an increasing number of questions about Model Y.

I expect Model Y reservations within 24 hours to exceed 500,000.

This is an open bet for first come, first serve.

I'll bet you a T-shirt of winners choosing from the Tesla store that they do not get 500,000 reservations for Model Y in the first 24 hours.

*Fine print: 24-hour period starts when reservations are available to the general public. A winner can only be determined if Tesla releases official numbers confirming the total is above or below 500k for the first 24 hours. Winner responsible for notifying loser at the conclusion of the bet. If reservations aren't taken then the bet is a push.

Deal?

 

[AcesDealt]

I would modify any Model Y gambles to be based on $ deposit totals because Tesla may very well increase those and offset unit counts, and I'm more interested in how much money they can bank by the sheer whiteness of Elon's smile.

 

[neroden]

Update: McConnell is reported to deny this report.

I looked at the treason clause in Article III of our constitution which has rather strict standards. Contravening the Constitution in the referenced post is not a permitted treason. "Treason against the United States, shall consist only in levying War against them, or in adhering to their Enemies, giving them Aid and Comfort. No Person shall be convicted of Treason unless on the Testimony of two Witnesses to the same overt Act, or on Confession in open Court." I wonder what thoughts Special Investigator Mueller has, or if Trump's attorneys have followed the logic of this argument and have motive to insist the investigation be shut down? It seems during the campaign Trump has already assisted our enemies.

I have no idea if violating oath of office is a crime. Should I warn my Congressperson and Senators about contravening the Constitution's protection of debt in Section 4 of the Fourteenth Amendment?

No, violating the oath of office is not a crime. It is an *impeachable offence*, however. That was made very clear by the people who wrote the Constitution.

 

[Buddyroe]

WSJ casts negative light on Tesla’s Autopilot ambitions

If you google search the WSJ title, you will be able to find the full article somewhere.

"Tesla’s Push to Build a Self-Driving Car Sparks Dissent Among Its Engineers"

It has a whole load of historical stuff which is irrelevant. The most pertinent info I found is this:

------

For $5,000, Tesla customers can buy an option called “Enhanced Autopilot.” That, Tesla said, would give them access to four of the car’s eight cameras as well as the radar, 12 sensors, and the new onboard computing system.

For another $3,000, drivers get the right to activate the rest of the cameras when Tesla enables a full self-driving system. Customers could wait to buy the option, but it would be more expensive.

The announcement shook up some engineers, because they believed that the product that was released wasn’t designed to be self-driving, according to several people interviewed.

The marketing was a factor in the decision by Mr. Anderson and at least two other engineers to leave the company, according to people familiar with the matter.

------

My confidence in Tesla's claims of Full Self Driving on existing hardware (even with upgraded computers) has flip-flopped over time to be honest.

I spoke with one person on the team, who is no longer there now. When I expressed my surprise at Musk claiming FSD in 2 years back then, he said "I have no idea either". That caused me to doubt Musk's claims to some degree. But over time as Musk continued to confidently claim the capabilities I started to believe him. Especially since Tesla started charging for it. Tesla has to be reasonably confident to be able to actually charge money I thought. But now we see this... I have no idea what to believe anymore.

I can only laugh when I read you guys debating ANYTHING pertaining to FSD and Tesla. AP2 has become a joke among Tesla owners. I'd be hard-pressed to believe Tesla will achieve FSD in the next 10 years when they can't even bring AP2 to parity with AP1. Heck, they can't even get TACC to work correctly with AP2. Every other car company is way ahead of Tesla with TACC.

The threads linked below are from this very web page. And the small sample of comments are from the last week (and from page 1 of 8 pages of people griping about AP2).

And you people think FSD is even a remote possibility? Do you not own an AP2 car? Do you not read these owners' threads? Talk to owners? Or do you just read press releases from Tesla?

AP2 Totally Sucks
AP2 Totally Sucks 2

"This morning when approaching a bridge overpass traveling at 65mph the car very quickly decelerated to 45mph, throwing everything in the passenger seat into the floorboard BUT MORE IMPORTANTLY causing everyone behind her to start braking hard and almost caused a major accident."

"Indeed! Most of us (AP2) have experienced this sudden slow down and/or lane departure when on AP. Its getting very tiring to always have to watch like a hawk and override brake/accelerator/steering when driving along in normal traffic/roads and the AP suddenly gets a wild hare in its 'brain' and does something dumb and dangerous."

"The 20/75/85 ramps and turns are horrible on "silky smooth". More like, "hug.the.inside.lane.marker" or "disengage.for.every.bend"

"While it has gotten much better with each release, when the road ahead is wide open and the car slams on the brakes, for what, invisible hazards?"

This one is a classic:
"But honestly, I think driving myself is less stressful than constantly being on edge waiting for the car to do the wrong thing. When driving myself I only have to worry about everyone else doing something stupid, not everyone else AND my own car doing something stupid."

"I tense [SIC] when using TACC with bridges and use AP only to test it by myself after it tried to follow a dark line on the road to an adjacent concrete wall instead of following the straight roadmarking lines ahead..."

"Yeah, I just don't use EAP or TACC anymore except for brief periods. Even plain old cruise would be better than TACC that does stuff totally by surprise."

"This week my EAP braked on I94 because of the shadow of a semi in an adjacent lane. It registered the shadow as a car entering my lane."

"Letting AP2 EAP drive is a lot like teaching a teenager to drive as they make noob mistakes.... except you are at least seated in the drivers position and don't have to reach so far to grab the wheel....."

"I understand keep hand on wheel for EAP but literally every car manufacturer in the world including tesla on ap1 had or has tacc function that doesn't try to get you in car accidents. I'm in the same boat with my wife not wajtijg to ever use eap or tacc because of this very reason. "

"I'm sorry for you guys. Don't understand why after so much time AP2 is still hit and miss. How can Tesla even speak of FSD when they can't keep a decent upgrade valid for longer than a week or so before it's replaced by a downgrade? "

"Forget the AP stuff, let's just look at TACC. On my Audi, it's TACC equivalent has been 100% flawless in the 13 months I've owned it. Zero occasions where it has randomly decelerated, or failed to see a car in front of it. On the flip side, my AP2 has had issues with both; it randomly decelerates on the freeway, and has almost rear ended cars in front of me (didn't slow down at all when approaching). At this point i do not trust it at all. "

"We'll another release and nothing special, this is getting down right sad."

"I don't want to brag or be seen as an elitist, but I own an exotic car with stars on it ... it's called a Subaru. It has TACC that is also flawless. And lane maintenance assist that has been flawless. And I can have plain old cruise instead of TACC if I push a "button." It's a shame that my tesla has its functionality set in stone such that I cannot disable the beta TACC functionality and just have basic cruise, but that would take dozens of lines of code."

 

[dennis]

So "on schedule" seems about right...

Maybe slightly ahead since what we've seen so far was probably manufactured first half of August.

There is also the issue of the EPA certification which is supposedly delayed a month. IIRC Tesla could deliver 100 M3's without the certification under some form of exemption. But the plan of 30 in July plus 100 in August would exceed that...

 

[neroden]

Tesla is really pushing out inventory sales this quarter. Even going as far as bringing back 2 year leases on unsold AP1 inventory (and obviously a large discount that reflects age and use of the car). Anyone up for making the traditional guess-the-deliveries-poll for Q3? I am voting 30k+ for this quarter : 25k newly produced, 3k sold out of inventory and 2k Model 3. Curious at what everyone else is thinking.

I think Tesla wants cash this quarter and next. I suspect they're coming close to being capital-constrained ramping up Model 3 production, Service Centers (they are way way behind on these), the Gigafactory, and the Buffalo factory -- not to mention plans for future Gigafactories, Hyperchargers, etc. I suspect cash this quarter is worth far more than cash in Q1 2018 -- probably allows advancement of at least one project by six months.

Q3 is, of course, the quarter with most of the Model 3 costs but very low production rates. Q4 will start to look more self-financing, and Q1 should be generating cash. Cashflow planning dictates that every dollar which can be extracted *now* is worth it. Which probably accounts for the bonus being offered to Jon McNeill for sales during Q3/Q4.

 

[neroden]

No. Just, no.

Tesla would be on the hook for all the bad pub of "firing" drivers. It would be a mess. Calacanis is only proposing this because of his investment in Uber.

Not to mention that the massive sex-discrimination scandal and sexual harassment scandals at Uber are exactly what Tesla DOESN'T need right now, given that Tesla seems to have some of its own problems in that arena. It would send exactly the wrong message to merge with a company with that bad a reputation.

 

[dennis]

I can only laugh when I read you guys debating ANYTHING pertaining to FSD and Tesla. AP2 has become a joke among Tesla owners. I'd be hard-pressed to believe Tesla will achieve FSD in the next 10 years when they can't even bring AP2 to parity with AP1. Heck, they can't even get TACC to work correctly with AP2. Every other car company is way ahead of Tesla with TACC.

The threads linked below are from this very web page. And the small sample of comments are from the last week (and from page 1 of 8 pages of people griping about AP2).

And you people think FSD is even a remote possibility? Do you not own an AP2 car? Do you not read these owners' threads? Talk to owners? Or do you just read press releases from Tesla?

AP2 Totally Sucks
AP2 Totally Sucks 2

"This morning when approaching a bridge overpass traveling at 65mph the car very quickly decelerated to 45mph, throwing everything in the passenger seat into the floorboard BUT MORE IMPORTANTLY causing everyone behind her to start braking hard and almost caused a major accident."

"Indeed! Most of us (AP2) have experienced this sudden slow down and/or lane departure when on AP. Its getting very tiring to always have to watch like a hawk and override brake/accelerator/steering when driving along in normal traffic/roads and the AP suddenly gets a wild hare in its 'brain' and does something dumb and dangerous."

"The 20/75/85 ramps and turns are horrible on "silky smooth". More like, "hug.the.inside.lane.marker" or "disengage.for.every.bend"

"While it has gotten much better with each release, when the road ahead is wide open and the car slams on the brakes, for what, invisible hazards?"

This one is a classic:
"But honestly, I think driving myself is less stressful than constantly being on edge waiting for the car to do the wrong thing. When driving myself I only have to worry about everyone else doing something stupid, not everyone else AND my own car doing something stupid."

"I tense [SIC] when using TACC with bridges and use AP only to test it by myself after it tried to follow a dark line on the road to an adjacent concrete wall instead of following the straight roadmarking lines ahead..."

"Yeah, I just don't use EAP or TACC anymore except for brief periods. Even plain old cruise would be better than TACC that does stuff totally by surprise."

"This week my EAP braked on I94 because of the shadow of a semi in an adjacent lane. It registered the shadow as a car entering my lane."

"Letting AP2 EAP drive is a lot like teaching a teenager to drive as they make noob mistakes.... except you are at least seated in the drivers position and don't have to reach so far to grab the wheel....."

"I understand keep hand on wheel for EAP but literally every car manufacturer in the world including tesla on ap1 had or has tacc function that doesn't try to get you in car accidents. I'm in the same boat with my wife not wajtijg to ever use eap or tacc because of this very reason. "

"I'm sorry for you guys. Don't understand why after so much time AP2 is still hit and miss. How can Tesla even speak of FSD when they can't keep a decent upgrade valid for longer than a week or so before it's replaced by a downgrade? "

"Forget the AP stuff, let's just look at TACC. On my Audi, it's TACC equivalent has been 100% flawless in the 13 months I've owned it. Zero occasions where it has randomly decelerated, or failed to see a car in front of it. On the flip side, my AP2 has had issues with both; it randomly decelerates on the freeway, and has almost rear ended cars in front of me (didn't slow down at all when approaching). At this point i do not trust it at all. "

"We'll another release and nothing special, this is getting down right sad."

"I don't want to brag or be seen as an elitist, but I own an exotic car with stars on it ... it's called a Subaru. It has TACC that is also flawless. And lane maintenance assist that has been flawless. And I can have plain old cruise instead of TACC if I push a "button." It's a shame that my tesla has its functionality set in stone such that I cannot disable the beta TACC functionality and just have basic cruise, but that would take dozens of lines of code."

I don't disagree that AP2 is a disaster. But the interesting thing about Tesla is that when they have a disaster they are willing to admit it (to themselves) and take action. A prime example is the Model X and your favorite falcon wing doors. They focused engineering resources on the problems and solved them. I suspect the same thing is happening with AP2, along with the engineering management changes in this area that we already know about.

AP1 was equal to or better than competitive offerings. Tesla suffered a huge setback with the Mobileye breakup, much bigger than they thought. They are climbing out of the hole much more slowly than they expected. But don't discount their will and their engineering focus. Fortunately Tesla is resiliant enough that the AP2 problems are not the end of the world.

 

[AcesDealt]

To be honest, my running supposition is that the actual FSD technology will arrive nearly simultaneously. In the vein of Leibnez and Newton co-discovering calculus. You have a highly liquid group of incredibly well paid engineers bouncing between these firms. And you have a decade+ old idea that 'IT is dead' from some Harvard dude I forgot which is nonetheless true. The bottomline is that drama behind FSD is inside-baseball. It's also fairly irrelevant WHEN it arrives since the longer it takes the more established Tesla can get in EV manufacturing which will in the long-run be the necessary competitive factor for robotaxis. All this being said, I do think that Elon has been erratic and inaccurate in his predictions regarding the timing of autopilot and AP2 is a joke and the whole department is a twisted array. I just hope it's a little better than say Waymo which appears to be happy and complacent about achieving nothing in particular.

 

[dennis]

I think Tesla wants cash this quarter and next. I suspect they're coming close to being capital-constrained ramping up Model 3 production, Service Centers (they are way way behind on these), the Gigafactory, and the Buffalo factory -- not to mention plans for future Gigafactories, Hyperchargers, etc. I suspect cash this quarter is worth far more than cash in Q1 2018 -- probably allows advancement of at least one project by six months.

Q3 is, of course, the quarter with most of the Model 3 costs but very low production rates. Q4 will start to look more self-financing, and Q1 should be generating cash. Cashflow planning dictates that every dollar which can be extracted *now* is worth it. Which probably accounts for the bonus being offered to Jon McNeill for sales during Q3/Q4.

@SBenson has a very credible hypothesis that the McNeil bonus and the move of Model S US custom orders to October this week are because there is going to be a hardware update to S & X in October.

Website wait times for delivery change

 

[Crowded Mind]

I can only laugh when I read you guys debating ANYTHING pertaining to FSD and Tesla. AP2 has become a joke among Tesla owners. I'd be hard-pressed to believe Tesla will achieve FSD in the next 10 years when they can't even bring AP2 to parity with AP1. Heck, they can't even get TACC to work correctly with AP2. Every other car company is way ahead of Tesla with TACC.

The threads linked below are from this very web page. And the small sample of comments are from the last week (and from page 1 of 8 pages of people griping about AP2).

And you people think FSD is even a remote possibility? Do you not own an AP2 car? Do you not read these owners' threads? Talk to owners? Or do you just read press releases from Tesla?

AP2 Totally Sucks
AP2 Totally Sucks 2

"This morning when approaching a bridge overpass traveling at 65mph the car very quickly decelerated to 45mph, throwing everything in the passenger seat into the floorboard BUT MORE IMPORTANTLY causing everyone behind her to start braking hard and almost caused a major accident."

"Indeed! Most of us (AP2) have experienced this sudden slow down and/or lane departure when on AP. Its getting very tiring to always have to watch like a hawk and override brake/accelerator/steering when driving along in normal traffic/roads and the AP suddenly gets a wild hare in its 'brain' and does something dumb and dangerous."

"The 20/75/85 ramps and turns are horrible on "silky smooth". More like, "hug.the.inside.lane.marker" or "disengage.for.every.bend"

"While it has gotten much better with each release, when the road ahead is wide open and the car slams on the brakes, for what, invisible hazards?"

This one is a classic:
"But honestly, I think driving myself is less stressful than constantly being on edge waiting for the car to do the wrong thing. When driving myself I only have to worry about everyone else doing something stupid, not everyone else AND my own car doing something stupid."

"I tense [SIC] when using TACC with bridges and use AP only to test it by myself after it tried to follow a dark line on the road to an adjacent concrete wall instead of following the straight roadmarking lines ahead..."

"Yeah, I just don't use EAP or TACC anymore except for brief periods. Even plain old cruise would be better than TACC that does stuff totally by surprise."

"This week my EAP braked on I94 because of the shadow of a semi in an adjacent lane. It registered the shadow as a car entering my lane."

"Letting AP2 EAP drive is a lot like teaching a teenager to drive as they make noob mistakes.... except you are at least seated in the drivers position and don't have to reach so far to grab the wheel....."

"I understand keep hand on wheel for EAP but literally every car manufacturer in the world including tesla on ap1 had or has tacc function that doesn't try to get you in car accidents. I'm in the same boat with my wife not wajtijg to ever use eap or tacc because of this very reason. "

"I'm sorry for you guys. Don't understand why after so much time AP2 is still hit and miss. How can Tesla even speak of FSD when they can't keep a decent upgrade valid for longer than a week or so before it's replaced by a downgrade? "

"Forget the AP stuff, let's just look at TACC. On my Audi, it's TACC equivalent has been 100% flawless in the 13 months I've owned it. Zero occasions where it has randomly decelerated, or failed to see a car in front of it. On the flip side, my AP2 has had issues with both; it randomly decelerates on the freeway, and has almost rear ended cars in front of me (didn't slow down at all when approaching). At this point i do not trust it at all. "

"We'll another release and nothing special, this is getting down right sad."

"I don't want to brag or be seen as an elitist, but I own an exotic car with stars on it ... it's called a Subaru. It has TACC that is also flawless. And lane maintenance assist that has been flawless. And I can have plain old cruise instead of TACC if I push a "button." It's a shame that my tesla has its functionality set in stone such that I cannot disable the beta TACC functionality and just have basic cruise, but that would take dozens of lines of code."

People sugared all over AP1 before the software came out, too. One week after the software came out, I posted a poll with a likert scale for owners to state how satisfied they were with AP1. 83% said they were satisfied or very satisfied with AP1.

Patience.

 

[JRP3]

Keep in mind that the number I suggested is nearly 3x of 180k Model 3 reservations in the first 24 hours.

Do you still agree with me?

Yes. 3x reservations 2 years from now should be no problem.

 

[RobStark]

Adient to purchase seating supplier Futuris Group for $360 million

Tesla's first seats where purchased from Futuris Group.

Then Tesla purchased Recaro brand seats from Johnson Controls for Tesla's "premium" seats.

Johnson Controls spun off their automotive seats and interiors business as "Adient."

Now Adient is buying Futuris Group.

Other than some/most of the parts for the Model X second row "monopost" seats all others as of Sept 30 are now from Adient I believe.

 

[Buddyroe]

A prime example is the Model X and your favorite falcon wing doors. They focused engineering resources on the problems and solved them.

Well, I don't know. I did a test drive this past Sunday of an S and an X. Obviously, I was incredibly impressed and my wife commented that they weren't cars - they were spaceships. She sounded like a little kid when after she had the opportunity to do a launch - she said, "I wanna do that again!"

But, there were 3 defects total between the cars. All were on the X. One was that the left side 3rd row seat would not fold down. The other two both pertained the FWDs. First, when left side door was just about completely opened, it made a fairly bad noise (kind of a screeching noise). The other issue was simply with alignment. My wife pointed out to me that the door was not aligned vertically with the rear quarter panel. She pointed at it and asked me if that was normal. I said, "yes AND no" (it's normal, but it shouldn't be).

 

[Buddyroe]

People sugared all over AP1 before the software came out, too. One week after the software came out, I posted a poll with a likert scale for owners to state how satisfied they were with AP1. 83% said they were satisfied or very satisfied with AP1.

Patience.

So...how long has AP2 been in production compared to AP1?

 

[NicoV]

Bros, I already solved the recharging problem for semis. It's a level 5 autonomous replacement battery pack on wheels that scurries around like a VW beetle and attaches\removes from the front of the tractor while still moving full speed down the highway. It also detaches before any weigh stations and rendezvous later.

edit: an optional charging fork can also skewer passing Leafs\Prius right in the battery and suck those dry when desperate\thirsty.

Great idea. Like refueling a fighter jet in flight!

 

[W84M3]

I see a lot of posts about how the autopilot team is in complete disarray and Elon doesn't act on it.

Elon is fully focused on production Model 3 scaling and I'm sure he'll sleep many nights at the factory floor to make sure they hit 5k/week in December. This is production hell.

It becomes clear they redesigned AP (HW 2.5) for Model 3 as well as building up a real navigation tool that should exceed all current navigation tools. Maps in the past have never been created with the purpose of FSD. It didn't really matter when a GPS showed 40mph but the street signs said 65mph; now we've got phantom braking.. Every little detail a navigation got wrong is now a significant problem.

The focus now is and should be on getting Model 3's out on the streets; for the company and for FSD.

Those will collect massive amounts of data and based on that, they will feed the FSD monster.

By now, they have maybe 70k Model S/X's with AP2 on the streets which is still peanuts IMO and the first months the data collected was almost inexistant.

In little more than a year from now, that number on the streets will quintuple. (current 70k + 100k X/S + 180k model 3's). That's when the real FSD monster will stand up.


Reliable data first. That's also why Tesla will maintain their advantage on FSD vs competitors. Nobody will come close to gathering data from 400k cars next year. Nobody.

Begin 2018, Elon will switch his focus to FSD & AI within Tesla and I'm sure 2018 will surprise many.

First things first, step by step. As it should.

 

[ValueAnalyst]

Yes. 3x reservations 2 years from now should be no problem.

If reservation level is an indication of demand, this may mean 2m annual year demand for Model Y since Model 3 may reach 700k annual demand per Elon. Would you agree with this as well?

 

[ValueAnalyst]

There is also the issue of the EPA certification which is supposedly delayed a month. IIRC Tesla could deliver 100 M3's without the certification under some form of exemption. But the plan of 30 in July plus 100 in August would exceed that...

Does this mean that Tesla is even further ahead of its production plan than I estimated?

 

[ValueAnalyst]

I see a lot of posts about how the autopilot team is in complete disarray and Elon doesn't act on it.

Elon is fully focused on production Model 3 scaling and I'm sure he'll sleep many nights at the factory floor to make sure they hit 5k/week in December. This is production hell.

It becomes clear they redesigned AP (HW 2.5) for Model 3 as well as building up a real navigation tool that should exceed all current navigation tools. Maps in the past have never been created with the purpose of FSD. It didn't really matter when a GPS showed 40mph but the street signs said 65mph; now we've got phantom braking.. Every little detail a navigation got wrong is now a significant problem.

The focus now is and should be on getting Model 3's out on the streets; for the company and for FSD.

Those will collect massive amounts of data and based on that, they will feed the FSD monster.

By now, they have maybe 70k Model S/X's with AP2 on the streets which is still peanuts IMO and the first months the data collected was almost inexistant.

In little more than a year from now, that number on the streets will quintuple. (current 70k + 100k X/S + 180k model 3's). That's when the real FSD monster will stand up.


Reliable data first. That's also why Tesla will maintain their advantage on FSD vs competitors. Nobody will come close to gathering data from 400k cars next year. Nobody.

Begin 2018, Elon will switch his focus to FSD & AI within Tesla and I'm sure 2018 will surprise many.

First things first, step by step. As it should.

Though I agree with the general sentiment of your post, IIRC Elon has said he already spends the vast majority of his time with FSD folks.