2017 Investor Roundtable:General Discussion

[anon]

It seems that the Livestream is not working. The announcement is unrelated to Tesla: two people paid for a flight on Crew Dragon around the Moon, launched by Falcon Heavy sometime in 2018.
Jeff Foust (@jeff_foust) | Twitter

 

[ggies07]

Couldn't find anything online, so I went straight to the source!

SpaceX to Send Privately Crewed Dragon Spacecraft Beyond the Moon Next Year

 

[mrdoubleb]

 

[Paracelsus]

Perhaps Goldman Sachs owns enough of Livestream to keep Elon off the air today?

 

[larmor]

Perhaps Goldman Sachs owns enough of Livestream to keep Elon off the air today?

So same leadership that can send people to the moon, that even the US govt has not been able to do in over forty years, but can't build a cheaper EV that is designed to be manufacturing wise easy to put together?

Where do i send my check to GS? They are sooo intelligent...

 

[Papafox]

I wouldn't really call that a best case scenario. We all know how many times they've failed before capturing that video to show-off. And to sell a product, one needs to be consistent. And arguably the product is as good as the worst one to potential customers.

You wouldn't call the Tesla video a "best case scenario"? We're just getting into a verbal exercise if you wanted me to say the Tesla video was a "better than best case scenario". My point is that the reality lies somewhere in between the Tesla video and this video taken on a street that is turning too sharply for the AP 2.0 software to work well with. Your reaction that of the two videos, the customer experience will be closest to the second video isn't rational. I have a Tesla with AP 1.0 hardware and software and know certain streets that exceed the design capabilities of the AP 1.0 hardware/software combination. I don't use the AP 1.0 on those short segments, but I get a tremendous amount of value out of the autopilot using it on the 95% of scenarios that don't exceed its current design parameters. Why don't you think AP 2.0 users will do the same? Granted, AP 2.0 needs to quickly mature to get highway driving up to reasonable allowed speeds, but once that step is reached, the hardware/software combination becomes very useful for the users. To believe that the usability will not be there is just silly.

 

[MitchJi]

That would be a worry indeed. But again, the competition is still playing catch-up. I still think that if Tesla gets AP2 on a significantly better level then AP2 in a time frame of 6 months it will all be fine. But you have a point that it will start to hurt very quickly if it takes Tesla in 2018 before they get there.

In 2018 when Tesla is producing 10k M3's per week it won't matter, at least the SP will be over $400 either way.

From the ER:
Our Model 3 program is on track to start limited vehicle production in July and to steadily ramp production to exceed 5,000 vehicles per week at some point in the fourth quarter and 10,000 vehicles per week at some point in 2018.

I consider 10k as an exit rate for 2018. I don't think it's that simple.

Do you really believe even if it takes Tesla until the last week of December 2018 to hit 10k per week that it's possible for that to happen without the SP already being over $400? How? Id rate the odds at zero. That could happen if they only hit something like 2k per week in November 2018, then hit 10k in December 2018.

We've seen two extreme case videos. On the one hand, we had the best case scenario, which is Tesla's "Paint it Black" video of the Tesla tooling safely around Palo Alto streets. Today, we saw the other extreme, a video taken on a winding road that contains too abrupt of turns for the current Tesla 2.0 software to adequately handle. That video was intended to make the AP 2.0 looks as bad as possible, just as the Tesla video was intended to make it look as good as possible. Clearly, reality lies somewhere in between.

It looks like a lot of the problems in the video were caused by the failure to recognize it was not a divided road, which should be relatively simple to fix.

 

[Runarbt]

Is it just me, or are others also feeling a bit relived about all these hopeless short articles popping up today?

Orange Peel with a stupid short article on SA, among others opening with a white lie about SP target below $200, when it has never been above.

Today certainly feels like a new Mark Spiegel moment, desperate short going "all in" trying for keep the downwards momentum, and hence marking the bottom? ;-)

 

[Waiting4M3]

The short term risk to Tesla is that they are now in about last place in the self driving car competition. That takes some luster off the glow.

"In last place" based on the new HW/SW that they've released into customers hands for ~ 2 months. We don't know what they have in their R&D. Eveyone else is showing off their R&D stuff, in controlled environment, with non-production ready HW.

 

[Fallenone]

You wouldn't call the Tesla video a "best case scenario"? We're just getting into a verbal exercise if you wanted me to say the Tesla video was a "better than best case scenario". My point is that the reality lies somewhere in between the Tesla video and this video taken on a street that is turning too sharply for the AP 2.0 software to work well with. Your reaction that of the two videos, the customer experience will be closest to the second video isn't rational. I have a Tesla with AP 1.0 hardware and software and know certain streets that exceed the design capabilities of the AP 1.0 hardware/software combination. I don't use the AP 1.0 on those short segments, but I get a tremendous amount of value out of the autopilot using it on the 95% of scenarios that don't exceed its current design parameters. Why don't you think AP 2.0 users will do the same? Granted, AP 2.0 needs to quickly mature to get highway driving up to reasonable allowed speeds, but once that step is reached, the hardware/software combination becomes very useful for the users. To believe that the usability will not be there is just silly.

It's not best case scenario because it's not a realistic scenario at all. It's an aspiration. The real scenario now is we're past 2 months when the initial estimate of AP1.0 parity with this performance.

 

[MitchJi]

What I mean with insufficient testing includes production testing.
1. Before rollout typical auto makers spend a good amount of time hashing out production issues during their testing. As you stated it appears Tesla is still ironing out production problems by delivering cars to customers diet and fixing problems second.
2. This can and should be avoided by fixing those problems earlier through more exhaustive testing and test production runs prior to shipping cars. If Tesla would have done that, the MX rollout would have been so much better.

1. Are you sure about that? CR states much higher problems with all OEM's with new models.

2. That's the crux of our differing opinions. It took Tesla almost a year of actually producing about 10k-15k to solve the MX production issues. You seem to think that Tesla could have solved those problems much more quickly and easily with "more exhaustive testing and test production runs". I doubt that Elon would agree, although I'm sure that they will do better in the future when confronted with similar challenges.

But that said, I do agree that it appears Tesla is taking a different approach with Model 3 as it'll be easier to manufacturer (so hopefully less quality issues) and they'll use their own employees as guinea pigs of sorts. So hopefully when Model 3 gets delivered to non-employee customers the quality will be good.

However, what I mean by "stellar" isn't exclusive to quality. Quality (in terms of reliability) is just part of the equation. The car has got to perform like no other car in its category. That's what the bulls are betting on. And what the bears are betting against.

I don't think it has to be that good to be wildly successful. But fortunately we will never find out, because it will definitely be superior to every other car in its category! The only reasonable question is will it be better than the current MS-MX's. There are good reasons they are waiting until very close to production to do the complete reveal (not the part 3 reveal that will probably happen in March). Because (they are concerned with cannibalization, which says a lot IMO.

 

[Papafox]

It's not best case scenario because it's not a realistic scenario at all. It's an aspiration. The real scenario now is we're past 2 months when the initial estimate of AP1.0 parity with this performance.

I completely agree with you that the Tesla video wasn't a realistic video of current capabilities. My point was that the video of the frequent failures of the AP 2.0 hardware/software combo was every bit as much a work of fiction. Whoever took that video knew that this particular street exceeded the capabilities of AP 2.0 and would produce horrible results. The video was an attempt to show AP 2.0 in the worst possible light, giving a performance that was far, far worse than what you would typically see with the Autopilot. I get great utility out of my AP 1.0 system, but I could also create a worse case scenario video that would make AP 1.0 look bad too, simply by using my knowledge of the system's weaknesses to choose the worst locations for the video. What I'm trying to share, as a Tesla autopilot user, is that the bad AP 2.0 video does not give a realistic view of utility that can be achieved from that hardware/software. You need to look somewhere between the two videos to find reality. Both extremes are fictitious distortions of the current capabilities of the system.

 

[TMSE]

I completely agree with you that the Tesla video wasn't a realistic video of current capabilities. My point was that the video of the frequent failures of the AP 2.0 hardware/software combo was every bit as much a work of fiction. Whoever took that video knew that this particular street exceeded the capabilities of AP 2.0 and would produce horrible results. The video was an attempt to show AP 2.0 in the worst possible light, giving a performance that was far, far worse than what you would typically see with the Autopilot. I get great utility out of my AP 1.0 system, but I could also create a worse case scenario video that would make AP 1.0 look bad too, simply by using my knowledge of the system's weaknesses to choose the worst locations for the video. What I'm trying to share, as a Tesla autopilot user, is that the bad AP 2.0 video does not give a realistic view of utility that can be achieved from that hardware/software. You need to look somewhere between the two videos to find reality. Both extremes are fictitious distortions of the current capabilities of the system.

A real positive that comes out of this, is that the AP team at Tesla are already on it. Like in medicine, proper identification of ailment or problem is a much bigger issue than coming up with remedy.

In a way, bug testing is being open sourced here. Comprehensive testing isn't easy, that's why we see many software companies sponsor hackathons to break their systems. The real intent of the guy who produced those AP2 failure videos not withstanding, this is actually helping Tesla.

 

[Fallenone]

I completely agree with you that the Tesla video wasn't a realistic video of current capabilities. My point was that the video of the frequent failures of the AP 2.0 hardware/software combo was every bit as much a work of fiction. Whoever took that video knew that this particular street exceeded the capabilities of AP 2.0 and would produce horrible results. The video was an attempt to show AP 2.0 in the worst possible light, giving a performance that was far, far worse than what you would typically see with the Autopilot. I get great utility out of my AP 1.0 system, but I could also create a worse case scenario video that would make AP 1.0 look bad too, simply by using my knowledge of the system's weaknesses to choose the worst locations for the video. What I'm trying to share, as a Tesla autopilot user, is that the bad AP 2.0 video does not give a realistic view of utility that can be achieved from that hardware/software. You need to look somewhere between the two videos to find reality. Both extremes are fictitious distortions of the current capabilities of the system.

I understand that particular video does present a challenging case for AP2.0 (night, curvy road). But after reading the thread, I do get the sense that many owners are not happy with the current status of AP2.0. And some claim AP1.0 could have done a better job under similar circumstances. This is where I'm worrying about. I heard about AP2.0 not performing well before but wasn't expecting this.

 

[schonelucht]

You have to admit that Tesla taking on responsibility for self-driving accidents or incidents would be unrealistic if it is not also making revenue from selling insurance for the feature. Elon has repeated said that self-driving will not be accident-free, it will just be much better than with a driver behind the wheel.

Tesla is making money : each car they sell is $30k gross margin. Stated goal is a 90% reduction in accidents (and therefore a 90% reduction insurance costs). Roll that into a one time $500 price increase for the Model S/X (same model as unlimited supercharger) and you are making value propositions to your customers from strength, not weakness.

I see Tesla's willingness to enter the insurance market as a positive. My insurance company has given me no discount after it was announced that Teslas with AP 1.0 hardware were 40% less likely to be involved in an accident than a non-autopilot-equipped vehicle.

Your insurance company knew that already long before Elon knew it. It didn't reduce your premiums either because it figured out it can keep the gain for itself or because it balances it out with the higher than average repair costs for Teslas. Likely a combination of both.

 

[schonelucht]

What I'm curious about is whether driving on local streets will allow Tesla to claim some of the self-driving revenue in Q1. Local streets was not available on AP 1.0, and so it is an enhanced capability.

No. Local streets is the enhanced part of EAP. Elon stated the the first features of FSD are only expected in maybe 3 months. So they will not recognise any FSD revenue this quarter. They will likely recognise full EAP revenue (if they manage to actually reach parity as planned by next month)

 

[dw4ngg]

Everyone worried about this pullback is trippin!! This company made $7b in revenues last yr, has a market cap of less than $40b.... and hasn't even sold a single model of its revenue-killer yet!!!

Yes there's risk, but every company has risk! Tesla has shown they can produce; do you really want to bet against their ability to scale with all the gigafactories in the pipeline? No sense in worrying about short term price either...

Worried about cash? Why??? They can raise capital easily!

It's all a Jedi mind f***. Don't fall for it. Keep your future intact. Keep your tesla. In fact, buy more. What's 500k times $40k? Oh yeah, $20b!!! (In 2018, conservative estimate)

 

[Krugerrand]

Please try harder.

Don't want to. I know what you said. And you said it again.

 

[neroden]

Looked at another way. Do you have the same disdain for 401k's?

Dunno whether the person you replied to has that disdain, but I absolutely do. They've proven to be great for the upper-middle class and to stink terribly for the lower classes. Most people simply aren't cut out to run self-directed investment accounts.

403(b)s have been around longer than 401(k)s -- they're like 401(k)s for college professors. I know a lot of college professors. My god -- practically none of them can manage their 403(b)s even half-competently. The few who do make a lot of money. The rest would have been better off with a defined-benefit pension.

Can you imagine how much more uncommon it is for assembly line workers to be able to invest competently in 401(k)s? Well, there have been studies about it.

Some opinion pieces which agree with me about 401(k)s:
The 401(k) World Sucks
401(k) Retirement Funds Are a Rip-Off

Hell, I have a friend who works at a university who didn't even bother to find out how much they were putting in her 403(b) or what it was being invested in, and couldn't be bothered to even after I pushed her. She might be doing better than the people getting scammed by fund managers, but I'm pretty sure she'd be better off with a defined-benefit pension.

I mean, I'm going to take advantage of these tax-advantaged retirement vehicles while they're available to me and my family, but on the whole I think they've been bad for most people.

Rant over.

 

[Krugerrand]

Yup. Until the moment when Elon goes on record using the lottery winnings in salary comparisons and saying that lottery winnings will continue. That is what is irresponsible and misleading, especially when you are talking to people who've never dealt with equity participation before.

They will be worth more. He's publically said he's going for a $1T market cap. Nothing misleading about that. It's his intent, whether you personally believe it or not is irrelevant. He'll do it or die trying. That's probably good enough for most Tesla employees.

And again, I'd hope to work at Tesla you'd have to be smart enough to understand equity participation...otherwise how did you get to work there in the first place?