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

[Cosmacelf]

I still don't understand why Tesla would then sell a car with a NPV of $200,000 for $35,000.

Or $25,000. I forget. Plus, my head hurts.

Because they don’t have the cash to buy millions of $25,000 cars. Right now, no one believes these cars have an NPV of $200,000 so you can’t borrow the money either.

 

[mongo]

This snippet isn’t entirely true. For example, if you look at Tesla, they have a few specific mount points for the cameras. All their data comes from those particular mount points. The network is pretty certain to overfit to those common parameters. Any deviation from that(which is going to be unavoidable in arbitrary other cars) is going to make the network fail spectacularly. This is very likely a large component of why Elon says retrofitting AP2+ onto AP1 cars would be hard/impossible.

My understanding is that they run the same NN on all the cameras.
To retrofit AP1 would require installing new wire harnesses, new mounting for HW3, cutting holes for B camera and front repeater, and likely the other changes for redundancy.

 

[Causalien]

Trip Chowdry
Seriously, tomorrow Trip will put out a gigantic stock upgrade, he was very impressed.

Yep. Trip seems to have the chip design competency to understand this. HOwever.... he was deemed a loony as an analyst before so his upgrades won't matter.

It'd be really nicer, if TSLA have a level 5 ai ready and say "this is available as of tomorrow" Boom you see a big explosion in every trader's mind.

 

[Unpilot]

Ok I just got a break and managed to read through several pages.
Couple questions....anything on S/X refresh?
Any link to watch the demo drives?

Any idea why the SP is not $1000?

 

[anthonyj]

The market cap of Tesla right now is around 50 Billion. The market cap of Uber is 100 BILLION. Uber is going to have an IPO VERY SOON. Tesla at around $1000 a share would have a market cap of between 150 and 200 BILLION. IMO Uber is worthless and Tesla should be close to $1,000 a share now.

I agree that Uber is poop. But Uber has proven ride sharing revenue, Tesla doesn't. That's why it isn't being factored into SP right now

 

[Cosmacelf]

Is that camera moving?? Like the director was counting down...

Oh wow yeah. That camera was hand held. What the heck?

 

[jerry33]

AThen a year in, Waymo gets it act together and is offering packages to car makers, individuals and fleet managers. The industry happily buys licenses to produce and implement Waymo approved computer and sensor kits plugging into cars with the hardware controls in place for any kind of autonomy. Waymo prices the hardware at zero, takes a lower cut per ride or mile than Tesla. The hardware kits end up dirt cheap and convert dirt cheap econoboxes to compete with relatively costly Teslas. Now suddenly Tesla needs to become a premium company again, someone create value in a market that is all about price. What then?

This is somewhat like the recipe for rabbit stew: First you catch a rabbit.

It's very hard for a company to change directions. I doubt Waymo can do it, and certainly not without making their stock sink to practically zero because they are basically saying "We were wrong, sorry". Even if they can, they will need to get millions of cars on the road to catch up, develop their own boards and chips, etc., etc.

 

[Doggydogworld]

Pretty sure that’d only be for new sales. They still recognize those features as EAP ones for those who bought that prior to the switch.

Exactly, you can't retroactively change what you promised to deliver and start counting all that deferred revenue.

I'm shocked they didn't show demo rides on the Live Feed. I enjoyed the tech details from Bannon and Karpathy, but it was far from an investor-friendly webcast.

 

[Anstandswauwau]

Ok I just got a break and managed to read through several pages.
Couple questions....anything on S/X refresh?
Any link to watch the demo drives?

Any idea why the SP is not $1000?

Not a word, demo drives are under NDA methinks, TSLA is totally decoupled from Tesla

 

[Nocturnal]

Buyout or merger. But assuming Lidar solution is wrong. No one will take Waymo taxi.

My guess is Waymo will find a way to properly deploy it. It only has to compete in jurisdictions that approves robotaxis. So the rollout speed can be controlled while google lobbies for the approval to slow down. They can have the legislature slowly control robo taxi rollout as they finish testing each geofenced area.

If I'm reading the tech correctly, geo-fencing won't help if the system can't account for seeing a person on their cell phone and not anticipating that they are going to walk in front of traffic.

 

[geturchiru]

Lol and how much power does the Nvidia system pull and how much does it cost... googling didn’t turn up cost but I’m guessing 10-15k and 500watts seems to be ballpark for a dual high power GPU board.

A little old industry leading numbers are close to 1TOPS/W and Nvidia/Google are the worst performers in terms of TOPS/W at close to 0.5. It's not just about cost or power but about the controllability in terms of HW and SW basically vertical integration of FSD things

 

[Cosmacelf]

No, that is not the problem.

As long as they can run a 10 car fleet in San Jose for a week, banks will line up to provide funding.

Yeah, after you’ve convinced the banks of that. That won’t happen now, and won’t happen day one of FSD either. Maybe 6 months after FSD is released and no accidents...

 

[Cloxxki]

The market cap of Tesla right now is around 50 Billion. The market cap of Uber is 100 BILLION. Uber is going to have an IPO VERY SOON. Tesla at around $1000 a share would have a market cap of between 150 and 200 BILLION. IMO Uber is worthless and Tesla should be close to $1,000 a share now.

How did Facebook do in the years after IPO? Tesla already had its.
Uber is doing well. Tesla is a startup that stopped growing for now. Or it's a stable car maker with a crazy revenue multiple considering the lack of profit and dividend.

If Tesla limits it's FSD to self made cars, it will be self limited. We all know about their production woes. Once the competition has software that works, any non-Tesla already on the road or still to be built can undercut Tesla, pretty much. By the time this all goes down, the non-Tesla fleet of BEVs that are ready to become Level 5 taxis will be quite significant. And it can be grown as fast as cars can be built by the whole of the market. Offer 10% over sticker for any BEV with FSD controls coming out of factories, as any needed sensors and computer, start undercutting Tesla.

 

[MarcusMaximus]

Seems the narrative went from unachievable to too safe to allow manual driving any longer in a heart beat. I'm not convinced. And to my it's not FSD if it works only for certain countries or regions.

IMO, anyone who ever said it was unachievable was being, to be polite, short sighted. It’s clearly a tractable problem; we all have computers in our heads that handle it with ease.

For too safe to allow people to be manually driving, I’ve always found that to be inevitable at some point. Vast majority of accidents occur because of lack of attention(whether just distracted, inebriated or unconscious), and computers don’t have that problem. Reaction times can also be far superior to that of humans. Almost certainly will still be private tracks(and off-road/dirt areas where people can drive), though.

 

[Cosmacelf]

Ok I just got a break and managed to read through several pages.
Couple questions....anything on S/X refresh?
Any link to watch the demo drives?

Any idea why the SP is not $1000?

No, no, and people are stupid?

 

[T-Will]

Oh wow yeah. That camera was hand held. What the heck?

It looks like a handheld camera recording video playback from a security camera monitor.

 

[Anstandswauwau]

This is somewhat like the recipe for rabbit stew: First you catch a rabbit.

It's very hard for a company to change directions. I doubt Waymo can do it, and certainly not without making their stock sink to practically zero because they are basically saying "We were wrong, sorry". Even if they can, they will need to get millions of cars on the road to catch up, develop their own boards and chips, etc., etc.

To be fair, Google does already produce their own NN chips and boards.

 

[capster]

So a path to prevention of financial success for Tesla is to prevent legislation that allows autonomous cars. This may be do-able for years. And it's a good fit with populist politics.

Wouldn't vehicle autonomy be as desirable or even more so than, say, oral sex for many?

So and speaking of democracy, it's going to be a real challenge for governators to deny to consumers/constituents.

BTW, market insistence is one of the forces I hinted at some time ago as to why regulatory approval will come (much) sooner rather than later (assuming Tesla has the actuarial data which it will undoubtedly).

 

[Artful Dodger]

I will never understand how this crap works EVERY SINGLE TIME! One VERY suspect clip conveniently comes out hours before what is sure to be a positive event for Tesla and the share price tanks. EVERY SINGLE TIME! Will it rebound after the autonomous event? Probably...a little, but the net effect will most likely be a red day. I just don't get it. How a stock can be so effected by negative news (even if it suspect or proven false) but yet oblivious to any positive news?

This, my friends, is why I am without a doubt clueless when it comes to the market and quite frankly, why most people don't mess with it (which is just the way the big boys want it I think).

Dan

Yeah, it's really not hard to understand. Just hard to fathom. The shortz just prefer having an excuse to pile in. But its no secret what they're doing:

Aggregate Short Volume Data Reported to FINRA: 2019-04-22

 

[SOULPEDL]

this does not make sense to me. If they really can run robotaxi, why sell those cars? Don't tell me about sharing the profit with consumers, those cars will be gobble up by big investors with deep pockets, ordinary Joes won't be able to buy them.

Why if you lease they want the car back.