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

[Remus]

So not surprisingly some believes the FSD is near and some don't. But most agree if it is really near then it's yuge.

I doubt the market would move though

 

[HG Wells]

When bureaucrats and politicians replace engineers, the outcome is obvious.

When engineers and scientists replace politicians, the out come is obvious.

Vote rational, vote technocrat !

 

[davecolene0606]

GF1 is not Automotive for Panasonic, it's Energy.

Ya, so there is a HW3 (aka APH4, aka FSD computer)

I, for one, will keep using HW3 as it is a lot faster to type than FSD computer (since FSD alone is ambiguous). Then you have HW2 + FSD computer upgrade as another case where the sensor HW is different (2.5 also changed radar suppliers).

Who’s the new radar supplier? No more Bosch?

Fire Away!

 

[Renew]

So Elon is ~100% confident on his approach towards autonomy but why is Google/Waymo following a different approach? The answer, I think, lies in the genesis of their efforts towards autonomy.

Elon's approach:

• Elon said he felt he needed to pursue autonomy because in his mind non-autonomous vehicle would be less valuable than autonomous vehicle in the future. So not to pursue it may even defeat his primary mission of electrification.
• Once he realized the need to pursue it, as usual, he followed first principles thinking.
• Broke down the problem and started solving for each individual lego piece the best way possible -
• Collecting huge data (the very basic requirement for NN training) by putting in required hardware from the very start on every vehicle even if the customer decides to not pay for AP/FSD,
• Developing a very specialized NN chip for over 3 years,
  
• Hiring the best AI researchers to build the algos (probably the easiest decision).

Google's approach:

• On the other hand, I think Google went into autonomous vehicles primarily because of one of their expertise - MAPS.
• Google probably thought because they have spent so much of their resources on building out the maps of the entire world - what next, what other ways to montize this expertise?
• Answer - build even higher resolution maps and leverage their own AI researchers (also best in the industry) to work on rest of the stuff to pursue autonomy.
• I think they will achive autonomy but it will be stuck in Geofencing and it's just not easily scalable.
• Also not to mention the LIDAR approach which will help them to achieve local maxima, but still really have to solve vision to really grind towards global maxima.
• I think the partial reason to select LIDAR was because they think solving vision was very difficult - primarily becuase of unavailability of huge amounts of data (including weird edge cases that simulation just can't predict and only real world data can provide).

I am just guessing if Google had started on automomy from a blank sheet of paper, they would have followed a similar approach as Elon's (maybe by partnering up with an auto manufacturer to setup the hardware to collect the real world data).

 

[MP3Mike]

Who’s the new radar supplier? No more Bosch?

Continental Automotive.

 

[ZachF]

He’s trying to hype and get more people to buy FSD now. It makes sense. If April 22nd shows us that level 5 is possible soon, then the feature should cost $20+k. Otherwise it is more profitable for Tesla to build the cars for themselves and have them drive people and keep 100% of profit. “Demand problem” solved too. This of course assuming that FSD exists, big assumption and I have my expectations very low

Ditto to all of this.

FSD, if it works, is a giant lever that would allow one car to do the work of many. 500k cars per year can replace the equivalent passenger miles of 2-5 million cars per year.

A car that can charge 0.25 per mile will make $125,000 over a 500k mile lifetime, with electricity only costing about $15-20k.

Of course this is all with a big fat if..

 

[Singer3000]

I for one ain’t letting it get within a block of me until I live-view the inside via the hailing app.

It ain’t rocket surgery (that was last week).

I wonder what is a more complex problem for a neural net. Driving itself across busy intersections with broken traffic lights, drunk jaywalkers and safely around horseback riders etc...? Or using the internal camera to see if somebody left a pile of empty red bull cans and vomit on the back seat?

 

[Singer3000]

Isn't this essentially bad news? I read it as pretty serious constraints on growth in the short term. Potentially even constraints on Tesla Energy when you have to buy from outside suppliers.

I wonder if this is why Tesla has to source Chinese production locally versus making batteries at GF3.

Maybe Tesla should cut out Panasonic entirely at some point and build their own batteries.

This is kind of fascinating and I need to revisit some of my own assumptions. Everyone went wild about the unplanned pull forward for the SR launch date, assuming this was a demand issue. But quite possibly it’s because Pana haven’t been holding up their end, so Tesla need to use the limited number of cells more intelligently. This doesn’t explain Model SX of course.

 

[jerry33]

I wonder what is a more complex problem for a neural net. Driving itself across busy intersections with broken traffic lights, drunk jaywalkers and safely around horseback riders etc...? Or using the internal camera to see if somebody left a pile of empty red bull cans and vomit on the back seat?

The bigger problem is how to prevent the second scenario from taking place. James Bond style ejection seat perhaps?
I'm thinking that infra-red sensors would go a long way to helping with the first.

 

[Todd Burch]

I wonder what is a more complex problem for a neural net. Driving itself across busy intersections with broken traffic lights, drunk jaywalkers and safely around horseback riders etc...? Or using the internal camera to see if somebody left a pile of empty red bull cans and vomit on the back seat?

Karpathy’s earliest research paper was entitled “On the Evolution of Convolutional Neural Image Networks of Stomach Bile”.

It’s now been open-sourced as VomitNet

 

[jerry33]

This is kind of fascinating and I need to revisit some of my own assumptions. Everyone went wild about the unplanned pull forward for the SR launch date, assuming this was a demand issue. But quite possibly it’s because Pana haven’t been holding up their end, so Tesla need to use the limited number of cells more intelligently. This doesn’t explain Model SX of course.

S/X is fairly easy. They removed the popular 75 kWh option so as not to cut into M3 sales. My understanding is that most of the shortfall was with the S not with the X.

 

[Krugerrand]

“Buying a car today is an investment into the future. I think the most profound thing is that if you buy a Tesla today, I believe you are buying an appreciating asset – not a depreciating asset.”​

Why foot in mouth?

Go big or go home. You’ve no more idea than us what they’ve really got. It might be you chewing on your foot when all is said and done.

 

[heltok]

Pana is bad news for short term production but good news for future demand situation. Lets say they can go from 24GWh to 35GWh in 6months, that is a 50% increase in revenue with minimal capital expenditure.

I also assume they had a bit extra supply of cells in Q3-Q4 thanks to flufferbut bottleneck Q1-Q2.

 

[K3jSe7Llp7]

Do I understand correctly that the reduced numbers in q1 was due to shortage of cells? I don't have a handle on cells per car for the various models and all like some of you do. But didn't Elon just say that that's the constraint?

 

[StarFoxisDown!]

I will say just from the mood and assumptions from practically all the media and message boards over this weekend, especially today with elons tweet.....if they do actually show off full FSD, it will shock everyone. Literally no one besides a few people here actually think they can do it. Really hope Elons not hyping up just like 1 or 2 new features.

 

[mongo]

Karpathy’s earliest research paper was entitled “On the Evolution of Convolutional Neural Image Networks of Stomach Bile”.

It’s now been open-sourced as VomitNet

Spinal Tap will be very interested to hear that since you can't dust for vomit.

 

[davecolene0606]

OT




Falcon Heavy is awesome, however that payload capability doesn't look quite right to me.


Weight Of Boeing 747 | What Things Weigh

A Boeing 747-400 will weigh 403,500 pounds (183,024 kg) when empty. Fully loaded, it weighs around 875,000 pounds (396,893 kg).

Capabilities & Services Falcon Heavy. Low earth orbit 140,660 lbs

Fuel vs payload

Fire Away!

 

[VValleyEV]

Fuel vs payload

Fire Away!

Yup was wrong, FH can lift a 737, not a 747.

My confusion was that in the very next paragraph on SpaceX website it says that FH thrust is equivalent to eighteen 747s. So in my haste I jumbled the two together. My bad. Too late to edit the original post, although I suppose I could remove it and repost it.

 

[Pras]

I will tape my fingers crossed till Apr 22nd!

If they really pull this off.... it will be like sugar!

 

[Dan Detweiler]

I will say just from the mood and assumptions from practically all the media and message boards over this weekend, especially today with elons tweet.....if they do actually show off full FSD, it will shock everyone. Literally no one besides a few people here actually think they can do it. Really hope Elons not hyping up just like 1 or 2 new features.

Why not get excited about one or two big features? If they come out and demonstrate Nav On Autopilot type functionality on surface streets that would be huge. Why do we all want to set ourselves up for disappointment? We, as bulls, should be excited about any major development. None of the competition seems to be doing anything significant. Let the bears do all the unrealistic expectations. That's what they're good at.

Dan