Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

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

This site may earn commission on affiliate links.
It takes more than traditional smarts and hard work to be exceptional. Her quoted words make it clear she does not practice first principles thinking very well. Her mistake was assuming that batteries need to be serviceable like an ICE engine. Yes, Tesla made this mistake too but I know that is not necessary and I don't even direct the technical direction of a multi-billion dollar company.

It's more important that an engineer knows what they don't know than it is for them to be smart and think they know things they don't.
Considering GM's battery reliability thus far, I think there's a strong case for designing with serviceability in mind. They have no other choice. Tesla doing that for the roadster and early S was a good call, not a mistake. The real mistake was taking too long to switch to a cell-to-pack architecture once reliability had improved. Elon said this in an interview a while back, mentioning that this kind of stuff has inertia and they neglected to revisit the premises behind their design rationale. In 2021, Tesla can get away with gluing the cells together and making them structurally integral because their cell reliability and BMS performance is so amazing. This just illustrates the massive advantage of Tesla over GM in pack design.
 
Last edited by a moderator:
UK Model Y order page is live, delivery early 2022
 

Attachments

  • Screenshot_20211015_185119_com.brave.browser.jpg
    Screenshot_20211015_185119_com.brave.browser.jpg
    133.6 KB · Views: 65
CyberTruck page updated, details and pricing about different trims removed from both the page and the pre order page, also you can no longer select FSD or lock in price for orders.

FSD pricing step changes coming, or turning to subscription only soon?

Let speculations begin!
 
So quick recap so far:

-Leo is planning to buy another 1M shares in Q4.
-UK has opened up Model Y orders.
-Texas Insurance is now live.
-FSD Beta for score of 99 and under in 2 weeks.
-Accountant & Rob Maurer -expecting > 2.00 /2.15 for earnings next week.
-even Gordon Johnson believes good results this quarter
-return of bullish sentiment overall in market
-covid cases dwindling back down
-tsla apparently at 91% from ATH

Any reason I shouldn’t pick up more shares today at the usual 10 o’clock dip?
 
CyberTruck page updated, details including pricing about different trims removed from both the page and the pre order page, also you can no longer select FSD or lock in price for pre orders.
FSD pricing step changes coming or turning to subscription only?

Let speculations begin!
Wow. No configuration at all. No single, dual, or tri motor. I'm glad I ordered a second one just days ago.
 
You both make it sound so easy to borrow $8 Billion cash. This isn’t a line of credit, this is cash that must be then transferred to the IRS and the California Dept of Revenue. EM has borrowed a few hundred million dollars against his shares so far from banks like Goldman, but $8 Billion is something else. I predict he sells shares to pay the tax bill.
there's also a SpaceX secondary transaction underway right now
 
So quick recap so far:

-Leo is planning to buy another 1M shares in Q4.
-UK has opened up Model Y orders.
-Texas Insurance is now live.
-FSD Beta for score of 99 and under in 2 weeks.
-Accountant & Rob Maurer -expecting > 2.00 /2.15 for earnings next week.
-even Gordon Johnson believes good results this quarter
-return of bullish sentiment overall in market
-covid cases dwindling back down
-tsla apparently at 91% from ATH

Any reason I shouldn’t pick up more shares today at the usual 10 o’clock dip?
Berlin and Austin starting production this Q
 
CyberTruck page updated, details and pricing about different trims removed from both the page and the pre order page, also you can no longer select FSD or lock in price for orders.

FSD pricing step changes coming, or turning to subscription only soon?

Let speculations begin!
I pre-ordered a second Cybertruck about a month ago wondering about the possibility of FSD pricing going up or just general pricing increases.

Of course the changes go beyond that. Can’t update existing configurations, they don’t even mention there are 3 configurations any more.
 
Definitely is a marketing effort but Vanderbilt is not financially desperate.

Vandy is one of the richest schools in the world with a $6+ billion endowment and a very wealthy alumni network. The pandemic hasn't helped but they're still doing fine.
Vanderbilt was literally founded with the wealth of railroad tycoon Cornelius Vanderbilt, one of the richest Americans who ever lived.

 
As someone who has been playing with FSD beta over the last few days, I would be interested in an explanation of how Tesla intends to overcome what appears (to me) to be a limitation with their FSD approach...

So, no dependency on hi-def maps using vision, right?

I had FSD take me on a route that required it drive out of a parking lot at a traffic light. This exit had 2 lanes on my side- the left lane went left and the right lane went straight or right. It was also at the top of a small incline at a traffic light. My route required being in the left lane to turn left. So FSD began to execute this maneuver, fairly slowly and turned right to begin the short drive up the incline where it needed to be in the left lane. Only it turned right into the right lane and then realized it needed to be in the left lane, so it cautiously started moving into the left lane, only it didn't have enough room before the light. So I sat at the red light kind of in both lanes.

So, here's the limitation I concluded: The car drove exactly like a first time driver that had never been at that intersection would have. But, once a "meat" driver learned that intersection, they would have learned how to handle it better and safer on subsequent drives. Their first encounter was probably riskier and not ideal but they would likely not perform the risky version on that intersection again. But FSD will be "seeing" that exact intersection for the first time...every time and likely making the rookie mistake because it's always seeing it for the first time. Tesla can't learn every unique intersection IMO, and I doubt it's poor performance would even get flagged as incorrect?

Am I missing something? How will Tesla learn the unique subtleties of the countless intersections across the world?
 
As someone who has been playing with FSD beta over the last few days, I would be interested in an explanation of how Tesla intends to overcome what appears (to me) to be a limitation with their FSD approach...

So, no dependency on hi-def maps using vision, right?

I had FSD take me on a route that required it drive out of a parking lot at a traffic light. This exit had 2 lanes on my side- the left lane went left and the right lane went straight or right. It was also at the top of a small incline at a traffic light. My route required being in the left lane to turn left. So FSD began to execute this maneuver, fairly slowly and turned right to begin the short drive up the incline where it needed to be in the left lane. Only it turned right into the right lane and then realized it needed to be in the left lane, so it cautiously started moving into the left lane, only it didn't have enough room before the light. So I sat at the red light kind of in both lanes.

So, here's the limitation I concluded: The car drove exactly like a first time driver that had never been at that intersection would have. But, once a "meat" driver learned that intersection, they would have learned how to handle it better and safer on subsequent drives. Their first encounter was probably riskier and not ideal but they would likely not perform the risky version on that intersection again. But FSD will be "seeing" that exact intersection for the first time...every time and likely making the rookie mistake because it's always seeing it for the first time. Tesla can't learn every unique intersection IMO, and I doubt it's poor performance would even get flagged as incorrect?

Am I missing something? How will Tesla learn the unique subtleties of the countless intersections across the world?
Dojo