I don't see it as a "lie" at all. I see it as a demonstration that Tesla tried to make look as realistic as possible. Maybe there was a green screen involved. Maybe it's almost entirely CGI. Maybe it's mostly or completely real video. We'll know for sure which at some point, I'm certain.
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Tesla, TSLA & the Investment World: the Perpetual Investors' Roundtable
- Thread starter AudubonB
- Start date
lafrisbee
Active Member
I agree with much of what you state. But I agree with the "Nikola" factor" of if it was CGI then this is deception, and it was presented at an INVESTOR EVENT. So Elon would get exactly what Trevor(?) got from the FEDS, But multiply the Financial Penalty several times over. And the damage to Tesla/TSLA would be off the charts.... The bears would die from gluttony.
But The video is too artsy. Some is in focus, and some is out of focus. Some of it is not real. It does create a false image, but not false facts about the product.
My resolution to it all is that What the Robots do is real. The video might have some scenes actually performed by Tesla bots, and anything the Tesla bot is shown to do it can indeed do regardless of whether an actual bot was used in the video. But most of the scene is CGI, manipulated pixels, or distortion done in an attempt to make it appear more appealing than a sterile video in a lab of bot stumbling around holding his tool.
petit_bateau
Active Member
US-IRA has also created a weird market in Europe. One where the 4680 equipment gets unbolted from Berlin and sent to Austin. And where as a result Berlin keeps sucking on LFP out of China. At this rate they'll be switching Fremont to 4680, and redirecting Fremont's ex-China LFP over to further ramp Berlin.
If I recall correctly Tesla still hasn't drawn on any of the German factory grants. They were conditional on battery manufacture in Berlin. Hard to see when Berlin is actually going to be able to get its hands on any 4680 kit to execute on that.
Since Mexico can act as a cell factory supplying into the US-IRA market then it is also likely that Monterrey will be ranked ahead of Berlin for 4680 equipment. A bit ironic that Berlin can't get 4680 equipment even though (??) a fair chunk of it is likely made in Germany.
The EU is figuring out the details on how to respond to US-IRA. I rather suspect it will have a solution for this in the corresponding EU package.
So the puzzle pieces may get shuffled again.
Oh no it's a straight on lie because Elon talked about the demo as if it's a real demo. First thing he said was "few months ago optimus couldn't even walk" as optimus was walking in the demo. We have seen optimus walk in CGI simulation from AI day.
How does that let a 800v+ Tesla i.e cybertuck charge at existing super chargers that are 400v? You wouldn't be able to charge anywhere that hadn't been upgraded yet, and while it might be cheaper in the long run to upgrade all super chargers to 800v+ capable than to make all high voltage Tesla's dual voltage capable, it would take years realistically to upgrade them (trained labor being the bottleneck), meanwhile you can't drive your shiny new Tesla anywhere beyond it's range.
If the cybertuck is higher voltage than existing superchargers, then they must choose either a DC to DC voltage doubler or ability to split the pack for series rather than parallel operation.
In that case, being able to split the pack is clearly superior, unless they can make a full power DC to DC doubler for really really cheap. I can't imagine a 300+ kW voltage doubler being cheaper or more mass efficient than a few more contactors and extra pyro fuse.
On the other hand, it would be neat if you had a DC to DC stage that was bidirectional and had a wide range of operation. Would provide a path to a interoperable v2h for those that want it, charging other vehicles that are stranded, or perhaps aren't cars (charge your e boat from your truck?), or hook up a string of DC output solar panels directly to the truck... But those are unreasonable features to expect.
Some more people maybe cancelling because the 800v architecture doesn't allow them to charge beyond 50kwh at a Tesla magic dock supercharger station.
Yes, 65% compared to a 50% average IIRC. Historical data. What else was there to buy? A Bolt? Tesla is the only volume EV producer and most people have a great experience.
But how many times am I rebuying a model Y? I do have a CT preorder, which I may take if it has a 400+ mile option and some form of power out like the Lightning. But the large majority of somewhat expensive vehicle buyers are not CT buyers. Then look at the low sales of the S/X over the last few years. Some more savvy S/X buyers are waiting for HW4, but the neighbors are no longer impressed with the falcon wing doors. The second CT is almost certainly a three row SUV, and that will probably sell well. But increments of a few hundred thouseand vehicles doesn't contribute much to reaching 20 million.
Musk may want buyers to fall in line and accept his idea of commoditized vehicle categories, but that not going to happen. Regardless of whether senior managers agree with musk to his face they likely know how they will build a minivan on the 3/Y platform.
From an investment perspective I view TSLA as undervalued as I expect CT this year and then buildup to the excitement of gen 3 launching. When Tesla is properly feeding the zeitgeist the stock rises. If Tesla drags on building out Mexico I will sell after CT pumps the stock
That's just like your opinion, man.
That's not at all what I took away from his comments. I think he intentionally left it open for interpretation. I would call it slightly misleading at worst. Anyhow, I think we've beat this topic enough for the time being.This makes sense to me. I have read here in the past few days that the plan is to have Giga Monterrey built in about 12 months. I also read that a Tesla executive said it will be producing Gen 3 vehicles in 18-24 months. What I have not read is, what will the factory be producing in months 12-18?
insaneoctane
Well-Known Member
While the videos show the competition mowing over pedestrians / cyclists while Tesla appears to stop and avoid them all, the VRU green numeric test scores in the corners of the videos hardly discern between the vehicles at all. I think Tesla's VRU score was an 82% and I think the Mercedes MQS that plowed through pedestrians / cyclists showed an 80%???? What's up with the VRU scores for the NCAP EAB?
insaneoctane
Well-Known Member
^^This^^
The best part is no part.
The problem is not with the 800v architecture per se, but rather the electronics on the cars equipped with 800v architectures. For instance, the 800v Porsche Taycan can pull 130 kW on a V3 supercharger.
In the case of the CT, yes. In the case of the Gen3 I might disagree with this.
Gen3 is going to be made in MUCH more numerous volumes than CT will, by 2030 they will likely be making about 10 million Gen3's per year. These cars are going to be everywhere on the roads, you won't be able to drive anywhere without seeing many Gen3's out and about.
If all Gen3's are simply bare stainless steel finish then roads are going to get very monotonous looking. With a car so numerous it would be very beneficial to have some diversity in how they look, and colors are the best way to do that.
I do like the notion they won't be painting the Gen3 body panels but rather wrapping them right after stamping. Wrapping can be automated, is much cheaper to apply than paint, and would allow for a wider variety of color choices. It even makes it much easier to change the colors up yearly.
Stainless stamped body panels which are factory wrapped is what I believe the Gen3 will ship with.
Yeah, split pack offers the faster charge, based on current infrastructure. That may not be the case once CT is rolling out in volume. If V3 cabinets use multiple modules per pedestal, 1000V could be fairly easy to retrofit. They already run dual independent feeds to the post so it may be super trivial and done in the CT pedestal vs cabinet.
While it would be power limited at 400V chargers without a split pack, needing vehicle specific chargers for full performance has been a Tesla thing in the past.
In the meantime, while much lower power, the existing on board AC to DC charger can also operate from DC. Since the Cybertruck is higher Wh/mile, it would follow that the on board charger would be higher rate than current vehicles (still loeer power Level 2 rates vs Level 3).
Going way old school (Roadster), they could use the drive unit inverters as chargers; however spilt pack likely wins over that.
And I agree with much of what you say- except the deception part. I think it would be extremely difficult/nearly impossible to make a good case in court that the demo (assuming CGI/animation) constituted deception anything near the order of the Nikola pinewood derby event.
insaneoctane
Well-Known Member
This graphic was impressive. Retail investors are buying TSLA at 3x the rate of S&P500 index!
The discussion about the Optimus demo reminds me of the Moon landing argument that says it would be more work to fake it than just do it.
In other words, the effort it would have taken to make this level of "as if real" simulation would be insane.
Note the tool: either they put red tape over a Milwaukee driver, or they downloaded a high res 3D model of the driver then chose to model tape over the logos instead of deleting the logo from the model.
Contrast with the Ingersoll Rand vertical support where they would need to download a model with logo and choose not to obscure it.
Then there is the portrait throwback to Airplane where the video has the flash of the camera shot that is then in the frame (as a taped on printout, not properly matted).
Plus blinkin' LEDs.
In other words, the effort it would have taken to make this level of "as if real" simulation would be insane.
Note the tool: either they put red tape over a Milwaukee driver, or they downloaded a high res 3D model of the driver then chose to model tape over the logos instead of deleting the logo from the model.
Contrast with the Ingersoll Rand vertical support where they would need to download a model with logo and choose not to obscure it.
Then there is the portrait throwback to Airplane where the video has the flash of the camera shot that is then in the frame (as a taped on printout, not properly matted).
Plus blinkin' LEDs.
Whether Tesla can sell 20m cars by 2030 reminds me of the goal to have robotaxi's available this year. A great goal but not happening.
If Tesla produces more cars then Toyota (10.5m) that would still be a success IMO. It will very challenging to produce 5m cars by 2025 let alone another 15m cars annually on top of that just 5 years later. None of this though should stop Tesla from trying to accomplish what seems to most an impossible goal. Along the way Tesla will make great things happen no matter if they make the goal or not.
If Tesla produces more cars then Toyota (10.5m) that would still be a success IMO. It will very challenging to produce 5m cars by 2025 let alone another 15m cars annually on top of that just 5 years later. None of this though should stop Tesla from trying to accomplish what seems to most an impossible goal. Along the way Tesla will make great things happen no matter if they make the goal or not.
Knightshade
Well-Known Member
In 1969 that was certainly true.
Today making such a video would be trivial on a reasonably powerful home computer- let alone by a company with one of the largest GPU clusters in the world.
None of which presents any opinion on how real the demo shown at investor day was-- but the objection a fake video would be hard with modern CGI and video editing capabilities simply ain't so.
A third possibility is the footage is all real footage, but it's lots of different takes edited smoothly together to look like a single working-pretty-well take... we know for a fact Tesla did that in the 2016 FSD demo, and such editing is even more trivially done than CGI robots.
I'm not sure any of this really matters much investor-wise.... much like robotaxi, or any "Nobody has really gotten this to work in the general case at all" thing- I'm not going to care a whole lot until there's a real product ready to mass produce doing real work for real money. Prototypes, I've heard, are easy.
Last edited:
Similar threads
- Locked
- Locked
- Locked
