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

[kbM3]

It's really not.

They ran out of compute on one node in mid-2020 and have had to spill over to use the second node ever since. See the full thread here for where Green mentions this has been the case since then and not getting better.

And that's a system VASTLY less capable than the L4 system promised to pre 3/19 buyers.

So the idea they'll somehow magically add a ton of functionality that currently does not exist at all and ALSO reduce required compute by at least 50% to fit back in a single node (since you need redundancy as per AI day presentation to run without a human in the car) is magical thinking unsupported by any evidence.

https://twitter.com/x/status/1409299851028860931

<citation required>

And ideally a citation that has more substance than "they'll somehow just make everything way more efficient *waves hands*"

Again, nobody knows how much compute is actually needed for L4 driving on a Tesla. Not Elon, and certainly not James Douma who AFAIK has even less visibility to Tesla code than Green does....and that will be true of everyone until they actually achieve it. Which they have not.





Douma is the guy who told us how they didn't need LIDAR but that RADAR was super useful to FSD and it totally made sense they used it.... right up until Tesla announced they're removing radar... at which point he did a Dave Lee video telling us how no, just kidding, Radar wasn't really super useful at all and it totally makes sense to remove it.

But we do know they're WAY past using all the compute available in a single HW3 node, even at L2, and have been for roughly 3 years now.

In fact, hilariously, Douma is one of the original sources for HW3 having run out of compute years ago:

https://twitter.com/x/status/1409527523231285249

Like I said, watch his videos.

1. You have zero evidence they “had to” use the 2nd processor. There’s tons of ancillary uses and capabilities they can be running on the 2nd processor. They wisely CHOSE to use processing on both chips. it would be dumb to let that processing power go to waste. If you do have such evidence that they were forced to, please produce it.

2. You also probably don’t have evidence the original plan was to have absolute full redundancy of every feature rather than a less capable safety fallback. If you do have evidence of that being the original plan, please link that as well.

 

[kbM3]

It's really not.

They ran out of compute on one node in mid-2020 and have had to spill over to use the second node ever since. See the full thread here for where Green mentions this has been the case since then and not getting better.

And that's a system VASTLY less capable than the L4 system promised to pre 3/19 buyers.

So the idea they'll somehow magically add a ton of functionality that currently does not exist at all and ALSO reduce required compute by at least 50% to fit back in a single node (since you need redundancy as per AI day presentation to run without a human in the car) is magical thinking unsupported by any evidence.

https://twitter.com/x/status/1409299851028860931

<citation required>

And ideally a citation that has more substance than "they'll somehow just make everything way more efficient *waves hands*"

Again, nobody knows how much compute is actually needed for L4 driving on a Tesla. Not Elon, and certainly not James Douma who AFAIK has even less visibility to Tesla code than Green does....and that will be true of everyone until they actually achieve it. Which they have not.





Douma is the guy who told us how they didn't need LIDAR but that RADAR was super useful to FSD and it totally made sense they used it.... right up until Tesla announced they're removing radar... at which point he did a Dave Lee video telling us how no, just kidding, Radar wasn't really super useful at all and it totally makes sense to remove it.

But we do know they're WAY past using all the compute available in a single HW3 node, even at L2, and have been for roughly 3 years now.

In fact, hilariously, Douma is one of the original sources for HW3 having run out of compute years ago:

https://twitter.com/x/status/1409527523231285249

James explained why LOW RESOLUTION radar was harming them.

Once again. Get off that useless FSD thread and pay attention to someone who’s extremely capable and brilliant.

 

[Atomsla]

Fellow investors: Tesla needs your help here.
When successful we will all benefit from this help with respect to the TSLA stock price.

As you may know, there is only an enormously crippled version of FSD here, due to EU regulations.
However, Tesla owners here should be able to obtain the same FSD Beta version as the USA and Canada.
In order to get this discussed in the European Parliament, a petition for this must have 15,000 electronic signatures.
So, let’s go global and please help Tesla/TSLA and sign this petition for FSD Beta AND share the petition on online platforms.
We are almost there (14,446 till date)!

And an extra push in motivation to sign and get this enabled here in Europe too, might be given by the safety results of FSD Beta:

Tesla Reveals FSD Beta Accident Rate for First Time; Compares It to Autopilot and National Average

Tesla's recent Impact Report presents new data on the safety performance of its Full Self Driving (FSD) Beta and Autopilot systems, emphasizing the company's de

www.notateslaapp.com

This is a four year old petition that I'm not sure makes sense at this point, given the upcoming regulation changes:

Tesla's FSD Beta Could Be Unleashed in Europe Sooner Than Expected

Tesla's Full Self-Driving (FSD) Beta could become available in Europe as early as January 2024, thanks to the European Commission's accelerated timeline for the

www.notateslaapp.com

The European Commission intends to prioritize the United Nations Economic Commission for Europe (UNECE) Regulation on Driver Control Assistance Systems (DCAS), which could help speed up the arrival of Tesla's FSD Beta in Europe or even allow Tesla to re-enable some Autopilot features that have been removed in parts of Europe.


According to a recent tweet by @elektrotimmy, the final draft for the DCAS was originally expected to be completed by 2024, with approval for FSD potentially coming in January 2025. However, the European Commission has now decided to give the DCAS the highest priority and aims to complete the regulation this year. The final draft is expected to be ready as early as September 2023.

Expedited Timeline May See FSD Beta in Europe by January 2024​

This change in priority means that FSD Beta could become available in Europe at the earliest by January 2024, and at the latest by the summer of 2024. The acceleration of the timeline will likely be welcome news for Tesla owners and enthusiasts in Europe who have been eagerly awaiting FSD capabilities.


To further speed up the process, the European Commission plans to separate hands-on (Phase 1) from hands-off (Phase 2) functionalities. This means that even if hands-off is not yet allowed in the United States by that time, only hands-on will be permitted in Europe until Phase 2 is completed. Additionally, discussions are underway to determine if a closed testing phase can be allowed before the January 2024 approval date.

This is Speculative, but Positive News​

It is important to note, as @elektrotimmy emphasizes, that the accelerated timeline for the UNECE Regulation on DCAS is not a guarantee that Tesla will roll out FSD Beta across Europe at that time. However, it will be allowed for Tesla from then on, and it is expected that the company will not wait to make the beta available to European customers.


While the expedited timeline is promising, there is still a possibility of unexpected delays. Working groups and committees meet only on specific dates, and a significant delay could occur if documents are not ready by the meeting date. Nonetheless, this development signifies a major step forward for Tesla's Full Self-Driving technology in Europe and, potentially, a faster rollout for European drivers.

 

[Causalien]

I thought you were quoting Gordon Johnson for a second..

Ya, I don't know who that is. I come from a time when only 2 analyst covered tsla. Andrea James and Adam Jonas. All the weed that popped out later dissappeared just as fast.

 

[Krugerrand]

I also think that many of those whining about Elon "blathering" about autonomy on the quarterly calls now, will in the future be complaining—because they are out of position—when the discontinuity due to autonomy hits the stock price.

"But..but...but..he didn’t tell us."

"Why yes. Yes he did. On pretty much every opportunity and every call."

You trying to get me in trouble again? I can do it quite well on my own without anyone throwing me a bone.

 

[Knightshade]

James explained why LOW RESOLUTION radar was harming them.

yes... mere weeks after he'd told us how it was beneficial because it can see in situations vision can not (he specifically cited fog as one example)

The only thing that had changed was tesla announced radar bad, so James changes his story to match them.

Once again. Get off that useless FSD thread and pay attention to someone who’s extremely capable and brilliant.

This is the appeal to authority fallacy.

Especially weird when, as I cited, he's one of the original sources that proved Tesla had run out of HW3 compute in the first place.

I cited the source for that.

YOU claimed you had a source saying otherwise from him- but have been unable to actually provide it.

Like I said, watch his videos.

I'd love to- except you said you "couldn't remember" which one he made the claim you say he did- so how can I?

If you can't even recall where you heard it, it's entirely possible you also don't correctly remember what you heard.

1. You have zero evidence they “had to” use the 2nd processor.

Of course I do. I posted it.

Hilariously one of the sources was James Doumas own findings.

Since I guess you didn't bother to read it here's the exact words

@jamesdouma profiled the NNs and they are too big to run at full fps on a single NPU

They HAD TO use the 2nd processor because the NNs had grown too large to run otherwise.

Per YOUR "expert" on the topic.

2. You also probably don’t have evidence the original plan was to have absolute full redundancy of every feature rather than a less capable safety fallback. If you do have evidence of that being the original plan, please link that as well.

Of course I have such evidence. Again I previously cited it as being called out by tesla on ai day as one of the architectural goals of the FSD computers entire design

You have Jim Keller on stage stating the intent is for both chips to recieve and process all of the video from the cameras entirely independent from each other....right after Elon explains the whole point is redundancy so one entire node can fail and the car is still fine to drive.

Keller goes on to say the two indepedent chips would then design a plan for driving from the input, then compare those independent plans to insure they match as a sanity check... which again requires full stack running redundantly on each node

All of this is in the roughly 8-10 minute range in the AI day video depending whose copy you use.



You seem deeply determined to ignore all actual evidence and just make up your own narrative that ignores all the available facts.

it'd be great if you could at least do so in the correct place- which is:

AI, Autopilot, & Autonomous/FSD

Discussion about AI, Tesla Bot, Tesla Autopilot (AP), the promise of Full Self Driving (FSD), as well as other Autonomous Vehicles.

teslamotorsclub.com

 

[Grifon]

Sorry, but sharing your IP (to improve your own ecosystem) doesn't equalize breaking the terms of use for other IP you're using yourself.

It's a fact that Tesla (like many other companies) is breaking the terms of open source software.
Doesn't really matter who's pointing it out...

AFAIK they would have to do just two basic things to comply:
- Provide a list of all the open source software used in the car together with their licenses
- upon request provide all modifications to those applications.

I can't see any reason not to provide the first one.
And appart from security and HW specific changes (which might not even be affected) there's little reason to ignore the second...


It's essentially a more forgiving version of Teslas own fair use policy.

Regarding documentation of the use of open source software,Tesla does have an acknowledgement link within the Tesla app under the profile section (at the very bottom). Not sure if this is for the app only, but it’s a very long list, so it does look like Tesla has put some effort into recognizing open source contributors.

 

[Maarten]

Tesla's FSD Beta Could Be Unleashed in Europe Sooner Than Expected

Tesla's Full Self-Driving (FSD) Beta could become available in Europe as early as January 2024, thanks to the European Commission's accelerated timeline for the

www.notateslaapp.com

Yes, of course it has taken quite some time to get to this point with the petition.
And we could just wait and see what happens with a new projected timeline for Regulation on Driver Control Assistance Systems (DCAS).

But a very true quote from the article you link to:
"While the expedited timeline is promising, there is still a possibility of unexpected delays. Working groups and committees meet only on specific dates, and a significant delay could occur if documents are not ready by the meeting date."
Let us not be naive: old auto will not stop to grab any opportunity to slow things down for Tesla; these companies are in a death struggle.
So, better to not sit and wait what ‘promising timeline’ will happen or not, but let’s bring it to the European Parliament with 15,000 signatures.

 

[Curt Renz]

Mesbahi - this afternoon:

 

[wipster]

MODS, please move this FSD argument to another thread, the weekend is over.

 

[Chunky Jr.]

I don't know much about options, but this seems like a good thing to me...

$TSLA's option chain is DEAD! Implied volatility is 43%, easily 10-20% lower than it has been historically and what would be consistent with realized volatility.There is NO manic and delusional call buying. No bubble. Stock is simply grinding higher in a very healthy manner.

source :

https://twitter.com/x/status/1683576078747725824

 

[2daMoon]

Okay, this is hilarious. (Elon did say they might try advertising...)

Tesla appears to have wrapped a Cybertruck to look like an F-150 Lightning

Tesla appears to have wrapped a Cybertruck to look like a Ford F-150 Lightning for some strange reason.

www.teslarati.com

Look at the wrapped faux grill with headlights.

Hoping to see photos of this in the wild soon.

 

[ByeByeJohnny]

a'

The only thing that had changed was tesla announced radar any change bad, so James changes his story to match them.

Why be surprised? A lot of folks here does the same.

 

[KBF]

Okay, this is hilarious. (Elon did say they might try advertising...)

Tesla appears to have wrapped a Cybertruck to look like an F-150 Lightning

Tesla appears to have wrapped a Cybertruck to look like a Ford F-150 Lightning for some strange reason.

www.teslarati.com

Look at the wrapped faux grill with headlights.

Hoping to see photos of this in the wild soon.

https://twitter.com/x/status/1683552185454297088

Video as well.
Edit: I do think of this as marketing - create some controversy/noise/attention; it gets headlines for essentially free... Tesla has mastered this approach. The haters can't resist trying to bash it (generating attention), and the fans eat it up. It's amusing to me...

Good to see we had a green day - I kind of expect a full recovery relatively soon since the Q2 numbers were really fine (the casino just needed to make sure they didn't lose too much money on Friday - and scare day traders into selling)

 

[Gigapress]

It's really not.

They ran out of compute on one node in mid-2020 and have had to spill over to use the second node ever since. See the full thread here for where Green mentions this has been the case since then and not getting better.

And that's a system VASTLY less capable than the L4 system promised to pre 3/19 buyers.

So the idea they'll somehow magically add a ton of functionality that currently does not exist at all and ALSO reduce required compute by at least 50% to fit back in a single node (since you need redundancy as per AI day presentation to run without a human in the car) is magical thinking unsupported by any evidence.

https://twitter.com/x/status/1409299851028860931

<citation required>

And ideally a citation that has more substance than "they'll somehow just make everything way more efficient *waves hands*"

Again, nobody knows how much compute is actually needed for L4 driving on a Tesla. Not Elon, and certainly not James Douma who AFAIK has even less visibility to Tesla code than Green does....and that will be true of everyone until they actually achieve it. Which they have not.





Douma is the guy who told us how they didn't need LIDAR but that RADAR was super useful to FSD and it totally made sense they used it.... right up until Tesla announced they're removing radar... at which point he did a Dave Lee video telling us how no, just kidding, Radar wasn't really super useful at all and it totally makes sense to remove it.

But we do know they're WAY past using all the compute available in a single HW3 node, even at L2, and have been for roughly 3 years now.

In fact, hilariously, Douma is one of the original sources for HW3 having run out of compute years ago:

https://twitter.com/x/status/1409527523231285249

For what it's worth, our former member Discoducky has told me that James Douma knows what he's talking about when it comes to Tesla's AI approach. Discoducky actually worked on the Autopilot software development team in 2014-2015, had weekly meetings with Elon and worked with Ashok Elluswamy to build Tesla's AI team. He has also told me that Green often thinks he knows more than he actually does. James probably gets stuff wrong sometimes and has biases and blind spots like everyone else, but in general I am personally inclined to give him credibility.

I believe @kbM3 's rebuttals were valid as well. On what basis can we confidently conclude that necessity is the reason that both nodes of the HW3 computer are being used to run the net? With FSD Beta being a level 2 ADAS that still requires active human oversight, is computer redundancy even a priority right now? In the event of a core failure, the driver should be the 2nd layer of protection.

Tesla have been redesigning the neural net architecture frequently and it would not be surprising if they were deliberately allowing bloat to exist in order to save engineering time and training compute so as to speed up iteration cycles. Premature optimization is the root of all evil. It is a fact that neural nets can be shrunk with optimization, but how much FSD can be compressed is uncertain. Considering that none of us here are working for Tesla AI, we are left with no option but handwaving about the possibility of squeezing a future level 4 or 5 version to fit in a single HW3 node.

 

[space_s3x]

 

[FutureCTOwner]

A little off topic, but I have found Randy Kirk on youtube to be an incredibly useful resource regarding tesla as he takes a more macro approach and how it relates to Teslas shareprice. Since I have been following him for the last few months, he has not missed a beat. Rob Maurer is laser focussed on Tesla, Randy Kirk offers a broader perspective and together they been incredibly useful combination for the overall Tesla perspective.

 

[Patrick66]

I have to say today's price movement in TSLA was surprising for me. Welcome, but surprising. Any explanation out there that I am just not getting? The news is all neutral to negative at the core, and the macro isn't responsible...

 

[winfield100]

I have to say today's price movement in TSLA was surprising for me. Welcome, but surprising. Any explanation out there that I am just not getting? The news is all neutral to negative at the core, and the macro isn't responsible...

@Patrick66
I would respectfully suggest you read Papa Fox's forum daily for cogent information

Papafox's Daily TSLA Trading Charts

TSLA chart above QQQ chart above On Monday I think we did see Cybertruck production give TSLA a boost as the stock gained 3.20% compared to the NASDAQ's gains of 0.93%. Unexpected good news for investors was a settlement by TSLA BOD members that returns $735M of compensation to the company...

teslamotorsclub.com

 

[ZeApelido]

A little off topic, but I have found Randy Kirk on youtube to be an incredibly useful resource regarding tesla as he takes a more macro approach and how it relates to Teslas shareprice. Since I have been following him for the last few months, he has not missed a beat. Rob Maurer is laser focussed on Tesla, Randy Kirk offers a broader perspective and together they been incredibly useful combination for the overall Tesla perspective.

Randy Kirk once interviewed me, so that's 3 strikes against him.