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

[Mengy]

In my opinion there is zero chance of anything approaching RoboTaxi to be delivered in a generalized form within this decade. That will NOT happen, in my opinion.

Given the competency of FSD v12.3 and given the rate of progress lately, I have to admit I find this a silly notion. We still have almost six years for this decade, do you truly think FSD won't be solved to a high enough level to enable RT's within that timeframe? Honestly?

 

[GatorMeat]

Elon is known to be overly optimistic and plans change. What's your point?

"Already started DC to NY" So where is it if it's already started? That is a lie, not an optimistic statement

 

[skybluecgreen]

0 chance in the next decade for robotaxi? I'm definitely betting against that premise. Reading your entire post I can see why you've sold your tesla holdings as you see them as a failing auto company with software that will never achieve what Elon aspires it to become.

But sure call it hype. That's the beauty of investing and trading. My estimation of what Tesla will achieve is unbelievable to some people and in the end time will tell who was on the right side of the trade.

He said there is zero chance this decade, not in the next decade. 5.5 years vs 10.

 

[larmor]

Given the competency of FSD v12.3 and given the rate of progress lately, I have to admit I find this a silly notion. We still have almost six years for this decade, do you truly think FSD won't be solved to a high enough level to enable RT's within that timeframe? Honestly?

If I can train a 16 yo to drive a car, then an AI engine with cameras all over the car should be able to do this task and learn it much faster. If I get the AI engine to train on videos of people driving cars, then this should happen even faster.

AI can take a word phrase and make a movie clip that is close to a human made movie clip...

 

[GatorMeat]

The most of the 500 that were laid off on the supercharging team located potential sites and negotiated with the site owners. Local contractors including electrical engineers pulled permits and did the job site construction. Obviously Tesla supplied the charging equipment.

Locating and negotiating with site owners would also be better left up to third party pros. Commercial real estate brokerages who already have relationships and knowledge of all the potential sites would handle this task much more effectively

 

[Todd Burch]

That doesn't seem unreasonable to think that since Elon liked a post stating that Tesla is pivoting from selling cars to selling rides

Yes it does.

Are they just going to throw away tens of billions in infrastructure and capital investments?

Do you think he could just throw away one of Tesla’s core businesses without shareholder approval? All the technology they developed and in-housed? They spent two decades getting to where they are. Especially with other OEMs choosing to not take the reigns of EV production, I give almost zero chance they stop producing vehicles.

It seems quite unreasonable.

 

[larmor]

He said there is zero chance this decade, not in the next decade. 5.5 years vs 10.

The key to Sam Altman's TED talk from 10 years ago from the example of humans stepping on ants: using multiple AI engines can equal a team of medical researchers working for 2 plus years just by increasing the number of AI processors. So that 200 AI processors might complete the medical research task in 2 weeks...

 

[DarkandStormy]

Mildly OT
@DarkandStormy
a _casual_ =>glance<= at the graph shows the the =>top<= of the Y axis is ~10%, would not 2% then actually be 20% of 10%? 1/5th take rate seems more like it unless I have lost reading skills

=>(simpler, 2, 4, 6, 8, 10) is the total Y axis)<=

may i suggest Edward Tufte's book on "visual display of data..." the one with Charles Joseph Minard's graph of Napoleons march to Moscow & back on the cover

Take up your issue with Yipit Data.

 

[2daMoon]

The key to Sam Altman's TED talk from 10 years ago from the example of humans stepping on ants: using multiple AI engines can equal a team of medical researchers working for 2 plus years just by increasing the number of AI processors. So that 200 AI processors might complete the medical research task in 2 weeks...

We all know how long 2 weeks can be...

 

[DarkandStormy]

Anyone know someone who works at the factory in Texas? This allegation made in The Information should be easy to refute if it's not true.

 

[Chunky Jr.]

Yes it does.

Are they just going to throw away tens of billions in infrastructure and capital investments? All the technology they developed and in-housed? They spent two decades getting to where they are. Especially with other OEMs choosing to not take the reigns of EV production, I give almost zero chance they stop producing vehicles.

It seems quite unreasonable.

if you take what Elon liked at face value (pivoting from selling cars to selling rides), I can only think of two interpretations:
1) Tesla will continue to make vehicles but not sell them to consumers
2) Elon is redefining what a "car" is and will be selling those to consumers

 

[Knightshade]

Take up your issue with Yipit Data.

The bigger issue is the lack of any detail on where their data is from.

They claim they know how many folks with free FSD trials subscribed.

How does anyone other than Tesla have that number?


Is it just they took a poll of some owners? If so, how many? With what sampling and methodology and confidence level it represents the fleet?

 

[EVMeister]

"Already started DC to NY" So where is it if it's already started? That is a lie, not an optimistic statement

I don't know the history of that one and this is wildly OT in my view. But it is possible that it was in fact started when he Tweeted that but never finished or scrapped.

 

[Max Plaid]

Perhaps people shouldn't be so quick to take his forward looking statements as the truth.

Forward looking statements are never "the truth", they're always aspirational based on current knowledge and certain assumptions

Elon has a track-record of erring towards the optimistic, and is also well known to commit to things off-the-cuff

It's neither untruthful or lying, but investors would do well to treat what he says with some caution and not as a given

 

[2daMoon]

if you take what Elon liked at face value (pivoting from selling cars to selling rides), I can only think of two interpretations:
1) Tesla will continue to make vehicles but not sell them to consumers
2) Elon is redefining what a "car" is and will be selling those to consumers

Let's paraphrase the statement after applying an Elon translator.

"Our primary focus for all new projects and technology will pivot from selling cars to selling rides. This will be a positive move for the company. In a week or so, for those who misunderstood my perspective, I'll post on X how this R&D pivot will not be expected to affect the existing production and delivery of cars at Tesla, which will continue to ramp up toward the goal of 20 million by 2030."

 

[tslagreen]

it is becoming very clear we have changing of the guard for Tesla Investors ...

1st generation ... EV /AP /SuC play
2nd Generation ... FSD/AI/Robots/Energy

the car business is a terrible business.. I would never invest in a car company ... i invested in Tesla in 2017 as a technology company ... the vehicles are just one way to deliver the Software ... Megapacks and Powerwalls are another way ... these are all industries ripe for disruption by SW ....Bots will be next ... Tesla best growth is ahead of it ..

sorry to see the Car company investors dropping out ... i think in hindsight this will have been a huge mistake

one other anecdote

after driving FSD beta and FSD Supervised for around 2 years ... one thing it has highlighted for me is just how bad human drivers are ...it forces you to notice this ... at some point soon it will be negligent to allow humans to drive ... sorry to say this but it is a fact ...

a world where RT/FSD vehicles are all over actually obeying traffic controls and the rules of the road ... will be a much safer world ... i know the path to get there is a long one

Yeah, if this was "just" a car company, the whole investment dynamic would change. As we all know (at least most of us), Tesla is so much more - autos, energy, technology and more led by the most consequential CEO and entrepreneur of his time. I recently saw my first Cybertruck in the wild and I felt like a kid with excitement. Not too many products can invoke that kind of response. I get that Tsla, the stock, has disappointed of late which accounts for much of the grumbling. I would not be as sanguine if I had invested recently instead of a decade ago. I truly expect more great things from Tesla the company as well as Tsla the stock.

 

[advocate8]

Locating and negotiating with site owners would also be better left up to third party pros. Commercial real estate brokerages who already have relationships and knowledge of all the potential sites would handle this task much more effectively

Within reason. Someone in coprorate using proprietary data and planning really would need to give those local and regional contractors say, a 3-mile radius circle for each location and pretty firm guidelines for access to key roads etc. It would have to be pretty tightly controlled, not sure overseeing the QC for a host of outside contractors would be cheaper or more effective than doing it in house.

 

[Discoducky]

on FSD as a likely significant contribution to profits:
First, I still count nothing for RoboTaxi precisely because even with technology 'perfected' there are innumerable impediments to widespread adoption, all well documented;
Second, since 2015 with first Autopilot I have considered Tesla well ahead of competitors. Nothing has changed in that respect, Tesla still is ahead in generalized non-geofenced operation;
Third, I have now had some time with FSD 12.3.6. It is a huge advance, no question. Now it is smooth and consistent, mostly. It does have major weaknesses in even highway routing within complex intersections (my measure of merit on this is Miami's Golden Glades (look it up). FSD even now cannot turn correctly, even for the easiest parts. It is totally incapable of final entry and exit from urban building access.
Fourth, With these limitations I still think buyers would pay for it now were it promoted well. It is not!
Fifth, Constant overselling since 2015 is making existing impediments worse.

In my opinion there is zero chance of anything approaching RoboTaxi to be delivered in a generalized form within this decade. That will NOT happen, in my opinion.

Further I considered the original labeling of Autopilot was perfectly appropriate. It was and remains as definite aid to drivers and has improved in almost every update. It is not anything remotely approaching Full Self Driving.

Were this product properly positioned and promoted I think it was have very high acceptance. However it is promoted as something it is not and the company seems bent on positioning further to something that will not happen anytime soon.

Finally, I use FSD on every drive and am happy I bought it. I did not buy hype, I bought reality. The reality is very, very good.

As for those who claim others are equivalent right now I have a short comment: Don't be ridiculous! Anybody who's actually tried several others and present-issue FSD would need to be delusional to say Tesla is not better.

The HYPE is the problem. not the product! This is a gigantic marketing failure, a major failure in Positioning. IN sum, sell the product you have, not the product you sometime hope to have.

Thank you as this is the kind of content that makes this forum great! I highly value your feedback and know that when/if I disagree we can have a good balanced discussion and healthy back-n-forths.

First off, 12.3.6 is *only* E2E NNs on surface streets, not highways/freeways. You are experiencing legacy V11 heuristics and a not-great hand off when this takes place.

And I think this comment ("It is totally incapable of final entry and exit from urban building access.") will be satiated with summon/banish that should be in 12.4.

How would we (1) expand permitted contributors to the Exalted Few?, including those who Maarten overlooked?

I always have maintained this thread is not a democracy. A meritocracy, though? Hmmm.

By the way, @unk45 most certainly would be one of them.

Also, I'd love to know where this idea is in the mix as I skipped the last 18 pages.

 

[sunwarriors]

Tesla best growth is ahead of it ..

So you think when Tesla grew 50x from 2018 -> 2024 was not their best growth and it will grow even further than 50x from now till maybe, 2030?

Tesla will then be a $27 trillion company with FSD/Robotics/AI/Optimus dominating because selling EVs are not the solution? I think this is a big part of my 'doubt'. Not FUD, but pure law of large numbers since Tesla is simply too big of a company to match their growth of the past 5/6 years.

 

[ggies07]

So you think when Tesla grew 50x from 2018 -> 2024 was not their best growth and it will grow even further than 50x from now till maybe, 2030?

Tesla will then be a $27 trillion company with FSD/Robotics/AI/Optimus dominating because selling EVs are not the solution? I think this is a big part of my 'doubt'. Not FUD, but pure law of large numbers since Tesla is simply too big of a company to match their growth of the past 5/6 years.

Totally get your doubt, but Elon just took care of the part I put in bold. It's all there, set up and ready to go...just rides on execution, just like in the past...Also, you forgot Energy Storage in your domination.