Subaru began testing solid-state batteries in industrial robots at its Oizumi plant in August. The solid-state batteries were manufactured by the Japanese company Maxell, which has been producing solid-state batteries for industrial use since 2023. According to Subaru, the robots’ current batteries normally last 1-2 years. The lifespan of solid-state batteries has been estimated at 10 years. The question of when we will see a Subaru equipped with solid-state batteries received a vague answer. So, the wait continues…
1 Like
If one tries to prove the system’s capability for self-driving, no amount of videos or stories tells anything. The CDE number and its development tell a lot. In the opposite direction, a couple of videos are enough, of course.
Tesla is indeed a great stock from a bull-case perspective. The number of failures, or results, or growth doesn’t matter, because there’s always a new story around the corner.
It’s very easy for me to say when I’m wrong about this robotaxi matter. As soon as Tesla removes safety drivers (tele-monitoring can be done by 5 people for each car) and drives a relevant amount, say, as much as Waymo does now in a month, without serious accidents, then I will have been completely wrong.
Then, when I try to ask the bulls when they are wrong, no answer comes. It’s like believing in God, impossible to prove wrong. Does any Tesla-believing username dare to say they were wrong about Tesla’s robotaxi capabilities if safety drivers haven’t been removed by the end of the year? Regardless of Musk’s latest story at that point.
14 Likes
Musk said Tesla “will geofence” its robotaxis in Austin to start, meaning the company will limit where those Model Y vehicles can drive. But there won’t be a human safety driver in the cars, Musk promised.
Elon Musk confirms Tesla plan for robotaxis on Austin roads in June.
6 Likes
So, in Europe, the first driverless robotaxi rides will be started by the Chinese Pony.AI, which is backed by Chinese car companies and Toyota. Luxembourg will be the launch country. In Europe, the biggest problem is outdated legislation. Even Germany will allow autonomous driving on certain sections before the end of the year, but even then, the car must be remotely controlled. England is likely the only European country that will allow autonomous driving very freely from the beginning of next year.
You have now continued this same FUD about decimals, safety drivers, invitation-only, and geofencing for dozens of messages. What’s bothering you about this? Is it that someone is buying the stock, or that Tesla is expanding its robotaxi trial - or do you still not dare to short? It feels like the robotaxi in Austin has stepped on your toes.
I don’t have an answer either as to when to remove the safety driver or when to apply for FSD permits in India. Tesla’s operational management knows better. When the robotaxi trial and FSD usage expand significantly, things are moving in the right direction - and there’s still room in the market for robotaxis, and competitors aren’t progressing. Other business brings positive cash flow. A taxi customer doesn’t ask about edge-case decimals and CDE figures if the alternative is to take a taxi driven by someone like Arian Juhana Rezaji.
1 Like
This discussion is a perfect description of why one should not short Tesla. I have never considered shorting Tesla, nor do I intend to do so in the future. The stock will very likely crash at some point, but history has taught that Musk’s stories can be believed for years.
Instead, I intend to buy Tesla as soon as the stock seems affordable to me. In the current situation, this would probably happen at around $15.
18 Likes
Well, but when it’s pouring rain and there’s no visibility, autonomy doesn’t help much. Of course, if it’s enough to be as bad as a human driver to get permission, they can very well start the business faster.
That’s right, JP Morgan Chase has believed since 2013 and just bought a hefty 6 million more shares in Q2 - quite a bit more than management sold. Perhaps Musk’s stories about reusable rockets, Giga Shanghai’s schedule, large battery factories, electric car growth, and the future of AI/robotics have convinced JP Morgan Chase and many other financial institutions.
What do these bulls think about the timeline for removing the safety driver? They’ll probably dump all their shares if it’s not removed by the end of the year. That’s when the crash will happen, and we’re waiting for it.
1 Like
Hi!
There are many good messages and content here, but some messages may not be sufficiently related to Tesla as an investment.
Sometimes it’s okay to briefly go slightly off-topic, but please remember to focus especially on essential matters that affect Tesla as an investment.
Thank you for your understanding and happy discussions! 
EDIT:
Thuile, there are many good messages here
12 Likes
Here are statistics on new car registrations in the USA from the beginning of the year until August 12, according to Kelley Blue Book. Even the much-maligned Tesla sold 250,000 cars with two models. The effects of Trump’s electric car policy and tariffs are not yet very visible in these figures. Asian car brands (Japan and South Korea) have a strong grip on the market.
The state of California plans to offer state purchase support for electric car buyers next year and, at the same time, state funding for infrastructure construction. It also intends to set its own limits for emission levels, and the goal is still to ban the sale of new internal combustion engine cars by 2035. A legal battle will surely ensue, but the federal court recently overturned Trump’s Department of Transportation’s decision to stop funding electric vehicle infrastructure. And the Department of Transportation yielded.
- Ford F-150 421,900
- Chevrolet Silverado 284,000
- Toyota RAV4 239,500
- Honda CR-V 212,600
- RAM 1500 174,300
- GMC Sierra 169,200
- Chevrolet Equinox 157,600
- Toyota Camry 155,300
- Tesla Model Y 150,200
- Toyota Tacoma 130,900
- Honda Civic 128,200
- Toyota Corolla 120,100
- Chevrolet Trax 119,600
- Hyundai Tucson 113,300
- Nissan Rogue 109,600
- Ford Explorer 104,900
- Tesla Model 3 101,300
- Jeep Grand Cherokee 99,700
- Subaru Forester 96,000
- Subaru Crosstrek 90,300
2 Likes
Tesla’s allegedly Level 4 autonomous system can’t figure out how to proceed, so the co-driver has to take the wheel.
12 Likes
“In 6 months time it will continue to get even better. I mean if it’s anything like the rate of change in LLMs there’s kinks now but they get worked out very quickly especially when you have more data in the mix.”
This was a very apt concluding summary. Really interesting to see how it works in 6 months.
From the same data, you can also get the headline if you wish:
Tesla’s management team - excluding Musk’s family - has increased its share ownership
or
Tesla’s management team has increased its share ownership
You can always draw favorable headlines from the data.
4 Likes
Forecasts for the end of the year? So much is happening with Tesla’s sales right now that the year-end numbers are at least a mystery to me. I haven’t seen anything from Troy Teslike regarding 3Q, even though he usually actively posts numbers in the second month. On the other hand, the 2Q forecast went so wrong that he might be moving cautiously.
China is showing good QtQ momentum (+46%) and daily reporting countries in Europe are QtQ +19%. The Model Y L received 40,000 orders on its first day, and the delivery time in China has been moved to October. Inventories are already depleted in the US, and price increases are being hinted at – is it just a pull-forward from the credit decision, when at the same time Musk talked about taking the Model Y L off sale to avoid the Osborne effect?
Cybertruck is coming to South Korea, and an FSD release seems to be coming to Australia as a Friday announcement (“no regulatory road blockers” - or perhaps CT or Y L). With a quick calculation, even the small Australian market alone, with a reasonable 8% FSD take rate, would generate over 30 million more to the bottom line (6500x 60,000x 0.08) - software distribution doesn’t incur many costs.
That Ark video about the intervention was a good indication of confidence in robotaxis. They have a calculation for a huge stock value (PT $2600) in their back pocket and yet they actively publish such a video.
https://x.com/r_jegaa/status/1959965519219437640
5 Likes
Xpeng is also abandoning lidar.
1 Like
Austin’s geofence is being expanded again. Could we get more cars too? It’s convenient to call a taxi directly to work from the factory, which is now within the area:
7 Likes
The worst part is that Musk himself seems to become convinced by his own stories, no matter how crazy they are.
Always remember to use either earplugs or a blindfold when moving around to avoid contradictory data from two separate sensors. This way, you also improve your own safety:
I would consider it very likely that Tesla, with its incredibly vast resources, would have already solved most of the major problems related to autonomous driving if Musk hadn’t been obsessed with trying to solve autonomous driving purely camera-based, and if they were willing to approach the problem technology-neutrally and, if necessary, ready to make hybrid solutions, because multimodality is a huge source of strength when training AI.
Even in indoor camera AI automation, which are vastly simpler problems than driving, there are constantly situations where camera AI-related problems are most easily solved by adding another type of sensor to the space. Even in industry, it’s hard to think of an example where additional sensors would be a direct disadvantage to the final system. Perhaps some quantum industry is the only one that comes to mind when decoherence occurs 
Musk should next contact the aircraft industry, suggesting that they should remove unnecessary, redundant sensor-dispute-causing components from airplanes 
43 Likes
Regarding @Pohjolan_Eka’s post, a few more tweets on the topic:
5 Likes
Elon Musk is talking shit once again.
There’s Waymo happily driving in heavy rain.
12 Likes
This example, by the way, is an excellent justification in itself against single-sensor systems and how one can argue crookedly to achieve a desired outcome.
Okay, so you have a situation where a car is driving on a dark road and a tire cannot be detected by lidar, but it can be detected by a camera. Let’s take a car that relies 100% on lidar. Does safety improve by adding a camera to this car? According to Elon Musk’s claim, no, because there will be sensor conflict when the camera detects an obstacle that the lidar does not.
In sensor fusion, one core idea is to lean on the strengths of different sensors in different situations and adjust the weight given to the system’s data accordingly, so that the overall benefit is greater than the sum of its parts. If software developers have trained the model to rely on lidar instead of the camera, on a dark road section, when specifically an object that does not reflect light appears, then the problem is not with the lidar but with having hired mouth-breathers for the job. “System decided to believe in lidar.” Wouldn’t it be better to code the system differently so that when the camera detects a tire object or another identified poorly light-reflecting object in dark driving, the camera’s detection is prioritized over lidar’s 
Let’s take another example from the perspective of a 100% camera-only car. This car is driving in fog, and the visibility of the cameras in all directions is non-existent. Does safety improve or worsen by adding a radar sensor to the car and relying on its sensor data in this situation instead of the camera? Although Elon Musk claims otherwise, the answer is probably obvious.
46 Likes