Search results

I've just never heard of a spacecraft/rocket launch being performed with issues known prior to launch that is also carrying humans onboard.

So from what I'm reading, one helium leak was discovered *prior* to launch, engineers discussed it, and they decided to go forward. Does anyone else feel that that they should have addressed those issues before they launched? Especially since two more developed enroute. I feel that because...

Ashok tangentially addressed this in the most recent quarterly conference call. In particular he was referring to training FSD differently for China. He actually made it sound like it wasn't that difficult to adjust for regional differences. So you may be right that state differences won't be...

Or at all, perhaps.

No one here talking about the helium leaks yet? There have been 3 helium leaks discovered onboard. I was hoping to come here to hear more about the implications. Apparently one of the helium links was discovered before launch (hard to believe)...unless it's just not a big issue, I'm not sure.

Yes but since I’m married 12.4 will definitely come first.

I don’t think we’ll see as much in the way of regressions now that we’re fully in NN-land. Every version will have more and better training and probably more neurons/layers. Not that there won’t be regressions, but with heuristic code we’d often give up in one place to gain in another. I’m...

I'll repeat this for the 307th time, and for the 7th year. Results from one country and one month are meaningless when trying to estimate total Tesla results. They may indicate something, but may not. It may have something to do with Tesla, or it may not. Literally sales could drop 90% for a...

The incumbents have a poor supply chain and lack economies of scale when it comes to electric drivetrains. I have literally zero concern about competitors in the electric class 8 truck space.

They "plan to debut" in 2026-2027. That means it debuts in 2028 or 2029 if at all. And that's just the debut. Tesla still has a massive lead. Nobody's even close. And Tesla's not been standing still. They've been gathering millions of miles of real-world experience, iterating and refining.

Class 8 Electric Semi Trucks: Daimler eCascadia: "Up to" 230 miles per charge. Up to 270 kW charging with dual ports. Kenworth T680: Up to 150 mile range, 3 hour recharge. Volvo VNR: Up to 275 mi range, 250 kW charging. Probably all of them cost more than the Tesla Semi. Some of this info...

Thank you for saying this, so I don't have to.

The goal was 20M per year. Always was. However, it was NEVER a promise. Forward-looking statements are just that. There's a disclaimer about them all the time indicating that situations may change. The 20M goal was stated before COVID, before significant inflation, before interest rate hikes...

Sure. Looking forward to the reddit poster's video. I'm sure it will be posted.

OK, now suppose you have 70% of your net worth in TSLA or whatever he claims. How does it benefit him to post the problems publicly? Is he trying to hurt his own stock value? If he truly is an internal beta tester, he'll have a direct line to the autopilot team and can tell them directly...

Could be anything. Stock manipulation was just one example of why someone might post BS on social media. Not saying that explains the poster. May I refer you to the last 15 years of Tesla history, in which BS was spewed daily? I lived through every day of it. Every smelly squishy BS turd. Maybe...

OK, so if both people who claim to have posted about it are legit, how do you explain the discrepancy that one claims (with words only) that it can't even hold the damn lane, whereas the other (posting video) shows essentially on-par performance with 12.3? I'm not saying it *can't* be true. But...

There are bulls who make themselves look stupid, and bears who make themselves look stupid. Some people will cheer the stupid bull stuff, others will cheer the stupid bear stuff. Some bulls will cheer a positive 12.4 review and share it as fact. Some bears will say it's nonsense and call it a...

Let me ask you this: If 12.4 really is drastically worse than 12.3, that would presume: 1. The autopilot team decided to release something for employee testing that is significantly worse than what's already released. That makes no sense. 2. The autopilot team ignored all of their own...

Not mine! :oops:

LOL why would you assume that "70% of his net worth is in tesla stock" is gospel? If that is a lie, then your question is irrelevant. If someone says "I have a massive yacht because I am a billionaire", you are basically saying "well, he MUST have a yacht...after all, he's a billionaire!"

So this guy says it's completely unusable. The earlier guy posted, apparently with an identical software version, and people thought it looked identical to 12.3. So one or both is likely lying. Time to stop trusting the internet!

I got a bumper sticker that just says: "I creep. A lot." People (women in particular) seem to keep their distance--I'm assuming so that if the car makes any strange maneuvers there won't be a collision.

Those dudes already work probably 60 hours a week, and you want them to work more?!?

Yeah I'm not saying that the only reason people aren't getting EVs is because legacies aren't committing. There are obviously practical reasons (people who drive a lot like traveling salespeople might not want to stop to charge, people may not have access to convenient charging, people who tow a...

Even Farley can't be trusted. I thought he had finally figured things out. I guess not. They won't commit until they figure out how to make EVs profitably. Which means Tesla is going to have to go it alone along with BYD and the other Chinese EV manufacturers until legacy gets their heads out...

That doesn't sound intentional--it could be that Tesla forgot to remove those ex-employees' vehicles from the "alpha" tester list. Seems like a potential lawsuit waiting to happen if a non-employee got a version to test with before releasing to the public and there was a problem.

I’m thinking that “leak” of 12.4 a few days ago was fake. If you followed the posts from him on Twitter, he was no longer a Tesla employee when the video was recorded and said a few other questionable things.

I live near several multi-mile tunnels that go quite a bit under the Chesapeake Bay. I’ve never had trouble with FSD holding the lanes while in these tunnels without a GPS signal, so I think the wandering over the shoulder is not related to the GPS signal message.

Too much added cost and complexity. Don't think we'll see this on the higher volume cars.

I'm a bull but of the view that we'll see "only" limited supervised Robotaxi by the end of the year. Meaning Tesla will have a pilot program with a safety driver. If we're very lucky (and by then we'll be quite a bit wealthier if so) the safety driver will come out sometime next year.

You are aware that Elon's other companies "help" Tesla immensely too, right? It's not just a one-way street. Examples: -Cybertruck's cold-rolled stainless steel 30x alloy originated at SpaceX -SpaceX helps with a lot of materials science tasks for Tesla. They trade people and knowledge...

How does 95% compare to diesel semis? I searched a bit and couldn't find a comparable number.

Agree. a semi truck FSD will require a significantly different training dataset than passenger vehicles. Tesla won't be able to start gathering a lot of this useful data until the semi fleet gets quite a bit larger.

Lots of people have financial models for robotaxis. What about for shipping? If, in a few years, Tesla had autonomous semi trucks, every shipping company in the world would want them. Suddenly you have no driver fatigue time limits, don’t have to pay drivers, don’t have to train them, etc...

But a step forward is a step forward. Let's at least appreciate that we're moving in the right direction. Hopefully it will get even better with 12.5.

They will continue to have a nag probably until the car is certified as a level 4 system.

"Downrezzed" was not a good term. Elon used "Emulated". I have no insight into what they're doing to emulate HW4 as HW3.

Typo, I think he meant "nag".

Because he feels they are unsafe for the riders due to a lack of safety structure.

The CEO of a public corporation cannot simply take intellectual property from that corporation. That is called theft. Elon would be spending a lot of time in prison. This will never, ever happen. Let’s stop considering it as a possibility.

Here’s the post:

Elon mentioned that HW4 is being downrezzed to HW3 in current evolutions of FSD--so there shouldn't be any significant differences between HW3 and HW4 right now.

Yes this is a new issue with v12. Indicating a lane change manually isn’t consistent. On city streets, sometimes it will change the lane. Sometimes it will waffle and go back, and sometimes it will just cancel the change altogether. I think that will be fixed in 12.4. I don’t recall having...

Just don't record a report and you should be fine.

Absolutely, I see a lot of silly stuff like this too. Tesla needs to open up a way to crowdsource corrections for these routing issues, and be responsive to fixing them. I'm sure it's on their roadmap, but maybe not their priority until we get a little closer to robotaxi-ready.

The Litmus test: If you can look at video from Chuck’s cameras and determine the right time to go purely from that video, the cameras are sufficient and so can AI. By looking at the output of the cameras you remove the advantage eyes have over the camera. Since we can all do this, the...

Yeah some laws are stupid. Forcing traffic to the right lane even when no other cars are around means there is uneven wear on the asphalt--more wear in the right lane than the left, increasing the need for road construction (just like you get increased wear on rear tires of a RWD car if you...

I don't think you're wrong. And I suspect that's exactly Leo's point for posting his vote publicly. I think he's doing this so that he can talk to someone on the board and make his point. I think ultimately once that happens he'll switch his vote back.

Lol sorry Chuck, that’s only evidence that some Teslas are driving through your neighborhood. Evidence? Really? How do you know it's 12.4 and not 12.5? How do you know it’s validation and not data gathering or experimentation? Wayyyyyy too many people post speculation as fact.