Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register
This site may earn commission on affiliate links.
Interesting idea, but I suspect it's just changing it's mind, something I've seen it do with no other cars around.


Mine does this a lot. I was about to calibrate the cameras, but it doesn't happen when running v11 on highways, so I don't think that's it.


Yes, I've seen it.
Sadly bouncing over the yellow line was on the interstate, so pretty sure that’s on the V11 stack.
Interestingly I’m seeing similar issues on my Model 3 which is basic AP running 2024.8.9
 
  • Like
Reactions: PianoAl
Sadly bouncing over the yellow line was on the interstate, so pretty sure that’s on the V11 stack.
Interestingly I’m seeing similar issues on my Model 3 which is basic AP running 2024.8.9
I'm becoming fairly convinced that with 8.9 they've replaced the original AP with a hobbled version of 11.4.x. I've suddenly noticed more phantom braking (with occasional swerving!), poorer lane line detection, and poor speed limit detection and management (e.g. seeing a 8 ton limit sign on a bridge in a 80km/h zone and suddenly hard braking to slow down to 8 km/h). But perhaps most tellingly, I noticed one time the "moving chevron" animation in the visualization when closing distance to another vehicle (even though it was actually pulled over on the paved shoulder of the highway), which from my experience with the 3 month trial as a 11.4.x thing.

I hope they fix the issues because suddenly I don't even really like using AP anymore and have been deferring to just TACC.
 
Last edited:
  • Informative
Reactions: primedive
Oh cool, looks like next release (V14) will contain V12 (not to be confused with V12)
 
Oh cool, looks like next release (V14) will contain V12 (not to be confused with V12)
Hoping this has V12 FSD (and not just V12 UI...LOL). If it does then the entire fleet can move to it. Those on 8.X will get their trial. Those on 3.15 will get the 8.X and the 14.X features. Win, win.
 
Hoping this has V12 FSD (and not just V12 UI...LOL). If it does then the entire fleet can move to it. Those on 8.X will get their trial. Those on 3.15 will get the 8.X and the 14.X features. Win, win.
Not the entire fleet, the new features in the UI require a Ryzen CPU.

So those of us with cars older than 2022 won't see the new pretty UI and won't get some of the new 14.x features.
 
Oh cool, looks like next release (V14) will contain V12 (not to be confused with V12)
Totally off topic, should get a new thread: Major UI update says "Ryzen MCU" so Atom gets a pass? I'm ok with that.

I'm really totally and completely over UI updates. Please just stop, or let me freeze my UI. Very bad idea, unless its a recall for a safety update.
 
Oh cool, looks like next release (V14) will contain V12 (not to be confused with V12)
I am lost.
 
  • Like
Reactions: old pilot
Praying this next version fixes curbing. Every right turn on FSD is an anxiety filled adventure.
Yeah, I had an incident today at the rotary in Concord MA... Leaving the rotary, there are two lanes and I was in the left lane. The car cut the corner, so we went about 50% into the right hand lane. Then as we straightened out, FSD put us back centered in the lane.

Not sure what would have happened if there had been a car right next to me, but if someone had been in the right lane slightly overlapped or just behind me, I feel like we would have cut them off and maybe caused a little road rage...
 
My point is you want to be last into the intersection and proceed the minimum distance, if there is a red light runner. That is why you establish visibility by moving promptly and pulling forward. You go quickly so you can see, expecting to stop immediately.

Unless you stay stopped for some time, you are more likely to be further into the intersection and involved in the collision (another shielding car pushed into yours) if you use other traffic as a shield. Red light runners can be very delayed - you have to see them coming, not wait for them to clear.

Having good visibility is key, so you can stop sooner and not be as far into the intersection. If you’re lucky you’ll be so far back that you are not involved.
I pull forward without fsd until I can see oncoming traffic when I'm first in line. Sometimes that means going into the crosswalk when there are no walkers. 2nd or 3rd in line is different. Not sure what I do, other than try to get a view as I enter.
 
  • Like
Reactions: AlanSubie4Life
What is intelligence, anyway ? It’s not a trivial question.
No, it's a very good question.

Here's one (of several) definition I found to be helpful:

The mental quality that consists of the abilities to learn from experience, adapt to new situations, understand and handle abstract concepts, and use knowledge to control an environment.

The "AI" we are discussing in this context is a long way from being able to learn from experience or handle abstract concepts. I don't think my Tesla will ever "understand" what a bear is, or ever come up with a new idea. It might be very fast at approximating a match to prior information it has been fed, and generate a response, but a worm can do that.
 
  • Like
Reactions: Ben W
No, it's a very good question.

Here's one (of several) definition I found to be helpful:

The mental quality that consists of the abilities to learn from experience, adapt to new situations, understand and handle abstract concepts, and use knowledge to control an environment.

The "AI" we are discussing in this context is a long way from being able to learn from experience or handle abstract concepts. I don't think my Tesla will ever "understand" what a bear is, or ever come up with a new idea. It might be very fast at approximating a match to prior information it has been fed, and generate a response, but a worm can do that.
For what it's worth, various CS types and/or hucksters of various flavors have been using, or maybe I should say misusing the term "Artificial Intelligence" whenever there's a buck or concept to be made.

Back in the 80's people would call predictive software systems "Artificial Intelligence" if they had a sufficiently large number of if/else-if statements using probability calculations to solve real-world problems. The one that comes to mind was where a Very Good diagnostician (think, "TV Show House") was followed everywhere by people with clipboards. Every time said diagnostician made a decision or guess about something medical, the people with the clipboards would ask question upon question about what drove that decision. Eventually, this was placed into a software framework and unleashed onto the medical community.

Reviews were mixed. The package was said to be better than a bad diagnostician and would occasionally give a decent (but not great) diagnostician a run for their money. As a doublecheck upon a real diagnostician's guesses, well, better than nothing, I guess. But.. intelligent?

People use AI to describe LLM's, which are big, complex systems that, more or less, predict what the next word to come up in a sentence or paragraph might be. But given the fact that these things "hallucinate", where the definition of hallucination is not the same as Webster's, raises questions about whether these things are actually self aware and actually adaptable. Not to say that fast-talking hucksters haven't convinced pointy-haired bosses (Sports Illustrated, anybody?) that Real Reporters can be swapped out for the "AI" kind with a Great Savings of Money. With predictable results.

My personal opinion: When people are running around screaming "AI" I get my feet off the floor first thing.
 
Oh cool, looks like next release (V14) will contain V12 (not to be confused with V12)
With all those changes it will probably end up being the Holiday update but here's to hoping....
 
  • Like
Reactions: zoomer0056
The "AI" we are discussing in this context is a long way from being able to learn from experience or handle abstract concepts. I don't think my Tesla will ever "understand" what a bear is, or ever come up with a new idea. It might be very fast at approximating a match to prior information it has been fed, and generate a response, but a worm can do that.

I think you might be overestimating the kind of intelligence that an AI system will need in order to safety drive a car. There's a computer scientist named Ajeya Cotra that does some interesting research on forecasting when AI will become transformative by comparing it to biological anchors. She was interviewed by Freakonomics, and the episode is definitely worth a listen: New Technologies Always Scare Us. Is A.I. Any Different? - Freakonomics

By her team's estimates, GPT2, with 1.5 billion parameters, had intelligence equivalent to a honey bee. And GPT4, with an estimated 1.8 trillion parameters, has the intelligence of a squirrel.

The last estimate of the model size of FSD we had came pre-V12, but it was 1 billion parameters at the time. So maybe moving planning and control has increased the model size, but we are currently working with something that's likely closer to honey bee intelligence than squirrel or even human.

All that being said, think about what bees can do with their limited brains. They can navigate a hive, with a complex series of chambers. They can fly at high speed, avoiding obstacles and seeking certain targets. They might not know what a bear is, but they know not to fly into one, or to sting it if it attacks their hive for honey.
 
They might not know what a bear is, but they know not to fly into one, or to sting it if it attacks their hive for honey.
The natural understanding can easily be classified as instincts where advanced understanding can be defined as knowledge.

I believe the honey bee does not know anything but is acting on instinct.
 
Last edited:
  • Like
Reactions: DrGriz