Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

New 40.50 Software

This site may earn commission on affiliate links.
Well, if this update was suppose to be the mother of updates, it is not. Yes, it recognizes stop signs and stop lights, but doesn't stop. It recognized a couple of trash cans sitting on the opposite side of the road and went into a full brake? In a three way intersection; left lane turns, center lane goes straight and right lane is a turn right, I needed to turn left and it stays in the center lane. And more damn games, I paid for FSD, not games. When will Tesla understand that games don't help me drive, get to work on making the car FSD. Very disappointed. Previous post make that 40.50.1

I thought the release notes and other statements were clear on what we were getting in this update, which isn't FSD.

But it is showing progress towards FSD - the first half of FSD is seeing and understanding what is around the car, and that's what they are showing you. The next step is responding to it, and that's not in this firmware yet.
 
I have HW 2.5 and this morning I’ve got notification for 2019.40.50. Anybody else in a similar situation?

I think it's going out to everyone. You probably won't have the option for the FSD visualizations when you get the software, and MCU1 folks likely won't get the new games and streaming options. Still a lot of useful new stuff even then, like the new voice options and text messages.
 
Yes, I can read.
Yes, but apparently understanding is lacking. This is a "sneak peek" of FSD for HW3 only, along with a bunch of other things presumably for everybody. So expect nothing substantive toward FSD behavior.

If you just want to complain that we don't have FSD yes, then you're pretty much the only one to be surprised.
 
I'm sure they are working on FSD diligently. They are just finding out that it is much harder than they realized. Tech people vastly underestimate the power of the human brain and how it processes information. That's also why the whole, "artificial intelligence is a huge worry and machines are going to take over" talk is nonsense. When a car can successfully negotiate a 2-way stop on a divided road come wake me up and we'll set off a flock of doves.
 
  • Like
Reactions: PrGrPa
saw video of Google auto taxis and they have around 400k worth of hardware on them and have a 50 mile radius mapped out to each bump on the road in high def to be able to drive around the city so not sure I am expecting full self driving from my Tesla. I am quite happy with highway self driving that alone, to me, was worth the 7k.

Super interesting. Link?
 
saw video of Google auto taxis and they have around 400k worth of hardware on them and have a 50 mile radius mapped out to each bump on the road in high def to be able to drive around the city so not sure I am expecting full self driving from my Tesla. I am quite happy with highway self driving that alone, to me, was worth the 7k.

I work along in an area of IT where I’m exposed to the kind of technologies that they need for driverless cars. There’s a multitude of problems that car makers have to overcome.

  • Lidar - The advances prototype cars use Lidar because of the 3D mapping of the environment it can create. Think not being blinded by the sun, driving at night, etc. Video cameras on their own currently aren’t likely to cut it.
  • Compute - There’s systems out there that can do things like recognise objects and events happening from video cameras. These run off servers packed with say Nvidia Tesla GPU’s. Those prototype cars normally have a boot for of compute / GPU’s to run what they are doing.
  • Power - The amount of electricity needed to run these systems currently requires more electricity than it takes for the motors to move your car. All the prototypes are fossil fuel cars because of the power draw currently needed, you’d have little range in an electric car. We need to get more compute for less power which mankind has been doing but we aren’t at a level where we have enough yet. We are also getting to transistor sizes that are becoming harder and harder to shrink. We might struggle soon.
  • Code - We haven’t yet written all the code that’s needed for self driving cars or even close. There’s also not enough people with the skills to do it significantly faster even if they wanted to.
There’s other issues of course but these I think are key ones. I wouldn’t expect an AP 4.0 car to get close to real self driving so your dreaming of you think your current car will ever pull this off. I know what Elon says but if you knew this area you’d just know his claims here aren’t true.

I think the solution in the end is roads will need to be adopted to communicate with cars, cars will talk to other cars, etc. It’ll happen but I think it’s still 10+ years away being a commercial reality.

Even knowing all this when I get my first Tesla (Not sure which model yet) I’d still tick the option. I like what they have achieved so far, it’s still very impressive.
 
  • Informative
  • Like
Reactions: aerodyne and MrKool
I work along in an area of IT where I’m exposed to the kind of technologies that they need for driverless cars. There’s a multitude of problems that car makers have to overcome.

  • Lidar - The advances prototype cars use Lidar because of the 3D mapping of the environment it can create. Think not being blinded by the sun, driving at night, etc. Video cameras on their own currently aren’t likely to cut it.
  • Compute - There’s systems out there that can do things like recognise objects and events happening from video cameras. These run off servers packed with say Nvidia Tesla GPU’s. Those prototype cars normally have a boot for of compute / GPU’s to run what they are doing.
  • Power - The amount of electricity needed to run these systems currently requires more electricity than it takes for the motors to move your car. All the prototypes are fossil fuel cars because of the power draw currently needed, you’d have little range in an electric car. We need to get more compute for less power which mankind has been doing but we aren’t at a level where we have enough yet. We are also getting to transistor sizes that are becoming harder and harder to shrink. We might struggle soon.
  • Code - We haven’t yet written all the code that’s needed for self driving cars or even close. There’s also not enough people with the skills to do it significantly faster even if they wanted to.
There’s other issues of course but these I think are key ones. I wouldn’t expect an AP 4.0 car to get close to real self driving so your dreaming of you think your current car will ever pull this off. I know what Elon says but if you knew this area you’d just know his claims here aren’t true.

I think the solution in the end is roads will need to be adopted to communicate with cars, cars will talk to other cars, etc. It’ll happen but I think it’s still 10+ years away being a commercial reality.

Even knowing all this when I get my first Tesla (Not sure which model yet) I’d still tick the option. I like what they have achieved so far, it’s still very impressive.
Apparently you know nothing whatsoever about Tesla's technology. What you've written sounds a lot like what somebody might write about the state of EV batteries and design if they knew nothing about Tesla and were just looking at what other companies have done. Tesla is years ahead. How many years depends on which technology you are considering. In autonomous driving technologies, I would estimate in aggregate about four years.
 
Apparently you know nothing whatsoever about Tesla's technology. What you've written sounds a lot like what somebody might write about the state of EV batteries and design if they knew nothing about Tesla and were just looking at what other companies have done. Tesla is years ahead. How many years depends on which technology you are considering. In autonomous driving technologies, I would estimate in aggregate about four years.

It’s not a case of if they are ahead or not. I still standby what I said that you won’t be getting FSD which is level 5 probably for at least 10+ years.

Honestly I’d love to be wrong. The complexities of it though are just mind boggling really. There’s so many edge cases they have to deal with. All they’ve achieved so far really is the easy highway driving.

The people that brought the first Tesla’s with FSD have been waiting years and years for this. Their hardware has been admitted that it’s not powerful enough. Then they came out with AP 2.0 hardware saying that would be powerful enough but again it’s not, AP 3.0 is now the one they say should do it. I’m pretty sure that also won’t be up to snuff in the end. We shall see.
 
  • Like
Reactions: KArnold
I received the new update, but voice commands don't work. I tried changing the temperature, turning on seats - nothing. I might have to read the release notes again, but I don't think there is anything in the settings menu that needs to be turned on. Anyone else?