Isn't the customer always right? How else do you make a company realise they are doing something wrong?
If you want change, there really is only one way to do when a company is led in the way Tesla is.
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
bjorn nyland's test of tesla vision
- Thread starter debrooker
- Start date
Isn't the customer always right? How else do you make a company realise they are doing something wrong?
If you want change, there really is only one way to do when a company is led in the way Tesla is.
_Redshift_
Member
Well, my next car is not going to be a Tesla.
pdk42
Active Member
Wow - this has turned into an epic thread. It’s interesting because I’m now half way through a European road trip in my 3 (currently in Denmark) so I’m getting a lot of time with the car. I’ve had it nearly three years now so I’m very familiar with it. Yes, the auto wipers are crap, auto headlights worse, AP suffers phantom braking all the time, the UI has some terrible design aspects, refinement could be better, ride is too hard on some surfaces, and bugs keep creeping into updates (currently suffering from the alarm going off bug when you open the driver’s door with the phone key).
But, and this is a big but … it’s still a great EV for journeys like this. It’s easy easy easy to do big journeys in Europe because the SuC network is everywhere with large sites, many at 250kW. The car’s sat nav works well for journey planning and it handles preconditioning and range planning extremely well. It even re-routed me to an alternative charger yesterday because the original one was busy! I feel very confident just pointing the car and going without dicking around with ABRP etc. And then it drives well, performs exceptionally well, and is pretty efficient (averaging 230 Wh/mile since I left the UK). And the app experience is great too.
All this is sharply contrasted by comparing it to my wife’s ID3. That’s a car with some nice things about it, but charging is subject to the vagaries of the public network, there’s no preconditioning so charging is usually slow, the software in the car is a nightmare, journey planning is nigh on impossible using the car’s satnav, efficiency is poor (amazing really given that it’s a smaller car than the Tesla), and the app/back end is basic, buggy and slow. But - the auto wipers are great and the headlights are amazing!
I know the Tesla ain’t perfect, but as a package it still works well and if I had to choose another EV as a family daily driver I’d buy another 3. Sure wish they’d fix the auto wipers and headlights though!
But, and this is a big but … it’s still a great EV for journeys like this. It’s easy easy easy to do big journeys in Europe because the SuC network is everywhere with large sites, many at 250kW. The car’s sat nav works well for journey planning and it handles preconditioning and range planning extremely well. It even re-routed me to an alternative charger yesterday because the original one was busy! I feel very confident just pointing the car and going without dicking around with ABRP etc. And then it drives well, performs exceptionally well, and is pretty efficient (averaging 230 Wh/mile since I left the UK). And the app experience is great too.
All this is sharply contrasted by comparing it to my wife’s ID3. That’s a car with some nice things about it, but charging is subject to the vagaries of the public network, there’s no preconditioning so charging is usually slow, the software in the car is a nightmare, journey planning is nigh on impossible using the car’s satnav, efficiency is poor (amazing really given that it’s a smaller car than the Tesla), and the app/back end is basic, buggy and slow. But - the auto wipers are great and the headlights are amazing!
I know the Tesla ain’t perfect, but as a package it still works well and if I had to choose another EV as a family daily driver I’d buy another 3. Sure wish they’d fix the auto wipers and headlights though!
Last edited:
Meanwhile........ The
Your post reminds me that compared to most if not all alternatives, Tesla excels at long distance road trips. Even when I was first coming to terms with phantom braking, whacky lane and speed changes, wipers and other niggles, driving through France to Switzerland I didn't once feel that the car would have to go.
But sat at home - supposedly ready for long trips - next to the Kona, it is usually the Kona that gets used and then getting back in the Tesla all the little niggles (that it's so hard to justify) take centre stage.
Last edited:
Zilla91
Active Member
Sorry, what? You must be in cloud cuckoo land. Why do you think cars aren’t depreciating in value these days - might it have something to do with long lead times?
Takes me a year (or much longer) to commit to a plan of action after endless research.... while knowing that even then there are no guarantees. But changing cars to (effectively) just any other car that people happen to be selling (also over priced and new not an option when supply is constrained) doesn't achieve anything except create more work and encourage dissatisfaction.
There are 2 ways - sell the car and vote with your feet or go on social media and forums and lobby for change
You may think the second method doesn't work, but isn't that how we got Pin to drive, waypoints on Nav etc, and not as a result of 100k Tesla owners taking their car back to Tesla?
"Tech company building cars". Still rings true and explains a lot.
They need to show they're making positive use of the gargantuan data pool they have though, and that they can move more quickly on fixes than the legacy manufacturers.
Also I can't imagine a group of owners right now who are more willing to engage in product improvement than Tesla owners - they should be leveraging this.
To quote monty python: "Strange men lying on their yacht distributing replies to selected tweets is no basis for a system of product feedback"
They need to show they're making positive use of the gargantuan data pool they have though, and that they can move more quickly on fixes than the legacy manufacturers.
Also I can't imagine a group of owners right now who are more willing to engage in product improvement than Tesla owners - they should be leveraging this.
To quote monty python: "Strange men lying on their yacht distributing replies to selected tweets is no basis for a system of product feedback"
Tesla treat software updates like some iphone app, except an iphone app probably is not capable of killing you. I would prefer a well tested update a couple of times a year. There are enough beta testers out there after all. But the endless fiddling has become annoying.
The tipping point for me was when they did the Xmas update a couple of years ago that shrunk the map and left a load of waster blank space. You cant even play beach buggy racing with an Xbox controller on a new Y, they can’t even be bothered to get bluetooth for that working after swiping the data-port usbs. Even the kids are getting fed up with him.
The tipping point for me was when they did the Xmas update a couple of years ago that shrunk the map and left a load of waster blank space. You cant even play beach buggy racing with an Xbox controller on a new Y, they can’t even be bothered to get bluetooth for that working after swiping the data-port usbs. Even the kids are getting fed up with him.
Some people (I’m definitely not one of them) look forward with bated breath to the holiday update every Christmas. It’s usually a major update, and as the day draws near people work themselves up into a frenzy as they wait for what their hero and best friend Elon is going to offer. Unfortunately, according to polls on this forum, most people have found the last two updates to be a big disappointment. I’m sure that won’t stop the frenzied anticipation if there’s a holiday update this year.
An update that fixed the crap windscreen wipers and crap auto headlights would indeed be welcome, but that’s an update I can pretty much guarantee we won’t get.
pdk42
Active Member
Gotta agree on the Christmas updates. I still look back fondly to the V.8 UI.
That UI, which I remember with regret, is MUCH better than the present one. Who said that software updates always mean progress? Clearly they don’t, especially as many updates introduce bugs and some updates even have to be withdrawn. And, of course, if Tesla says we have to update our cars then, as far as I know, we have no choice.
Zilla91
Active Member
To make more room for visualising cones and wheelie bins and display other cars. This was of course a completely sensible move as if your windscreen breaks and you can’t see out of it, Tesla know that replacements take months…
goRt
Active Member
No, that's the way to bankruptcy. The contract is always right, if you want to vary the contract then that's fine.
For Musky it's all about image, so social media pounding on him will eventually bring about change
FirstInTown
Member
Nope, they suck here too. Anyone that defends the wipers (not many, even among the fans) hasn't had a car with a WATER SENSOR to control them. Our Ford and Caddy both had and just leave on and forget about them, they just work. I think Tesla use to have this back in 2016 but cut it for whatever reason (probably to save $$$)
And the auto highs are SO bad. Too bad they got rid of the stalk on the 2021 MX refresh, as the UX to turn on highs manually is even worse than the auto mode (hold button on yoke for 2 seconds to latch high mode. Sucks BAD!). Auto lights are so bad, flash others, don't see taillights right, way too long stay on highs when someone comes over a hill, seconds longer than necessary to go back to high. Worse than 2015 Caddy SRX which I would give a D (below average and didn't use). By 2019 GM had improved them on our Volt to point where leave auto on 90 % of the time. A- rating. Tesla currently gets a D- on auto lights.
(My grading scale is purely subjective and not on a curve.
F is completely useless, should be illegal. Should never use them.
A is wonderful, leave auto on 100% of the time and always work.)
This is a scary video if you believe in AI especially with regard to imagine classification which is the core of Tesla Vision. It’s done by a university and I’ve found many of their videos informative in the past.
It’s easy to dismiss it as “that couldn’t happen”, and it might be pretty unlikely, but it does show that these things aren’t foolproof.
It’s easy to dismiss it as “that couldn’t happen”, and it might be pretty unlikely, but it does show that these things aren’t foolproof.
Its not scary or new. Neural Networks have no 'intelligence' they are simply amazing at pattern recognition and do it with much much better efficiency than humans. What Tesla are trying to do is what Google is/(have done?) is 'generalised AI', the idea/aim of developing true 'intelligence' within an IT system. Is the 'light bulb' movement the thing that give us humans the 'edge' over other living creatures.....If so am not sure how 'special' that moment is.
DALL·E 2
DALL·E 2 is an AI system that can create realistic images and art from a description in natural language.
openai.com
Nothing in regards to the push for AI currently is finished or even close to it, but its quickly becoming obvious there is nothing 'special' about the human mind, like turkeys voting for Xmas, we(humans) cannot help but peruse these developments regardless of the eventual outcomes.
Google DeepMind
Artificial intelligence could be one of humanity’s most useful inventions. We research and build safe artificial intelligence systems. We're committed to solving intelligence, to advance science...
Tesla is constantly having to adapt/change their approach to generalise AI as the pace of development in this area is mad, its not 'vision' that needs solving, its to a degree 'intelligence'. Having seen the hour cost of some AI techs for a recent work project, the amount of $$$$$$$$ Tesla must be spending on AI/FSD development will be mind boggling. A $5 rain/light sensor will almost certainly be cheaper, but it wouldn't deliver 'generalised AI'.
You cannot get any access to DeepMinds code without $$$$$$ and ALOT of IT hardware/experience, even their published work requires a 6 figure sum to replicate (I know this first hand). What Tesla is giving FSD beta tester to in the US is literally million $$$ worth of development.........How well it works right now is a different matter
.
Last edited:
Similar threads
