TSLA: What Changed?
-
- Posts: 349
- Joined: Fri Sep 03, 2021 1:26 pm
Re: TSLA: What Changed?
All this doesn't matter. Tesla delivered 934k cars in 2021. It will continue to grow 50% every year. So by 2031 it can sell all the 65M vehicles sold in the world and rest of the autos can shut down.
That's it. That's the only analysis one needs! SI
That's it. That's the only analysis one needs! SI
No individual stocks.
Re: TSLA: What Changed?
OK, I think Tesla is more likely to go the way of Blackberry instead of Apple - getting killed by competition.wtfire wrote: ↑Wed Jan 12, 2022 4:57 pmWhen iPhone first came out, Steve Ballmer said, "There’s no chance that the iPhone is going to get any significant market share. No chance. It’s a $500 subsidized item." We know how that turned out. Some of my paycheck-to-paycheck friends carry the "Pro" iphones for reasons beyond me.gougou wrote: ↑Wed Jan 12, 2022 2:18 am If you sell your cars for $42K on average you are not going to sell 10M cars a year. There are only about 65M new car sales per year in the whole world and most of those sales are in developing countries such as China where the average new car price is about $20K.
And good luck getting 45% gross margin on an economic car that most people can afford. Toyota's gross margin is 20% and they have some very economic cars at $20K price point and only then they are selling 10M cars a year. Imagine getting 45% gross margin on a $20K car, maybe the car is made with magic beans instead of physical material and labor?
The FSD software and all those other features are included in the gross margin calculation. The FSD software is a joke I'm telling you with firsthand experience. Tesla cars are just cars with big batteries and nothing special. Tesla's FSD is probably not a market leader and more likely to be a total flop. Tesla has very little competitive advantage which is also getting eroded quickly as other carmakers start to make electric cars with big batteries and fast charging networks.
50% growth every year for the next 7 years is a wet dream. Tesla is not a software company where the incremental product costs 0. Tesla has to spend enormous capital expenditures for all the new capacities to produce the cars, and it has to build every single car with physical material and labor. It is in a very ugly business that never deserves a P/E multiple of 30x or more. And we are not even sure the demand is even there.
So using my original highly optimistic projections I suppose Tesla could become a major carmaker with $500B+ market cap 10 years from now after massive capital expenditure and perfect execution. Current shareholders stand to lose more than 50% over 10 years even under such lofty assumptions.
60% of Tesla buyers are coming from non-luxury car market. They "upgraded" to get an Tesla. The whole "upgrading to iPhones" thing maybe at play here.
As for Apple, it was already dominating the luxury smartphone market in 2012, but it was trading at about 10x P/E for much of 2012 to 2016. I am a buyer of Tesla stock at 10x P/E if it's still the EV leader at that time.
And I'm not sure Tesla car is such a revolutionary product. iPhone is more than a phone which does a lot more things than phones in the past. Tesla car uses a different energy source but it still only moves people around.
The sillier the market’s behavior, the greater the opportunity for the business like investor.
Re: TSLA: What Changed?
Tesla has revolutionized the ability to charge folks $10,000 for a non-existent feature. Even Apple hasn't figured that out.gougou wrote: ↑Wed Jan 12, 2022 7:58 pm OK, I think Tesla is more likely to go the way of Blackberry instead of Apple - getting killed by competition.
As for Apple, it was already dominating the luxury smartphone market in 2012, but it was trading at about 10x P/E for much of 2012 to 2016. I am a buyer of Tesla stock at 10x P/E if it's still the EV leader at that time.
And I'm not sure Tesla car is such a revolutionary product. iPhone is more than a phone which does a lot more things than phones in the past. Tesla car uses a different energy source but it still only moves people around.
Re: TSLA: What Changed?
Sayonara to 2022 Cybertruck. It’s being reported that the release date and pricing pulled from TSLA website. Did anyone really think they were releasing a vehicle with 2x capacity battery for the same price as the Model 3?
-
- Posts: 349
- Joined: Fri Sep 03, 2021 1:26 pm
Re: TSLA: What Changed?
Promise:
Tesla has a reservation list of over 1.2M CyberTrucks. It was the first to display a prototype in 2019 and will be the Number one EV pickup truck.
Reality:
- The reservation is a $120M interest free loan to Tesla. Many Tesla bulls have boasted 10+ reservations.
- Ford, GM and likely even Rivian will outsell Cybertruck.
- When will CyberTruck ship its first 100k - not before 2025 IMO. And average price will not be less than $70k.
- They will never sell for $39,900.
Model X - With one crazy feature of Falcon doors - was delayedHere is timeline for Model X
-Prototype February 2012.
-First shipped in 2015
-100k cumulative in Q4 2018.-- 6 1/2 years after prototype displayed.
CyberTruck crazy features include
- Made of cold-rolled stainless steel body panels - 3 mm (1⁄8 in). Cannot be stamped like conventional automobile parts
- Bullet proof glass, Built in light bar instead of lights, (wipers and side mirrors will likely be added)
- No front crumple zone
- curb weight effect of towing/Payload capacity. Ford lightening as curbweight of 6200-6600lb much heavier than F150 (4-5k lb) mainly due to the 98kWh battery size. CyberTruck has SS will be heavier.
===========
Here is Ford's pricing tool. I wonder how much money they lose on the base price of $41,669 and if anyone is able to buy at that price.
https://shop.ford.com/configure/f150-li ... tomize/pro
No individual stocks.
Re: TSLA: What Changed?
The dealership model is a big one.TimeRunner wrote: ↑Wed Jan 12, 2022 5:43 pmThere's also range anxiety, no-fuss well-maintained Supercharger sites, superior UX, OTA updates, no dealerships to deal with, etc - all issues that current ICE OEMs are going to have to address or will eventually address to be competitive with Tesla's huge first-mover advantage.
Anecdote: our Chevy Bolt EV needed a software update for safety that can't be done over the air. With letter in hand from GM I call my dealer. They absolutely refused to do it, claiming there was no such update. Many other owners have reported the same experience. I had to find a different dealer myself who was willing to do it. For a safety update...
The Bolt has been a great short-range runabout car for us, and it was cheap (<$25k new with all the bells).
Couldn't even get a test drive in a Tesla last time I tried, so there is that.
-
- Posts: 2613
- Joined: Fri Mar 09, 2012 2:47 pm
Re: TSLA: What Changed?
Pain as in Tesla is no longer production constrained or automotive gross margin drops too low? Presumably to maintain their 50% delivery growth target, Tesla will add additional vehicles including trucks and smaller/lower cost vehicles.flyfishers83 wrote: ↑Tue Jan 11, 2022 2:59 pmWhere's the pain point where Tesla runs out of buyers? I think the lowest offering is at 45k
Assuming the $45k Model 3 has 30% margin (current price has delivery 6 months out, so looking 6 months ago, price was $40k), Tesla's costs would be ~$35k of which a guessed 55kWh pack @ $150/kWh is ~$8k. Potentially Tesla could reduce the battery cost and size for a smaller vehicle given their scale of EV manufacturing as well as leveraging their energy efficient vehicle engineering and designs resulting in ~$3k in battery for a ~$30k vehicle cost before any non-battery cost savings of the cheaper vehicle.
How much more demand would there be for a $35k Tesla that still has over 15% margin? Could Tesla focus on software/services revenue while selling the vehicle for $30k or even $25k?
I suppose inflation might be a factor depending on when these vehicles are available, but I believe $35k in 2016 would be over $40k today. Or what used to be $20k in 2016 could be $25k when Tesla is selling the cheaper vehicle, so could Tesla tap into the "$20k" vehicle market to avoid "running out of buyers?"
-
- Posts: 130
- Joined: Fri Jul 20, 2018 11:06 am
Re: TSLA: What Changed?
That word doesn't mean what you seem to think it means. Only your first point (reservation fees being a large free loan) meets the definition of that word: something this is happening or has happened. The rest is all conjecture/expectation, not (yet) reality.
Re: TSLA: What Changed?
search.php?keywords=Tesla&terms=all&aut ... mit=Search
search.php?keywords=TSLA&terms=all&auth ... mit=Search
The above two links are for search terms TSLA and Tesla on this forum, results start from early posts. Lots of evidence to ponder the question of what changed.
search.php?keywords=TSLA&terms=all&auth ... mit=Search
The above two links are for search terms TSLA and Tesla on this forum, results start from early posts. Lots of evidence to ponder the question of what changed.
Pale Blue Dot
Re: TSLA: What Changed?
More viewpoints, analyst interview: https://www.youtube.com/watch?v=al5NsfjGcF0
Pale Blue Dot
-
- Posts: 2613
- Joined: Fri Mar 09, 2012 2:47 pm
Re: TSLA: What Changed?
Seems like Pierre Ferragu (New Street Research analyst) has consistently had relatively high price targets for TSLA primarily based on the evaluation of the core automotive business and less so about AI, insurance or energy. Particularly he does see the recurring revenue from Full Self-Driving capability subscriptions being multiple times the revenue from selling the car with the assumption that eventually more people will want a more capable driver assist, but he believes Tesla will find reaching and deploying robotaxis will be much harder than anticipated.4nursebee wrote: ↑Fri Jan 14, 2022 2:14 pm More viewpoints, analyst interview: https://www.youtube.com/watch?v=al5NsfjGcF0
We've had FSD Beta for 100 days now, and there have definitely been pretty significant improvements in how it drives around here. It's currently a driver assistance system, and when used appropriately (i.e., not treating/expecting it to be an autonomous system), it is actually pretty pleasant to use. To be clear, there are still plenty of caveats and odd behaviors of FSD Beta, but the software is still improving and could soon expand more from innovator/early adopter Tesla population to a majority.
-
- Posts: 349
- Joined: Fri Sep 03, 2021 1:26 pm
Re: TSLA: What Changed?
People on FSD beta have posted videos of vehicles running into on coming traffic, missing red lights getting to wrong lanes, and fender benders. This has been happening for every update over the last 6-8 months since FSD beta started.harikaried wrote: ↑Tue Jan 18, 2022 11:55 am
We've had FSD Beta for 100 days now, ...
To be clear, there are still plenty of caveats and odd behaviors of FSD Beta, but the software is still improving and could soon expand more from innovator/early adopter Tesla population to a majority.
FSD, Full Self Driving, needs level 4/5 automation. And any Professor or Scientist will tell you that level 4 is not possible without Lidar, at the min radar. Tesla uses neither, cameras alone. This is not happening. 1M RoboTaxis or even reliable FSD is not possible without radar and Lidar.
There needs to be just one mistake for it to fatal. Ask families that have sued Tesla for accident deaths. Though these may or may not have been FSD.
No individual stocks.
-
- Posts: 139
- Joined: Wed Jan 20, 2021 7:40 pm
Re: TSLA: What Changed?
Can I order a Cybertruck with the limited edition Iomega interior trim package? And the K-Tel audiophile upgrade?
Re: TSLA: What Changed?
Only if you get the sex doll driver attachment.SeasOfCheese wrote: ↑Wed Jan 19, 2022 4:21 pm Can I order a Cybertruck with the limited edition Iomega interior trim package? And the K-Tel audiophile upgrade?
Re: TSLA: What Changed?
seems to have lost a comma
-
- Posts: 2007
- Joined: Mon Mar 02, 2020 4:33 pm
Re: TSLA: What Changed?
If I, human meatbag, can drive without lidar or radar, as I'm not equipped with any, then so, in principle, can a machine.impatientInv wrote: ↑Wed Jan 19, 2022 4:03 pm [...]
FSD, Full Self Driving, needs level 4/5 automation. And any Professor or Scientist will tell you that level 4 is not possible without Lidar, at the min radar. Tesla uses neither, cameras alone. This is not happening. 1M RoboTaxis or even reliable FSD is not possible without radar and Lidar.
[...]
Re: TSLA: What Changed?
People on FSD beta have also posted videos of successfully navigating from point A to B without any issues. If anyone is getting into an accident or breaking the rules with it, they're not using it correctly or monitoring it close enough. It's all clearly a beta and the user is warned of it as much.impatientInv wrote: ↑Wed Jan 19, 2022 4:03 pm
People on FSD beta have posted videos of vehicles running into on coming traffic, missing red lights getting to wrong lanes, and fender benders. This has been happening for every update over the last 6-8 months since FSD beta started.
FSD, Full Self Driving, needs level 4/5 automation. And any Professor or Scientist will tell you that level 4 is not possible without Lidar, at the min radar. Tesla uses neither, cameras alone. This is not happening. 1M RoboTaxis or even reliable FSD is not possible without radar and Lidar.
There needs to be just one mistake for it to fatal. Ask families that have sued Tesla for accident deaths. Though these may or may not have been FSD.
There's 2 schools of thought about FSD. Professors and Scientists will think that you need HD maps, LIDAR, radar for it to work and those that don't. Tesla employs the ones who think the latter. Comma AI is another company that's going for a third party vision only system. The vision only approach is significantly harder but will eventually need to happen for FSD anywhere to work. HD maps, LIDAR, radar all gives a lot more detailed information of the environment and makes it easier for the AI to drive the car in the surroundings. As a human, you don't need to know in advance there's a stop sign or where all the roads are but it helps. You also don't need to know that the curb or tree is 126 cm away to avoid it, but it would help. Ultimately you just need to see your surroundings, process it into the information that you need to drive, then steer the car through the system which is the easier part (but still extremely complicated and so far underestimated). The hard part is the navigating the unpredictable human drivers causing all the problems that you need to react to.
Families have tried to sue Tesla for accident deaths but none of them (to my knowledge) have won because it's always user error. Falling asleep, texting, not paying attention, using the feature incorrectly, mistaking the accelerator pedal for the brake pedal. It's still a long way from mainstream use and I'm in the camp of not knowing when it'll get there. The interesting part is that once it gets to a point where if you replace all drivers with robots, statistically it'll be safer but it still won't happen because humans don't like giving up control.
-
- Posts: 349
- Joined: Fri Sep 03, 2021 1:26 pm
Re: TSLA: What Changed?
If you are willing to spend $100k+ maybe in 2024. Anyone hope for $39,900, its not happening.SeasOfCheese wrote: ↑Wed Jan 19, 2022 4:21 pm Can I order a Cybertruck with the limited edition Iomega interior trim package? And the K-Tel audiophile upgrade?
Ford and GM will continue to outsell Cybertruck, maybe even Rivian. And EV Truck is small market IMO.
No individual stocks.
-
- Posts: 18502
- Joined: Tue Dec 31, 2013 6:05 am
- Location: 26 miles, 385 yards west of Copley Square
Re: TSLA: What Changed?
Spend ten grand. Rely on auto pilot or whatever Tesla calls it. Run a stop sign, hit someone and kill them. Get charged with manslaughter.
https://www.youtube.com/watch?v=jO2KQeTobUY&t=618s
https://www.youtube.com/watch?v=jO2KQeTobUY&t=618s
Bogle: Smart Beta is stupid
Re: TSLA: What Changed?
The machine's vision could be way better than human vision. Cameras can see more than human-visible light.Hyperchicken wrote: ↑Thu Jan 20, 2022 4:03 pmIf I, human meatbag, can drive without lidar or radar, as I'm not equipped with any, then so, in principle, can a machine.impatientInv wrote: ↑Wed Jan 19, 2022 4:03 pm [...]
FSD, Full Self Driving, needs level 4/5 automation. And any Professor or Scientist will tell you that level 4 is not possible without Lidar, at the min radar. Tesla uses neither, cameras alone. This is not happening. 1M RoboTaxis or even reliable FSD is not possible without radar and Lidar.
[...]
Tesla has a problem, however. There's no way untrained drivers should be given "Full Self Driving" capabilities until it has been proven to work as well as or better than a human driver. I think they have overpromised and need to be seen to get something in the hands of at least some owners, before people start demanding their FSD money back.
I'd be interested in the "$25k Hatchback", if it ever happens. Even if it's more like $35k. I won't be paying an extra $12k for FSD.
-
- Posts: 349
- Joined: Fri Sep 03, 2021 1:26 pm
Re: TSLA: What Changed?
Review of Tesla FSD by Brad Templeton of Robocars.com
https://www.forbes.com/sites/bradtemple ... b8dd444b32
https://www.youtube.com/watch?v=jRrf11M ... obocars%2B
" pretty scathing review from a Tesla owner who is also a Google car veteran. Brad points out a number of dangerous behaviors and walks down the parts of an autonomy system (perception, prediction, planning ...) with comments."
==========
Remember this: When a car is in a accident the driver is responsible 100%. FSD or not
https://www.latimes.com/california/stor ... ?_amp=true
First felony charge in the USA involving Tesla Autopilot (or any L2 system):
"CA prosecutors have filed two counts of vehicular manslaughter against the driver of a Tesla on Autopilot who ran a red light, slammed into another car and killed two people."
https://apnews.com/article/tesla-autopi ... 1b5c2462ae
None of this matters as TSLA is going to $3,000!
https://www.forbes.com/sites/bradtemple ... b8dd444b32
https://www.youtube.com/watch?v=jRrf11M ... obocars%2B
I've owned a Tesla since 2018 and I'm a huge fan of the car. My background is in software and self-driving cars. In addition to writing about them for 15 years, I’ve worked for and advised a wide variety of companies including Waymo, Zoox, Cruise, Starship, giant car OEMs and many others. So naturally been very interested in both Autopilot and the promised future system Tesla calls "full self driving." Tesla has released a prototype (which they call a “beta”) version of that system out to some Tesla owners, and I finally got it recently. I've watched many videos of it in action, some good, some bad, and wanted to see it for myself. So here’s my review, including a video of a sample drive with sadly, too many mistakes.
I have great respect and admiration for Elon Musk, so sorry to say this but ... it's terrible. I mean really bad. After all those videos I didn't expect a lot, but I expected more than this. My first drive home after activating it was frightening. You're going to see the second loop I did, one around Apple's Headquarters in Cupertino California. I've now driven this loop a dozen times with the system on, and each drive is different, with a different pattern of errors, several of them serious.
This is not a complex set of roads. It's typical Silicon Valley Suburbia, the same valley where Tesla's HQ and most of its developers are. There's a fast arterial and some afternoon traffic, but aside from the straight sections, there's no turn or other complexity on the route it didn't screw up at least once in my loops.
" pretty scathing review from a Tesla owner who is also a Google car veteran. Brad points out a number of dangerous behaviors and walks down the parts of an autonomy system (perception, prediction, planning ...) with comments."
==========
Remember this: When a car is in a accident the driver is responsible 100%. FSD or not
https://www.latimes.com/california/stor ... ?_amp=true
First felony charge in the USA involving Tesla Autopilot (or any L2 system):
"CA prosecutors have filed two counts of vehicular manslaughter against the driver of a Tesla on Autopilot who ran a red light, slammed into another car and killed two people."
https://apnews.com/article/tesla-autopi ... 1b5c2462ae
None of this matters as TSLA is going to $3,000!
No individual stocks.
Re: TSLA: What Changed?
The above is the exact reason why there's no real race (or shouldn't be) to get to advertise widespread level 3+ autonomous driving. It would be more for clicks and advertising but it's a huge shift between liability from driver to autonomous company. This is why Comma AI will likely remain level 2 forever, no matter the capabilities, because they doesn't want any of the liability. Anyone that follows any of these FSD capabilities should know that anything the car does is on the owner/driver's fault. Anyone that believes otherwise is completely mistaken.
Once a company does announce widespread level 3+ autonomous driving, they had better be darn sure their capabilities far exceed normal driving needs as there's a lot of dumb people out there that will come up with ways to challenge capabilities no one could see coming.
As for TSLA going to $3000, completely possible from just car manufacturing alone as long as they meet all of their lofty expectations/goals (high volume, high revenue, high margin). If a company does ever solve fully autonomous driving, everywhere, for every vehicle, that company will revolutionize the world (and valued as such) as the TAM is so mind bogglingly large.
HD maps, LIDAR, radar is a crutch for full level 5 autonomous driving. For geofenced limited level 3+ driving, these tools will help get the software to the goal faster without a doubt. It's logistically impossible to HD map the entire world and every road and keep it updated in real time. To get to full level 5 autonomous driving, you need a good enough AI/neural net to navigate the world. If you have that ability, why do you need HD maps or LIDAR?
Once a company does announce widespread level 3+ autonomous driving, they had better be darn sure their capabilities far exceed normal driving needs as there's a lot of dumb people out there that will come up with ways to challenge capabilities no one could see coming.
As for TSLA going to $3000, completely possible from just car manufacturing alone as long as they meet all of their lofty expectations/goals (high volume, high revenue, high margin). If a company does ever solve fully autonomous driving, everywhere, for every vehicle, that company will revolutionize the world (and valued as such) as the TAM is so mind bogglingly large.
HD maps, LIDAR, radar is a crutch for full level 5 autonomous driving. For geofenced limited level 3+ driving, these tools will help get the software to the goal faster without a doubt. It's logistically impossible to HD map the entire world and every road and keep it updated in real time. To get to full level 5 autonomous driving, you need a good enough AI/neural net to navigate the world. If you have that ability, why do you need HD maps or LIDAR?
-
- Posts: 349
- Joined: Fri Sep 03, 2021 1:26 pm
Re: TSLA: What Changed?
1. Tesla has charged over $3B from customers promising Full Self driving. This is 4 years behind. Why do they continue to sell this?
2. Promised 1M RoboTaxis in 2020 on the road.
3. There are "beta" FSD -full self driving vehicles on the road with untrained drivers. And they will get into more accidents. This is likely breaking many laws.. will add links.
4. What happens when vision only doesn't work? And millions of cars sold (without radar) will not have FSD. Remember, Tesla themselves released advance radar back in 2015.
PS: Humans use sound, vast experience of situations, weather, road conditions and judgement to drive. Not just vision.
Promises of Autopilot and FSD video
2. Promised 1M RoboTaxis in 2020 on the road.
3. There are "beta" FSD -full self driving vehicles on the road with untrained drivers. And they will get into more accidents. This is likely breaking many laws.. will add links.
4. What happens when vision only doesn't work? And millions of cars sold (without radar) will not have FSD. Remember, Tesla themselves released advance radar back in 2015.
PS: Humans use sound, vast experience of situations, weather, road conditions and judgement to drive. Not just vision.
Promises of Autopilot and FSD video
Last edited by impatientInv on Sat Jan 22, 2022 7:44 am, edited 1 time in total.
No individual stocks.
Re: TSLA: What Changed?
Just wait until he promises 300 different blood tests using only the Tesla nano-tainer.
70% Global Stocks / 30% Bonds
- PicassoSparks
- Posts: 417
- Joined: Tue Apr 28, 2020 5:41 am
Re: TSLA: What Changed?
Please enjoy this supercut of Elon Musk promising fully automated self driving “next year” every year since 2014.
https://www.reddit.com/r/videos/comment ... ving_cars/
https://www.reddit.com/r/videos/comment ... ving_cars/
-
- Posts: 349
- Joined: Fri Sep 03, 2021 1:26 pm
Re: TSLA: What Changed?
PicassoSparks wrote: ↑Sat Jan 22, 2022 8:16 am Please enjoy this supercut of Elon Musk promising fully automated self driving “next year” every year since 2014.
https://www.reddit.com/r/videos/comment ... ving_cars/
Top comment
"I'm extremely confident that by the end of next year Elon Musk will state that they'll achieve level 4 autonomy by the end of next year."
Alternate link for video.
https://youtu.be/o7oZ-AQszEI
No individual stocks.
Re: TSLA: What Changed?
1. Tesla has only realized a portion of the money they take in from customers' FSD purchase. The rest of it remains deferred. They continue to sell it because customers still buy it now instead of the future as they raise prices. The original FSD was $5000 and now costs $12000 as more functionality gets added. If that's a good idea or not to purchase is up for debate as more features and usability gets rolled out. If they are able to solve it in the near or distant future and then it costs $40000 for full robotaxi functionality, would it have been a good purchase now?impatientInv wrote: ↑Sat Jan 22, 2022 7:37 am 1. Tesla has charged over $3B from customers promising Full Self driving. This is 4 years behind. Why do they continue to sell this?
2. Promised 1M RoboTaxis in 2020 on the road.
3. There are "beta" FSD -full self driving vehicles on the road with untrained drivers. And they will get into more accidents. This is likely breaking many laws.. will add links.
4. What happens when vision only doesn't work? And millions of cars sold (without radar) will not have FSD. Remember, Tesla themselves released advance radar back in 2015.
PS: Humans use sound, vast experience of situations, weather, road conditions and judgement to drive. Not just vision.
2. Elon has greatly underestimated FSD challenges and is constantly enthusiastic with the internal progress they make (warranted or not).
3. I'd be curious for any links you provide how they're breaking (m)any laws.
4. They're going to keep trying for vision only. If it doesn't happen in the distant future and they throw up their hands with futility, they'll refund the portion of the money that has been deferred and call the product 'not FSD'. I paid $4000 EAP and very happy with that purchase.
"A person is smart. People are dumb, panicky dangerous animals and you know it." People also speed/overtake in unsafe weather conditions causing winter pileups because they drive faster than they can safely see in front of them. They text, get angry, eat, put on makeup, race, cut people off, brake check, stop/reverse on highways because they miss their exit all while driving. I've seen a person with a phone with one hand and using a squeegee with the other to clean their windshield driving down the road.
Replace all humans with robots and now there's no human unpredictability to worry about, just rules of the road to follow. The vast experience of situations and judgement could be arguably in favor of the AI with billions or soon trillions of driving miles to compare to/learn from. In the distant future, an autonomous car will be able to get you from point A to B without being distracted while detecting the appropriate safe speed due to road traction/conditions.
-
- Posts: 349
- Joined: Fri Sep 03, 2021 1:26 pm
Re: TSLA: What Changed?
In this video there is compilation of Elon Musk expecting full autonomy, level 4 FSD within a year. Starting 2014 till 2021.impatientInv wrote: ↑Sat Jan 22, 2022 11:07 amPicassoSparks wrote: ↑Sat Jan 22, 2022 8:16 am Please enjoy this supercut of Elon Musk promising fully automated self driving “next year” every year since 2014.
https://www.reddit.com/r/videos/comment ... ving_cars/
Top comment
"I'm extremely confident that by the end of next year Elon Musk will state that they'll achieve level 4 autonomy by the end of next year."
Alternate link for video.
https://youtu.be/o7oZ-AQszEI
He is a CEO of a public company of $1T company.
No individual stocks.
-
- Posts: 3633
- Joined: Thu Jul 09, 2015 7:00 pm
Re: TSLA: What Changed?
A CEO of a public company that oversells and under-delivers? That can’t be right. They put you in prison for that kind of thing.impatientInv wrote: ↑Sat Jan 22, 2022 4:55 pmIn this video there is compilation of Elon Musk expecting full autonomy, level 4 FSD within a year. Starting 2014 till 2021.impatientInv wrote: ↑Sat Jan 22, 2022 11:07 amPicassoSparks wrote: ↑Sat Jan 22, 2022 8:16 am Please enjoy this supercut of Elon Musk promising fully automated self driving “next year” every year since 2014.
https://www.reddit.com/r/videos/comment ... ving_cars/
Top comment
"I'm extremely confident that by the end of next year Elon Musk will state that they'll achieve level 4 autonomy by the end of next year."
Alternate link for video.
https://youtu.be/o7oZ-AQszEI
He is a CEO of a public company of $1T company.
The most precious gift we can offer anyone is our attention. - Thich Nhat Hanh
Re: TSLA: What Changed?
Difference between over selling and just flat out lying which is what Elon does.finite_difference wrote: ↑Sat Jan 22, 2022 5:26 pmA CEO of a public company that oversells and under-delivers? That can’t be right. They put you in prison for that kind of thing.impatientInv wrote: ↑Sat Jan 22, 2022 4:55 pmIn this video there is compilation of Elon Musk expecting full autonomy, level 4 FSD within a year. Starting 2014 till 2021.impatientInv wrote: ↑Sat Jan 22, 2022 11:07 amPicassoSparks wrote: ↑Sat Jan 22, 2022 8:16 am Please enjoy this supercut of Elon Musk promising fully automated self driving “next year” every year since 2014.
https://www.reddit.com/r/videos/comment ... ving_cars/
Top comment
"I'm extremely confident that by the end of next year Elon Musk will state that they'll achieve level 4 autonomy by the end of next year."
Alternate link for video.
https://youtu.be/o7oZ-AQszEI
He is a CEO of a public company of $1T company.
-
- Posts: 349
- Joined: Fri Sep 03, 2021 1:26 pm
Re: TSLA: What Changed?
He is untouchable - above scrutiny. Way bigger than Tesla.
2021 person of year. By Time magazine
None of this actionable for Tesla for a long time.
https://time.com/person-of-the-year-202 ... sk-choice/
2021 person of year. By Time magazine
None of this actionable for Tesla for a long time.
https://time.com/person-of-the-year-202 ... sk-choice/
No individual stocks.
Re: TSLA: What Changed?
And, it’s all priced in anyway. Everybody knows this stuff……. There aren’t. any secrets here…..impatientInv wrote: ↑Sat Jan 22, 2022 6:50 pm He is untouchable - above scrutiny. Way bigger than Tesla.
2021 person of year. By Time magazine
None of this actionable for Tesla for a long time.
https://time.com/person-of-the-year-202 ... sk-choice/
Re: TSLA: What Changed?
The biggest problem with FSD is that unless it's actually perfect, it's more stressful than driving on your own.
When you're driving, you're 100% in control. You're planning your next 2 or 3 moves. You're paying attention to traffic down way down the road. You're aware of who is driving like a jerk around you and can actively avoid.
On FSD, you have to guess what the machine is seeing. It doesn't have the same reaction time as you do. It is either more aggressive or more conservative than you are. You have to always be ready for it to make a mistake and to correct it, but you have no clue when it is making a mistake vs. when it is driving differently than you would.
Until you can lay back and take a nap while FSD is doing its thing, it remains a very expensive, silly, potentially dangerous gadget.
When you're driving, you're 100% in control. You're planning your next 2 or 3 moves. You're paying attention to traffic down way down the road. You're aware of who is driving like a jerk around you and can actively avoid.
On FSD, you have to guess what the machine is seeing. It doesn't have the same reaction time as you do. It is either more aggressive or more conservative than you are. You have to always be ready for it to make a mistake and to correct it, but you have no clue when it is making a mistake vs. when it is driving differently than you would.
Until you can lay back and take a nap while FSD is doing its thing, it remains a very expensive, silly, potentially dangerous gadget.
Re: TSLA: What Changed?
Lol.
I agree with the PS by another poster. People should be using an awful lot of 'stuff' to drive as safe as possible, and I just don’t see how a machine is going to get there any time soon. If we snapped our fingers and 100% of the vehicles on the road were self driving it could work, maybe. We could decrease speed (not average) but increase efficiency, getting people places faster while decreasing kinetic energy and collision possibilities.
Re: TSLA: What Changed?
So whats going to happen to all our portfolios if Tesla crashes?
-
- Posts: 2613
- Joined: Fri Mar 09, 2012 2:47 pm
-
- Posts: 349
- Joined: Fri Sep 03, 2021 1:26 pm
Re: TSLA: What Changed?
Folks thinking of shorting TSLA - PLEASE DONT SHORT. Smart folks have lost large portions of their savings doing this. Many believe TSLA should be way below $600, but it is not actionable.
That said Tesla speculators are feeling a bit of pain.
PS: For the Hedge fund contest I didn't short TSLA. Instead shorted other EV companies
viewtopic.php?f=10&t=366449&p=6440156#p6440156
That said Tesla speculators are feeling a bit of pain.
PS: For the Hedge fund contest I didn't short TSLA. Instead shorted other EV companies
viewtopic.php?f=10&t=366449&p=6440156#p6440156
No individual stocks.
Re: TSLA: What Changed?
Jags4186 wrote: ↑Sat Jan 22, 2022 7:30 pm The biggest problem with FSD is that unless it's actually perfect, it's more stressful than driving on your own.
When you're driving, you're 100% in control. You're planning your next 2 or 3 moves. You're paying attention to traffic down way down the road. You're aware of who is driving like a jerk around you and can actively avoid.
On FSD, you have to guess what the machine is seeing. It doesn't have the same reaction time as you do. It is either more aggressive or more conservative than you are. You have to always be ready for it to make a mistake and to correct it, but you have no clue when it is making a mistake vs. when it is driving differently than you would.
Until you can lay back and take a nap while FSD is doing its thing, it remains a very expensive, silly, potentially dangerous gadget.
Are you speaking from experience? Do you have a Tesla car? Have you used automatic features so far? Are you using FSD beta?
We have used all available features of the car. It took about a week and 800 miles to get a bit used to it, then it became priceless. Highway use of features has been amazing. There is no guessing the limitations, one learns them and anticipates them.
FSD beta became a natural progression. Some things got better, some things just not ready for primetime, but you get used to it and anticipate. More attention is required, the car won't let me play on the cell phone (a great safety feature). Mistakes? Sure there are several, but again, you learn and stay at the ready.
My view is this feature set is priceless, not silly, and potentially a life saver.
My spouse, who has only been a passenger for NOA has taken to FSD beta like a fish in water. Anxious, but willing and able to use.
I have family that owns a TSLA, had never turned on some of the auto features, took a long drive on a windy day and fell in love.
If the car let me, I am confident I could go to sleep on the highway now.
Pale Blue Dot
-
- Posts: 349
- Joined: Fri Sep 03, 2021 1:26 pm
Re: TSLA: What Changed?
What did you think about this drive below, his commentary and grades? Actual drive video starts at t=5:36.4nursebee wrote: ↑Mon Jan 24, 2022 5:06 pmJags4186 wrote: ↑Sat Jan 22, 2022 7:30 pm The biggest problem with FSD is that unless it's actually perfect, it's more stressful than driving on your own.
When you're driving, you're 100% in control. You're planning your next 2 or 3 moves. You're paying attention to traffic down way down the road. You're aware of who is driving like a jerk around you and can actively avoid.
On FSD, you have to guess what the machine is seeing. It doesn't have the same reaction time as you do. It is either more aggressive or more conservative than you are. You have to always be ready for it to make a mistake and to correct it, but you have no clue when it is making a mistake vs. when it is driving differently than you would.
Until you can lay back and take a nap while FSD is doing its thing, it remains a very expensive, silly, potentially dangerous gadget.
Are you speaking from experience? Do you have a Tesla car? Have you used automatic features so far? Are you using FSD beta?
We have used all available features of the car. It took about a week and 800 miles to get a bit used to it, then it became priceless. Highway use of features has been amazing. There is no guessing the limitations, one learns them and anticipates them.
FSD beta became a natural progression. Some things got better, some things just not ready for primetime, but you get used to it and anticipate. More attention is required, the car won't let me play on the cell phone (a great safety feature). Mistakes? Sure there are several, but again, you learn and stay at the ready.
My view is this feature set is priceless, not silly, and potentially a life saver.
My spouse, who has only been a passenger for NOA has taken to FSD beta like a fish in water. Anxious, but willing and able to use.
I have family that owns a TSLA, had never turned on some of the auto features, took a long drive on a windy day and fell in love.
If the car let me, I am confident I could go to sleep on the highway now.
In the drive below it veers towards a parked trailer & almost hits it (driver had to intervene); jumps a red light; veers onto oncoming traffic and more...
https://www.youtube.com/watch?v=jRrf11M ... obocars%2B
https://www.forbes.com/sites/bradtemple ... rive-an-f/
So I'm giving Tesla FSD an "F" when it comes to self-driving. In fact it clearly shouldn't have that name, as many have pointed out. It should have a driver-assist name, so I will call it "Street Autopilot". The problem is I have to give it a "D" as a driver assist product. Using it is a harrowing experience. It's definitely not relaxing or providing assistance. (To those who feel an “F” is too harsh, let me know where you can take a driving test and run 2 red lights, block the road crosswalk for long periods in multiple intersections while people honk at you and the tester has to grab the wheel twice to stop you from hitting things, and 3 times make random turns the tester didn’t tell you to make — and still not get an “F” on the test. In fact, the test would be stopped at the first one of most of these.)
To be fair, I and many wondered if Autopilot could be a pleasant and relaxing product. If you use it right, on the highway, it is. I saw that quickly after using it. It's hard to see that for Street Autopilot, even if it gets a lot better. With highway Autopilot, it takes a gently winding highway, and makes driving on it like driving on a beeline straight road with light traffic. You're still driving, but it's easier. Decisions are rarely quick. You have to focus, but on different things. In traffic, it does what any adaptive cruise control does to take the stress out.
Street Autopilot is much harder to make relaxing. The car is now making sharp turns and quick decisions. Other road users, including pedestrians are coming from all directions. You have to stay fully aware. When you drive yourself, you know your plan in advance, and you know when things are going according to your plan. With Street Autopilot, it's doing the planning so its moves can be a surprise to you. This probably gets a bit better when you get more used to how it drives, but I haven't reached that yet. I'll report more later on that.
About 5 minutes into the video you will see commentary on various situations where it had problems including:
1. Yielding too long at a 3 way stop, even though it was clearly there first (D+) T=5:36.
2. Veering towards a trailer on the side of a quiet street(F) T=6:20
3. Being very slow turning onto an arterial and getting honked at (D) T= 7:39
4. Pointlessly changing lanes for a very short time T= 7:39
5. Failing in many ways at a right turn to a major street that has its own protected lane, almost always freezing and not knowing what to do (F) T= 8:40
6. Jerky accelerations and turns (D)
7. Stalling for long times at right turns on red lights (F)
8. Suddenly veering off-course into a left turn that’s not on the route, then trying to take that turn even though the light is red! (F, and ends test immediately) T=10:55
9. Finding itself in a “must turn left” lane and driving straight out of it, or veering left into oncoming traffic (F, and ends test immediately) T=11:30
10. Handing a basic right turn with great uncertainty, parking itself in the bike lane for a long period to judge oncoming traffic (F)
11. Taking an unprotected left with a narrow margin, and doing it so slowly that the oncoming driver has to brake hard. (Possible F)
All of these in a simple 3.5 mile loop in a suburban residential/commercial area.
One thing he points out and you said as well - Autopilot on the Highway works and is a pleasant experience (mostly). But he says that Street FSD is very different than Highway Autopilot.
Here is what Elon has to say about difference between Street FSD (which is in Beta) and Highway Autopilot. Basically means that "Both are completely different software"
https://techcrunch.com/2021/08/24/musk- ... y-streets/FSD Beta 9.2 is actually not great imo, but Autopilot/AI team is rallying to improve as fast as possible.
We’re trying to have a single stack for both highway & city streets, but it requires massive NN retraining.
No individual stocks.
-
- Posts: 3633
- Joined: Thu Jul 09, 2015 7:00 pm
Re: TSLA: What Changed?
I disagree. It doesn’t have to be perfect. It has to be better than human drivers. Right now human drivers kill 30,000 people per year and maim hundreds of thousands of people. If AI driving kills 300 people/year that’s not perfect but is two orders of magnitude better. There will likely always be some accidents caused by human drivers (drunk, fall asleep, incapacitated, mistakes, etc.) running into AI cars. Or mechanical failure, e.g. tire blowout rounding a steep curve. I predict that by 2025 that a Tesla will drive better than the average human in most circumstances. (They claim they already do, but proof hasn’t been published yet.) By 2030, all modern cars will drive better than the average human in almost all circumstances. And by average, I mean an average driver human that is never drunk, tired or inattentive.Jags4186 wrote: ↑Sat Jan 22, 2022 7:30 pm The biggest problem with FSD is that unless it's actually perfect, it's more stressful than driving on your own.
When you're driving, you're 100% in control. You're planning your next 2 or 3 moves. You're paying attention to traffic down way down the road. You're aware of who is driving like a jerk around you and can actively avoid.
On FSD, you have to guess what the machine is seeing. It doesn't have the same reaction time as you do. It is either more aggressive or more conservative than you are. You have to always be ready for it to make a mistake and to correct it, but you have no clue when it is making a mistake vs. when it is driving differently than you would.
Until you can lay back and take a nap while FSD is doing its thing, it remains a very expensive, silly, potentially dangerous gadget.
The most precious gift we can offer anyone is our attention. - Thich Nhat Hanh
-
- Posts: 2613
- Joined: Fri Mar 09, 2012 2:47 pm
Re: TSLA: What Changed?
Not sure if you are intentionally being misleading in quoting that or didn't understand what he was talking about as it is some "inside baseball" level of detail going on. The tweet was specifically about FSD Beta 9.2 performing worse than FSD Beta 9.1 in their internal testing, and these versions are released about every 2 weeks. These fortnightly sprints often include new human coded behaviors and/or neural network training, and each of those can result in regressions where something that was working gets worse in attempts to improve some other behavior.impatientInv wrote: ↑Mon Jan 24, 2022 9:16 pmhe says that Street FSD is very different than Highway AutopilotFSD Beta 9.2 is actually not great imo
Basically what the tweet was conveying was that the usual 2 week cadence would probably be skipped this time as it wasn't good enough quality to release to the beta testers and not that FSD Beta overall is "not great." It's been about 22 weeks from that tweet and indeed there have been about 11 versions of FSD Beta each with various improvements that have allowed Tesla to expand the audience to 20k+ owners so far.
The "single stack for both highway & city streets" has yet to happen for FSD Beta, so yes currently they are two very different behaviors (that is very obvious for those who have used FSD Beta). Supposedly the combining of the two will happen with the next major FSD Beta version with a pretty high quality threshold for the new software to exceed the real-world tested production highway Autopilot.
Re: TSLA: What Changed?
The video where it "veered" into a truck and trailer can't be readily evaluated as he cut off most of the screen.impatientInv wrote: ↑Mon Jan 24, 2022 9:16 pmWhat did you think about this drive below, his commentary and grades? Actual drive video starts at t=5:36.4nursebee wrote: ↑Mon Jan 24, 2022 5:06 pmJags4186 wrote: ↑Sat Jan 22, 2022 7:30 pm The biggest problem with FSD is that unless it's actually perfect, it's more stressful than driving on your own.
When you're driving, you're 100% in control. You're planning your next 2 or 3 moves. You're paying attention to traffic down way down the road. You're aware of who is driving like a jerk around you and can actively avoid.
On FSD, you have to guess what the machine is seeing. It doesn't have the same reaction time as you do. It is either more aggressive or more conservative than you are. You have to always be ready for it to make a mistake and to correct it, but you have no clue when it is making a mistake vs. when it is driving differently than you would.
Until you can lay back and take a nap while FSD is doing its thing, it remains a very expensive, silly, potentially dangerous gadget.
Are you speaking from experience? Do you have a Tesla car? Have you used automatic features so far? Are you using FSD beta?
We have used all available features of the car. It took about a week and 800 miles to get a bit used to it, then it became priceless. Highway use of features has been amazing. There is no guessing the limitations, one learns them and anticipates them.
FSD beta became a natural progression. Some things got better, some things just not ready for primetime, but you get used to it and anticipate. More attention is required, the car won't let me play on the cell phone (a great safety feature). Mistakes? Sure there are several, but again, you learn and stay at the ready.
My view is this feature set is priceless, not silly, and potentially a life saver.
My spouse, who has only been a passenger for NOA has taken to FSD beta like a fish in water. Anxious, but willing and able to use.
I have family that owns a TSLA, had never turned on some of the auto features, took a long drive on a windy day and fell in love.
If the car let me, I am confident I could go to sleep on the highway now.
In the drive below it veers towards a parked trailer & almost hits it (driver had to intervene); jumps a red light; veers onto oncoming traffic and more...
https://www.youtube.com/watch?v=jRrf11M ... obocars%2B
https://www.forbes.com/sites/bradtemple ... rive-an-f/
So I'm giving Tesla FSD an "F" when it comes to self-driving. In fact it clearly shouldn't have that name, as many have pointed out. It should have a driver-assist name, so I will call it "Street Autopilot". The problem is I have to give it a "D" as a driver assist product. Using it is a harrowing experience. It's definitely not relaxing or providing assistance. (To those who feel an “F” is too harsh, let me know where you can take a driving test and run 2 red lights, block the road crosswalk for long periods in multiple intersections while people honk at you and the tester has to grab the wheel twice to stop you from hitting things, and 3 times make random turns the tester didn’t tell you to make — and still not get an “F” on the test. In fact, the test would be stopped at the first one of most of these.)
To be fair, I and many wondered if Autopilot could be a pleasant and relaxing product. If you use it right, on the highway, it is. I saw that quickly after using it. It's hard to see that for Street Autopilot, even if it gets a lot better. With highway Autopilot, it takes a gently winding highway, and makes driving on it like driving on a beeline straight road with light traffic. You're still driving, but it's easier. Decisions are rarely quick. You have to focus, but on different things. In traffic, it does what any adaptive cruise control does to take the stress out.
Street Autopilot is much harder to make relaxing. The car is now making sharp turns and quick decisions. Other road users, including pedestrians are coming from all directions. You have to stay fully aware. When you drive yourself, you know your plan in advance, and you know when things are going according to your plan. With Street Autopilot, it's doing the planning so its moves can be a surprise to you. This probably gets a bit better when you get more used to how it drives, but I haven't reached that yet. I'll report more later on that.
About 5 minutes into the video you will see commentary on various situations where it had problems including:
1. Yielding too long at a 3 way stop, even though it was clearly there first (D+) T=5:36.
2. Veering towards a trailer on the side of a quiet street(F) T=6:20
3. Being very slow turning onto an arterial and getting honked at (D) T= 7:39
4. Pointlessly changing lanes for a very short time T= 7:39
5. Failing in many ways at a right turn to a major street that has its own protected lane, almost always freezing and not knowing what to do (F) T= 8:40
6. Jerky accelerations and turns (D)
7. Stalling for long times at right turns on red lights (F)
8. Suddenly veering off-course into a left turn that’s not on the route, then trying to take that turn even though the light is red! (F, and ends test immediately) T=10:55
9. Finding itself in a “must turn left” lane and driving straight out of it, or veering left into oncoming traffic (F, and ends test immediately) T=11:30
10. Handing a basic right turn with great uncertainty, parking itself in the bike lane for a long period to judge oncoming traffic (F)
11. Taking an unprotected left with a narrow margin, and doing it so slowly that the oncoming driver has to brake hard. (Possible F)
All of these in a simple 3.5 mile loop in a suburban residential/commercial area.
One thing he points out and you said as well - Autopilot on the Highway works and is a pleasant experience (mostly). But he says that Street FSD is very different than Highway Autopilot.
Here is what Elon has to say about difference between Street FSD (which is in Beta) and Highway Autopilot. Basically means that "Both are completely different software"
https://techcrunch.com/2021/08/24/musk- ... y-streets/FSD Beta 9.2 is actually not great imo, but Autopilot/AI team is rallying to improve as fast as possible.
We’re trying to have a single stack for both highway & city streets, but it requires massive NN retraining.
He certainly has a number of things come up, shows exactly why a driver needs to stay in control. But it seems like he learns and understands the deficits, monitors for them, corrects them.
What I don't see is him reporting the problems as he should. And he acknowledges that he has not installed a pending update.
I am not sure what makes him an expert, but perhaps he has vested interests?
Pale Blue Dot
-
- Posts: 49035
- Joined: Fri May 11, 2007 11:07 am
Re: TSLA: What Changed?
Lawsuits are probably not a good test of whether the AI is working.Nysoz wrote: ↑Fri Jan 21, 2022 5:05 am
Families have tried to sue Tesla for accident deaths but none of them (to my knowledge) have won because it's always user error. Falling asleep, texting, not paying attention, using the feature incorrectly, mistaking the accelerator pedal for the brake pedal. It's still a long way from mainstream use and I'm in the camp of not knowing when it'll get there. The interesting part is that once it gets to a point where if you replace all drivers with robots, statistically it'll be safer but it still won't happen because humans don't like giving up control.
There are the inevitable problems with automated systems -- hacking, jamming, what happens when all the AIs make the same (wrong) decision simultaneously, etc.
So I am sceptical, still, that we will get to the heights of full autonomy.
And very concerned that the loss of driver alertness at lower levels will lead to horrible accidents. That woman walking her bicycle across the road & her horrified face ... that clip will live in the minds of those who have seen it. (I realise this was an Uber AV not a Tesla?)
But it's worth saying that this is the place commercial aviation is going. As I understand it, a pilot is increasingly a computer operator - the plane flies itself. Could even land itself. Airbus proved that this was safer, and Boeing has followed suit (see William Langeswische "Fly by Wire" about the Hudson River flyer (Sully Sullentrope)).
So it is possible. Driving is not actually a pleasure in the conditions many of us do it - high traffic congestion, difficult or uncertain road conditions. It's a constant test of alertness v boredom. We are not highly skilled professionals -- we are just amateurs who need to go places.
Road deaths are now the leading cause of death worldwide (ex Covid) I believe-- over 1m dead pa plus who knows how many permanently maimed. Non disease death in any case. The highways of emerging markets have turned into virtual slaughterhouses.
And our populations are aging - rapidly. That means many of us will be restricted in driving. I had the onerous task of explaining to a relative, who had driven for 70+ years, driven professionally for her work as a civil servant, that she was no longer safe on the roads, and many of her neighbours in the village had approached us at church to tell us this. It was a difficult conversation for both of us.
Reflexes, eyesight, road awareness go much faster than we realise.
So true AV would meet great acceptance - much faster than probably we think.
But the technical problem of true AV is a lot worse than we think. That seems to be at least one view out there, and supported at least from a 30,000 foot view by the history of technology hype cycles.
Re: TSLA: What Changed?
It has to be near perfect for universal acceptance. Sure, it may one day get to the point of being safer overall than the average human, but due to illusory superiority, almost everyone believes they are above average drivers. Therefore it needs to be a lot safer than average. And when an accident happens with human drivers, most of us are likely to think that an individual or individuals made bad decisions. Their decisions do not impact your decision making. But if a system you rely on caused an accident, you now have doubts about your own system’s decision making ability.finite_difference wrote: ↑Mon Jan 24, 2022 10:40 pmI disagree. It doesn’t have to be perfect. It has to be better than human drivers. Right now human drivers kill 30,000 people per year and maim hundreds of thousands of people. If AI driving kills 300 people/year that’s not perfect but is two orders of magnitude better. There will likely always be some accidents caused by human drivers (drunk, fall asleep, incapacitated, mistakes, etc.) running into AI cars. Or mechanical failure, e.g. tire blowout rounding a steep curve. I predict that by 2025 that a Tesla will drive better than the average human in most circumstances. (They claim they already do, but proof hasn’t been published yet.) By 2030, all modern cars will drive better than the average human in almost all circumstances. And by average, I mean an average driver human that is never drunk, tired or inattentive.Jags4186 wrote: ↑Sat Jan 22, 2022 7:30 pm The biggest problem with FSD is that unless it's actually perfect, it's more stressful than driving on your own.
When you're driving, you're 100% in control. You're planning your next 2 or 3 moves. You're paying attention to traffic down way down the road. You're aware of who is driving like a jerk around you and can actively avoid.
On FSD, you have to guess what the machine is seeing. It doesn't have the same reaction time as you do. It is either more aggressive or more conservative than you are. You have to always be ready for it to make a mistake and to correct it, but you have no clue when it is making a mistake vs. when it is driving differently than you would.
Until you can lay back and take a nap while FSD is doing its thing, it remains a very expensive, silly, potentially dangerous gadget.
Full Self Driving will definitely be a thing one day that everyone has. But it is way further down the line than being suggested in this thread. When young children regularly see and experience full self driving and grow up to become vehicle purchasers, then it will become ubiquitous. Lots of cars today will essentially drive themselves on the highway, so it’s today’s youngsters will push us to FSD ubiquity. We’re probably 20-30 years away.
Re: TSLA: What Changed?
I purchased "FSD" for my 2018 model 3 for $3K mostly to a) get the HW3 computer upgrade and b) support Tesla's march towards Level 4/5. Since my wife and I don't have children we will be counting on autonomous vehicles to drive us around at some point. Having used the Beta I agree it is further away than many think. I also have doubts that the side cameras have enough range and they clearly have no mitigation against rain or sun-blinding. I think once Tesla themselves really believe they are close to L3/L4/L5 they will remedy that. Until then, I don't believe Tesla is as close as they publicly advertise. I think they are staying tight-lipped to avoid popping the TSLA stock bubble. As for the public acceptance of autonomous technology, once the automobile insurance companies are convinced it is safer than human drivers they will price automobile insurance accordingly and people will vote with their wallets: Pay more for the privilege of manually driving one's automobile? We will see. As for me, once I am convinced an autonomous automobile can drive more safely than me I will happily hand over the "keys" to my robot overlords.Jags4186 wrote: ↑Tue Jan 25, 2022 6:59 amIt has to be near perfect for universal acceptance. Sure, it may one day get to the point of being safer overall than the average human, but due to illusory superiority, almost everyone believes they are above average drivers. Therefore it needs to be a lot safer than average. And when an accident happens with human drivers, most of us are likely to think that an individual or individuals made bad decisions. Their decisions do not impact your decision making. But if a system you rely on caused an accident, you now have doubts about your own system’s decision making ability.finite_difference wrote: ↑Mon Jan 24, 2022 10:40 pmI disagree. It doesn’t have to be perfect. It has to be better than human drivers. Right now human drivers kill 30,000 people per year and maim hundreds of thousands of people. If AI driving kills 300 people/year that’s not perfect but is two orders of magnitude better. There will likely always be some accidents caused by human drivers (drunk, fall asleep, incapacitated, mistakes, etc.) running into AI cars. Or mechanical failure, e.g. tire blowout rounding a steep curve. I predict that by 2025 that a Tesla will drive better than the average human in most circumstances. (They claim they already do, but proof hasn’t been published yet.) By 2030, all modern cars will drive better than the average human in almost all circumstances. And by average, I mean an average driver human that is never drunk, tired or inattentive.Jags4186 wrote: ↑Sat Jan 22, 2022 7:30 pm The biggest problem with FSD is that unless it's actually perfect, it's more stressful than driving on your own.
When you're driving, you're 100% in control. You're planning your next 2 or 3 moves. You're paying attention to traffic down way down the road. You're aware of who is driving like a jerk around you and can actively avoid.
On FSD, you have to guess what the machine is seeing. It doesn't have the same reaction time as you do. It is either more aggressive or more conservative than you are. You have to always be ready for it to make a mistake and to correct it, but you have no clue when it is making a mistake vs. when it is driving differently than you would.
Until you can lay back and take a nap while FSD is doing its thing, it remains a very expensive, silly, potentially dangerous gadget.
Full Self Driving will definitely be a thing one day that everyone has. But it is way further down the line than being suggested in this thread. When young children regularly see and experience full self driving and grow up to become vehicle purchasers, then it will become ubiquitous. Lots of cars today will essentially drive themselves on the highway, so it’s today’s youngsters will push us to FSD ubiquity. We’re probably 20-30 years away.
Adapt or perish
-
- Posts: 3633
- Joined: Thu Jul 09, 2015 7:00 pm
Re: TSLA: What Changed?
I don’t think it needs to be near perfect to have universal acceptance. If it’s proven to drive better than an average non-drunk, non-tired, and non-distracted driver then I think that would go a long way for acceptance. Let’s go a little further. Let’s say it is statistically proven to be as good as an average driver that is female, between 30 and 60 years of age, and that is paying attention and sober. I would use that all the time. I’d still pay attention and be ready to jump in as needed, and the system would require you to pay attention by monitoring your face and body, but I would definitely still find it very useful.Jags4186 wrote: ↑Tue Jan 25, 2022 6:59 amIt has to be near perfect for universal acceptance. Sure, it may one day get to the point of being safer overall than the average human, but due to illusory superiority, almost everyone believes they are above average drivers. Therefore it needs to be a lot safer than average. And when an accident happens with human drivers, most of us are likely to think that an individual or individuals made bad decisions. Their decisions do not impact your decision making. But if a system you rely on caused an accident, you now have doubts about your own system’s decision making ability.finite_difference wrote: ↑Mon Jan 24, 2022 10:40 pmI disagree. It doesn’t have to be perfect. It has to be better than human drivers. Right now human drivers kill 30,000 people per year and maim hundreds of thousands of people. If AI driving kills 300 people/year that’s not perfect but is two orders of magnitude better. There will likely always be some accidents caused by human drivers (drunk, fall asleep, incapacitated, mistakes, etc.) running into AI cars. Or mechanical failure, e.g. tire blowout rounding a steep curve. I predict that by 2025 that a Tesla will drive better than the average human in most circumstances. (They claim they already do, but proof hasn’t been published yet.) By 2030, all modern cars will drive better than the average human in almost all circumstances. And by average, I mean an average driver human that is never drunk, tired or inattentive.Jags4186 wrote: ↑Sat Jan 22, 2022 7:30 pm The biggest problem with FSD is that unless it's actually perfect, it's more stressful than driving on your own.
When you're driving, you're 100% in control. You're planning your next 2 or 3 moves. You're paying attention to traffic down way down the road. You're aware of who is driving like a jerk around you and can actively avoid.
On FSD, you have to guess what the machine is seeing. It doesn't have the same reaction time as you do. It is either more aggressive or more conservative than you are. You have to always be ready for it to make a mistake and to correct it, but you have no clue when it is making a mistake vs. when it is driving differently than you would.
Until you can lay back and take a nap while FSD is doing its thing, it remains a very expensive, silly, potentially dangerous gadget.
Full Self Driving will definitely be a thing one day that everyone has. But it is way further down the line than being suggested in this thread. When young children regularly see and experience full self driving and grow up to become vehicle purchasers, then it will become ubiquitous. Lots of cars today will essentially drive themselves on the highway, so it’s today’s youngsters will push us to FSD ubiquity. We’re probably 20-30 years away.
You think it will take 25 years (2055). I think it will take 10 (2030). Let’s see who is right.
The most precious gift we can offer anyone is our attention. - Thich Nhat Hanh
Re: TSLA: What Changed?
Lunch today: Left home, put features on once on the road. 20 min trip to destination through 3 very small towns. Stopped at a friends house, then 10 min trip through small but larger town to restaurant, then 20 min trip to home through 1 or 2 small towns. All features running.
Zero disengagements for safety.
I had to provide two inputs to get the car to make left turns, just tapping on the accelerator. Might could have waited, not likely.
I adjusted speed up to limit +5 mph regularly.
One passenger grabbed the handles for an aggressive entry on a curve, car slowed way down and handled it well.
Car stayed on the road, adjusted speed, used turn signals, obeyed the laws. It fussed at me to pay attention once when I looked over my shoulder. 4 way stops good. Traffic lights good. Right and left turns were slow and cautious. Stone driveways not so good!
Perhaps there needs to be a totally separate ongoing discussion for TSLA FSD experience and value?
Zero disengagements for safety.
I had to provide two inputs to get the car to make left turns, just tapping on the accelerator. Might could have waited, not likely.
I adjusted speed up to limit +5 mph regularly.
One passenger grabbed the handles for an aggressive entry on a curve, car slowed way down and handled it well.
Car stayed on the road, adjusted speed, used turn signals, obeyed the laws. It fussed at me to pay attention once when I looked over my shoulder. 4 way stops good. Traffic lights good. Right and left turns were slow and cautious. Stone driveways not so good!
Perhaps there needs to be a totally separate ongoing discussion for TSLA FSD experience and value?
Pale Blue Dot
-
- Posts: 349
- Joined: Fri Sep 03, 2021 1:26 pm
Re: TSLA: What Changed?
Is this Street FSD beta or Autopilot? Glad to hear you had good experience.4nursebee wrote: ↑Tue Jan 25, 2022 2:13 pm Lunch today: Left home, put features on once on the road. 20 min trip to destination through 3 very small towns. Stopped at a friends house, then 10 min trip through small but larger town to restaurant, then 20 min trip to home through 1 or 2 small towns. All features running.
Zero disengagements for safety.
I had to provide two inputs to get the car to make left turns, just tapping on the accelerator. Might could have waited, not likely.
I adjusted speed up to limit +5 mph regularly.
One passenger grabbed the handles for an aggressive entry on a curve, car slowed way down and handled it well.
Car stayed on the road, adjusted speed, used turn signals, obeyed the laws. It fussed at me to pay attention once when I looked over my shoulder. 4 way stops good. Traffic lights good. Right and left turns were slow and cautious. Stone driveways not so good!
Perhaps there needs to be a totally separate ongoing discussion for TSLA FSD experience and value?
No individual stocks.
Re: TSLA: What Changed?
We purchased FSD when we bought the car. The features I described don't work on just AP.impatientInv wrote: ↑Tue Jan 25, 2022 5:21 pmIs this Street FSD beta or Autopilot? Glad to hear you had good experience.4nursebee wrote: ↑Tue Jan 25, 2022 2:13 pm Lunch today: Left home, put features on once on the road. 20 min trip to destination through 3 very small towns. Stopped at a friends house, then 10 min trip through small but larger town to restaurant, then 20 min trip to home through 1 or 2 small towns. All features running.
Zero disengagements for safety.
I had to provide two inputs to get the car to make left turns, just tapping on the accelerator. Might could have waited, not likely.
I adjusted speed up to limit +5 mph regularly.
One passenger grabbed the handles for an aggressive entry on a curve, car slowed way down and handled it well.
Car stayed on the road, adjusted speed, used turn signals, obeyed the laws. It fussed at me to pay attention once when I looked over my shoulder. 4 way stops good. Traffic lights good. Right and left turns were slow and cautious. Stone driveways not so good!
Perhaps there needs to be a totally separate ongoing discussion for TSLA FSD experience and value?
Pale Blue Dot
-
- Posts: 349
- Joined: Fri Sep 03, 2021 1:26 pm
Re: TSLA: What Changed?
Is this FSD beta 10.9? What driving score did you have to be selected for FSD beta?4nursebee wrote: ↑Wed Jan 26, 2022 3:29 amWe purchased FSD when we bought the car. The features I described don't work on just AP.impatientInv wrote: ↑Tue Jan 25, 2022 5:21 pmIs this Street FSD beta or Autopilot? Glad to hear you had good experience.4nursebee wrote: ↑Tue Jan 25, 2022 2:13 pm Lunch today: Left home, put features on once on the road. 20 min trip to destination through 3 very small towns. Stopped at a friends house, then 10 min trip through small but larger town to restaurant, then 20 min trip to home through 1 or 2 small towns. All features running.
Zero disengagements for safety.
I had to provide two inputs to get the car to make left turns, just tapping on the accelerator. Might could have waited, not likely.
I adjusted speed up to limit +5 mph regularly.
One passenger grabbed the handles for an aggressive entry on a curve, car slowed way down and handled it well.
Car stayed on the road, adjusted speed, used turn signals, obeyed the laws. It fussed at me to pay attention once when I looked over my shoulder. 4 way stops good. Traffic lights good. Right and left turns were slow and cautious. Stone driveways not so good!
Perhaps there needs to be a totally separate ongoing discussion for TSLA FSD experience and value?
I have not driven FSD beta myself, don't own a Tesla. I own a plug-in Hybrid, though I have driven various EVs over many years owned by a company I worked for.
My knowledge of FSD beta is based on videos, reports in Consumer Report,NYT, LA Times, CNBC etc and posts/articles by professors/academics from CMU/Harvard etc.
No individual stocks.