-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
What the chances Tesla cars will be self driving in 3 years? Why do you think that way?
- Thread starter DanCar
- Start date
Although I don't think it will be FSD, hope it is, I do think we will have excellent level 3 driverless in a couple of years on the freeway.
A history of Elon self driving predictions:
2015: Elon Musk Says Tesla Vehicles Will Drive Themselves in Two Years - Tesla Vehicles Will Drive Themselves in Two Years
2016: Elon Musk on Twitter - Tesla driving itself coast to coast in two years
2017: Elon Musk clarifies Tesla’s plan for level 5 fully autonomous driving: 2 years away from sleeping in the car - Full self driving, level 5, two years away
2018: Tesla CEO Elon Musk: ‘self-driving will encompass all modes of driving by the end of next year’ - self-driving will encompass all modes of driving by the end of the next year.
2015: Elon Musk Says Tesla Vehicles Will Drive Themselves in Two Years - Tesla Vehicles Will Drive Themselves in Two Years
2016: Elon Musk on Twitter - Tesla driving itself coast to coast in two years
2017: Elon Musk clarifies Tesla’s plan for level 5 fully autonomous driving: 2 years away from sleeping in the car - Full self driving, level 5, two years away
2018: Tesla CEO Elon Musk: ‘self-driving will encompass all modes of driving by the end of next year’ - self-driving will encompass all modes of driving by the end of the next year.
diplomat33
Average guy who loves autonomous vehicles
Yes but look at what has happened recently:
1) AP3 computer now in production, dubbed "FSD computer".
2) AP3 computer has 4 NN accelerators and is only at 10% compute load running NOA with full redundancy.
3) AP3 can process full resolution images from all 8 cameras with no cropping.
4) AP3 upgrades for Model 3's will start "in a few months"
5) FSD price will increase dramatically after May 1.
6) Tesla expands FSD early access program.
7) Tesla announced April 22 event to show real test drives of FSD.
8) Tesla announces post-lease Model 3's will be used for "full autonomy" Tesla Network.
All signs point to Tesla making big progress on FSD with AP3.
These are low resolution cameras. If increase cameras to 4K resolution CPU demand will go up by 4x. Higher resolution cameras are helpful to determine hand gestures, and partially occluded objects like a child coming out from behind a bush or a minivan.
> 2) AP3 computer has 4 NN accelerators and is only at 10% compute load running NOA with full redundancy.
Wonder how much compute load will be used when it stops braking for shadows. Long ways to go.
Last edited:
The more I read you @diplomat33...
...the more I think Musk/FSD is not the hill *I* would choose to die on.
But you do you.
Shock-On-T
⚡️⚡️⚡️⚡️⚡️
To be fair:
2015: Elon wrong
2016: Elon could be out by only a year if it happens this year
2017: Elon could be right if it happens this year (unlikely
2018: Elon could be right if it happens this year (unlikely
diplomat33
Average guy who loves autonomous vehicles
Considering that my prediction was L3 in 3 years and I am on record as saying that I don't think Tesla will get to L4, I am not sure I would say that I am "dying" on the Musk/FSD hill. But I do think we will get some exciting progress this year with AP3.
I’m referring to the way you refer to recent Elon Musk quotes as your evidence for your conclusions. I doubt that would be my way, no matter my opinion on where Tesla’s FSD might be headed.
It really is rather odd to see you list bunch of stuff Elon Musk said as an indication of something big happening. But maybe it is just me. I can’t take the guy seriously anymore on this.
diplomat33
Average guy who loves autonomous vehicles
Uh, my 8 point bullet point specifically avoided Musk FSD quotes or predictions for this very reason. Instead I tried to stick with facts or things Tesla has officially announced.
Last edited:
Most of the eight points you listed is so far only based on Elon Musk’s announcements.
diplomat33
Average guy who loves autonomous vehicles
Well. Let’s look at every one from the point of your conclusion:
Do they point to that?
This may be factual in the sense that it is its name, but this is just a PR move. It is not pointing at anything other than Musk liking to use big words. This points at nothing though of course the hardware progress is probably good. Who knows about the state of the software though.
So far nothing more than big words from Musk, especially the 10% part. I do believe it has 4 NN accelerators. What they are running and when is another question.
Perhaps, but when and what is the difference? Also, source Musk.
Musk timeline, and vague at that, probably the most worthless point here.
Musk demand lever, likely to change or be reveresed... but most importantly not likely to be motivated by progress but by said need for a demand lever. Tesla needs buyers for FSD today, it seems.
Again related to a the demand lever of generating cash by discounting FSD. Also so far mostly a Musk-tweet pipedream. I’m guessing this one may eventually happen but again not proof of progress but something to appease customers after the cash-raise FSD discount.
This is the most tangible one, we shall see what the content is. Of course widely agreed to be a stock-market move mostly as even Tesla says any features are only coming late this year (and again that is Musk estimate).
This one looks like a PR move so far. Tesla can easily reverse course later on this one or just forget about it. I mean Tesla advertised Tesla Network back in 2016. They just silently forgot about the ”next year” announcement both for 2017 and 2018.
I wouldn’t put it past Tesla that the reason for this was more related to some leasing/financing math (and/or PR reasons) than anything to do with the Tesla Network.
Last edited:
wesley888
Member
3 years is a very long time, especially now that they actually have a FSD computer that's capable of doing it. I would say most certainly will happen in 3 years.
Interesting question.
Back in 2014, how many of us would pick the right answer to:
What are the chances that Spacex will successfully re-fly a previously flown and recovered orbital booster within three years?
People with extensive knowledge of rocketry would be more likely to be pessimistic.
Similar problem for our resident NN/ smart systems experts.
Back in 2014, how many of us would pick the right answer to:
What are the chances that Spacex will successfully re-fly a previously flown and recovered orbital booster within three years?
People with extensive knowledge of rocketry would be more likely to be pessimistic.
Similar problem for our resident NN/ smart systems experts.
Last edited:
There is this fool me once problem here.
Betting against Tesla making a great BEV or SpaceX succeeding with their reusable booster sure would seem silly now. This is also a reason why so many did believe Musk and Tesla on EAP and FSD back in 2016. Unfortunately that one turned out to be a ”fool me once” episode.
Is Tesla’s BEV success and SpaceX’s success really indicative of Tesla’s FSD success? Or is the fate of Tesla’s FSD a completely separate question?
Mo City
Active Member
If only the current sensor suite was adequate for legit FSD.
What do you think are the chances that within three years we will be told you need hardware 4 for FSD? The problem with Elon's statements about FSD is that he has been trying to change the definition of FSD to mean, yeah the car will drive itself but you still have to be vigilant. That is N O T FSD.
Similar threads
- Poll
