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

Elon: Maybe update first week of March

This site may earn commission on affiliate links.
I wonder if Elon actually believes what he says or just continues to play the con man?

If the former he is seriously out of touch with reality. That would not surprise me - I doubt many want to tell the king he isn't wearing any clothes. If the latter, well...

Neither is good.
I think its a mix of both. The lack of transparency is the problem. For example, the rewrite. He mentioned that a year and a half ago that that had been working on it, but never mentioned it before, so they clearly knew the initial method wouldn't work.

But I thijk now that eh believes they are on the correct path, he believes the time frame is accurate
 
Is L5 so important, though? I mean, what do you get with L5?

-- Robotaxis, aka "My Tesla has been trashed by passengers in only 24 hours" or "wait .. my insurance is HOW much???".
-- Ability to sleep on long journeys or on way to work, aka "I have to live further from work now and get up so early to sleep in car on the way."
-- Ability for car to drive to pick you up, aka "Damn my partner summoned the car again! Now I have to wait an hour for it to get back here to take me home."

But what do you get with L3/L4?

-- FAR fewer accidents as humans are taken out of the driving equation.
-- Lower insurance rates thanks to cars avoiding more and more collisions.

Sure, L5 is sexy, and one day will make manual driving as unusual as a stick-shift, but the big benefits it seems to me (safety) and mostly L3/L4.
 
He already told us since 2015 what he believes fully autonomy or complete autonomy consists of and what he thinks Level 5 means.

Level 5 = "no geofence", "cross country summon".
Complete / Full autonomy = "1 million robotaxis with no one in them", "look outside your window", "safe to fall asleep and wake up at their destination", "human intervention will decrease safety"
Put two and two together, its quite easy.
Elon is like the Bible (or any religious text).

You can pick and choose what you want, to interpret the way you want or prove a point you want to prove ;)

But, thanks for all the quotes, i’ll take a close look when I get some time.
 
Is L5 so important, though? I mean, what do you get with L5?

-- Robotaxis, aka "My Tesla has been trashed by passengers in only 24 hours" or "wait .. my insurance is HOW much???".
-- Ability to sleep on long journeys or on way to work, aka "I have to live further from work now and get up so early to sleep in car on the way."
-- Ability for car to drive to pick you up, aka "Damn my partner summoned the car again! Now I have to wait an hour for it to get back here to take me home."

Not sure why you are mentioning these for L5. These are not examples unique to L5. These would all be true for L4 as well.

But what do you get with L3/L4?

-- FAR fewer accidents as humans are taken out of the driving equation.
-- Lower insurance rates thanks to cars avoiding more and more collisions.

That is not what you get with L3/L4. L3 still requires a driver so the first one is not an example of L3.

Lower insurance rates could also be true for L4 or L5.

I don't see why L4 would have better insurance rates than L5. L4 is not inherently safer than L5.

Sure, L5 is sexy, and one day will make manual driving as unusual as a stick-shift, but the big benefits it seems to me (safety) and mostly L3/L4.

I do agree that L5 might not be as important as L4 but not for the examples you mention.

I think a better reason why L5 might not be as important as we think, is because statistically, most humans drive pretty close to home. Statistically, the average American only drives about 30 miles per day. So I think it could be argued that a L4 robotaxi that is geofenced to your city where you live, would be "good enough" for most daily driving. L5 that allows you to drive anywhere in the US is not as important since it would be pretty rare that you would need to drive cross country.
 
  • Like
Reactions: Microterf
I think a better reason why L5 might not be as important as we think, is because statistically, most humans drive pretty close to home. Statistically, the average American only drives about 30 miles per day. So I think it could be argued that a L4 robotaxi that is geofenced to your city where you live, would be "good enough" for most daily driving. L5 that allows you to drive anywhere in the US is not as important since it would be pretty rare that you would need to drive cross country.

I wonder how many people would really use L3-L5 because they can't be bothered to drive themselves. If you give me in a car that is declared 2x better than the average human, I am not going to let it drive me, my family, nor risk my insurance premium by having it drive somewhere by itself. Not worth it. It had better be much much better than that.
 
That is not what you get with L3/L4. L3 still requires a driver so the first one is not an example of L3.

Lower insurance rates could also be true for L4 or L5.

I don't see why L4 would have better insurance rates than L5. L4 is not inherently safer than L5.

I didn't say they were exclusive to L3, just that you get the benefit starting at L3. And while L3 requires a driver, it can still seriously lower accident rates while it is engaged, so I think it is an L3 benefit.

My was simply that L3 gets us a long way along the autonomous path, and many of the major benefits.
 
I didn't say they were exclusive to L3, just that you get the benefit starting at L3. And while L3 requires a driver, it can still seriously lower accident rates while it is engaged, so I think it is an L3 benefit.

My was simply that L3 gets us a long way along the autonomous path, and many of the major benefits.

Thanks for clarifying. Right now, L3 is pretty useless IMO since the only L3 systems we have so far can only be used in traffic jams on the highway. So I am not sure how many accidents they would actually prevent. But L4 and L5 could prevent a lot of accidents as soon as they become significantly better drivers than humans.

I wonder how many people would really use L3-L5 because they can't be bothered to drive themselves. If you give me in a car that is declared 2x better than the average human, I am not going to let it drive me, my family, nor risk my insurance premium by having it drive somewhere by itself. Not worth it. It had better be much much better than that.

I would like to see autonomous cars be a lot better than 2x safer than humans. However, I could see instances where only 2x safer would still be worth it. For example, if I am drunk at a party and need to get home, getting a robotaxi to take me home that is "only" 2x safer than human driving would still be much better than not being able to get home or trying to drive home drunk.
 
I wonder how many people would really use L3-L5 because they can't be bothered to drive themselves. If you give me in a car that is declared 2x better than the average human, I am not going to let it drive me, my family, nor risk my insurance premium by having it drive somewhere by itself. Not worth it. It had better be much much better than that.

L3 I would never trust so lets throw that one into the trash.

L4-L5 won't risk your insurance because the entire point of those levels is for the car to take responsibility for the entire drive, and not you. If it crashes its all on the car.

As to how many people would really use L4 to L5 I would argue that almost everyone who uses AP/NoA would benefit safety wise to move from L2 driving to L4-L5 driving. The reason for that it is it's really difficult for a human being to oversee a system when that system works really well. It's hard to maintain the same level of situational awareness. Scientist have done study after study on this kind of thing, and they all show that humans get bored and their mind wanders. Or they just fall asleep.

I would also argue that pretty much anyone who uses lyft/uber would also use L4-L5. It being 2x better than an average human means that it's likely on par with uber/lyft drivers.

With that being said I highly doubt 2x better than a human driver would ever get released as an L4-L5 system. There is so much liability, and resistance towards autonomous vehicles that it's likely going to have to be a lot better.

What I hope to see with autonomous driving is a push for humans to get better at driving or taking the license away. Like why was Tiger Woods even allowed to drive? Why didn't he lose his license a bunch of times before his last crash.

There really isn't an "average" driver just like most "accidents" aren't really accidents at all.

We have a huge number of drivers doing really dumb things that account for the majority of the accidents.

Things like running red lights
Excessive speeds (like massive differentials of speed)
Road rage
impaired driving
drowsy driving

Humans do so many stupid things that I would prefer there wasn't a comparison at all. Instead I'd rather see the approval process consist of thousands of simulated incidents followed by extensive on-road testing. Then once it was "licensed" there was the expectation that incidents would happen, and corrective action would take place. Where the corrective action could be anything from fixing the SW to account for some situation, fixing an infrastructure issue, or blocking the area where it failed from the approved areas list.

In fact I expect the initial parts of autonomous driving to be a bit frustrating where yesterday my car could take me somewhere, but something happened and now no autonomous driving. Where it simply gets grounded.
 
I wonder how many people would really use L3-L5 because they can't be bothered to drive themselves. ...

Most car owners purchased a transportation appliance. They spend most their time behind the wheel doing everything but driving the car. And we have the 'Graying of America' to deal with as well. Both have conspired to put more dangerous drivers on the road each day.

Autonomous systems are the only way we are going to improve auto safety or even maintain the current body count.

It is sad it has come to that, but society is decaying in all regards.
 
  • Love
Reactions: mikes_fsd
Thanks for clarifying. Right now, L3 is pretty useless IMO since the only L3 systems we have so far can only be used in traffic jams on the highway. So I am not sure how many accidents they would actually prevent. But L4 and L5 could prevent a lot of accidents as soon as they become significantly better drivers than humans.

And at this point of course we get into the "what is part of L3" debate .. my feeling is even the FSD beta, when released, will still be L3, hence my comments. If you feel it qualifies as L4, then you can mentally re-write my comments to say L4 .. but you get what I'm saying I think, regardless of terminology.
 
And at this point of course we get into the "what is part of L3" debate .. my feeling is even the FSD beta, when released, will still be L3, hence my comments. If you feel it qualifies as L4, then you can mentally re-write my comments to say L4 .. but you get what I'm saying I think, regardless of terminology.

Yeah, I think I get your point. You think we don't really need L5 because something like FSD Beta would provide a lot of benefits like reducing accidents and reducing insurance costs long before being L5. Did I get that right?
 
  • Like
Reactions: drtimhill
Elon has been wrong for the past 6 years on this topic: 2015 - 2020. He will continue to be wrong for rest of the decade.

I wonder if Elon actually believes what he says or just continues to play the con man?

Elon is a serial optimist. Most likely a super optimist. yes, I like how that rolls off the tongue... Super optimist.
Reminds me of Wile E. Coyote.
Most inventors and entrepreneurs are. They have to be, because who in their right mind would even start something that they want to do.
Elon may be worst than most... an EV company AND a space rocket company AND a tunneling company... Come on, man be realistic!

I have worked for people like him. Their favorite line is: "How hard can this be?"
They are great visionaries (like Steve Jobs was) and they do some incredible things.
Remember Steve's quote in 2007 saying "this will revolutionize our world" when introducing the Iphone...
He saw the future (I did not until much, much later). Some people just see things differently than others.
Jacques Cousteau was an aviator but a auto crash while going to attend a friend's wedding derailed those plans.
Eventually he re-focused his energies to underwater instead of above water and helped to invent the Aqua-Lung and do many, many cool things.
It is this diversity that brings about new ideas and products and makes this an interesting time.
This should be celebrated. Remember, if man was meant to fly, he would have had wings!

I think its a mix of both. The lack of transparency is the problem. For example, the rewrite. He mentioned that a year and a half ago that that had been working on it, but never mentioned it before, so they clearly knew the initial method wouldn't work.

Much like how Space X works, they try something and fix the issues. With FSD, I have heard we'll have self driving cars since the early 1980's. Most people don't realize how complex driving is because it is now simple for them. Humans are great at filtering out unneeded information. Computers are not. They have to be trained what to filter and what not to filter. When you are teaching your kids how to drive, hopefully you mention that in a neighborhood environment, a bouncing ball is extremely dangerous situation. That is because likely there is a kid or dog chasing that ball. Every sign (yard sale, speed limit, Yield, etc...) needs to get processed and filtered as needing more attention or not.

So I disagree that the "clearly knew the initial method would not work". They discovered the issues and are now working on a better solution. You need both a good ideas and also a lot of hard work and a unhealthy dose of optimism to succeed.
How many times did Wernher von Braun's rockets fail before he helped to design an build the Saturn 5?
 
Autonomous systems are the only way we are going to improve auto safety or even maintain the current body count.

.

100% not true. It's one way, and not even the best one, IMO. It just happens to be the flashiest / trendiest.

For instance, the aging problem. Introduce legislation that requires regular driving tests for drivers over 65. Set legislation for removal of license given a certain number of incidents over that age, much like drunk driving.Yes, draconian, but way less expensive than autonomous driving.

I don't advocate for this however it's just one way to address the issue you brought up.
 
Last edited: