Autonomous Car Progress

[DanCar]

You really think he doesn't know what L5 means?

He obviously has a problem, being wrong 5 years in the past and 5 years in the future.

 

[willow_hiller]

He obviously has a problem, being wrong 5 years in the past and 5 years in the future.

I think his biggest problem is making promises and predictions in the first place.

Just look at Waymo! They haven't promised anything, and they've kept all of those non-promises.

 

[DanCar]

Just look at Waymo! They haven't promised anything, and they've kept all of those non-promises.

They have, you just haven't seen them. For the last three years they said they would have driverless service the following year. They have failed in my opinion, but others disagree.

 

[diplomat33]

And you doubt his interpretation. You really think he doesn't know what L5 means? He showed cars last year without steering wheel. L5 has been the objective al along and uncontrolled left turns and roundabouts are pretty basic stuff. Perhaps there might be an issue with falling ash due to a vulcano eruption that needs to be addressed that is why he says almost L5.

I don't doubt that Tesla is sincere about L5 as an objective. But I do doubt Elon's interpretation. I do think Tesla has software to make right turns, unprotected left turns and roundabouts and our cars will be able to do a simple commute with minimal driver interventions. So I think Elon is being sincere about our cars being close to being able to do that. But Elon seems to think that is enough to be L5 and making L5 driverless will just be a matter of solving remaining edge cases. If Elon really thinks that, then yes, I think he is oversimplifying L5. Elon has consistently described L5 as just easy "low speed", somewhat hard "intermediate speed" that is basically just traffic light control and turning at intersections and easy "high speed". That's a very simplistic description of L5 autonomous driving.

They have, you just haven't seen them. For the last three years they said they would have driverless service the following year. They have failed in my opinion, but others disagree.

It's a partial promise. Waymo did remove the safety driver for some rides, just not for all rides and of course, the driverless rides are only available in a limited geofenced area. So they did do it, just not enough for most people to consider the promise completely fulfilled.

 

[Dutchie]

From another thread

FWIW I coded my own neural network from scratch and have been coding for a scary 39 years *cracks knuckles*.
Yes, teslas approach is not only the best, but IMHO the ONLY one that can possibly offer a truly universal solution, and the only one that can produce a really safe solution to FSD.

But hell... this stuffs hard. Neural networks become a lot more like art than science. Its a messy *fuzzy* business, and starts to feel more like voodoo than normal code. Despite all the memes, its not that hard to predict how far through a project you are when its normal object-oriented traditional C++, but the minute you have a massive neural net involved... things get way harder to predict.

FWIW my entirely subjective opinion is that they may still be another small rewrite away from FSD. I think the big gear change was karpathy. He understands that you need the code to be mostly neural net, with minimal procedural code. I think thats the only answer, and its a tough one to agree with unless you work on this stuff all day, because it feels SCARy as a coder to effectively cede control of decision making to a neural net nobody can understand.

From what I read, karpathys time at tesla has been one of progressively lobbying for the NN to take over more of the decision making. I think it will make up an overwhelming proportion of all the FSD code eventually.

I dont think we will get FSD until end of 2021 at best. But I am 100% certain nobody will beat tesla to it, or even get close. Persuading trad auto companies to let software control a car is very hard. persuading them to let software *nobody understands* to do it is almost impossible. Its only possible at tesla because elon is a coder with an interest in how brains work.

 

[Dutchie]

I don't doubt that Tesla is sincere about L5 as an objective. But I do doubt Elon's interpretation. I do think Tesla has software to make right turns, unprotected left turns and roundabouts and our cars will be able to do a simple commute with minimal driver interventions. So I think Elon is being sincere about our cars being close to being able to do that. But Elon seems to think that is enough to be L5 and making L5 driverless will just be a matter of solving remaining edge cases. If Elon really thinks that, then yes, I think he is oversimplifying L5. Elon has consistently described L5 as just easy "low speed", somewhat hard "intermediate speed" that is basically just traffic light control and turning at intersections and easy "high speed". That's a very simplistic description of L5 autonomous driving.



It's a partial promise. Waymo did remove the safety driver for some rides, just not for all rides and of course, the driverless rides are only available in a limited geofenced area. So they did do it, just not enough for most people to consider the promise completely fulfilled.

My goodness hundreds of the best minds work on Tesla FSD and you think they are all naive. What can I say..

 

[diplomat33]

My goodness hundreds of the best minds work on Tesla FSD and you think they are all naive. What can I say..

NO! I did not say that. I said Elon might be naive in his description of L5. I never said that everybody working at Tesla is naive in how they do their jobs.

I would also point out that hundreds of the best minds also work at Waymo. I guess they are all wrong and wasting their time?

 

[diplomat33]

It's remarkable that you guys are accusing me of being naive and arrogant yet you all are going around declaring that all the experts working on FSD for years are wrong, that nobody will even come close to FSD, that only Elon/Tesla will achieve FSD and that it will be the best, safest and only viable solution to FSD. WOW!

 

[john5520]

Somehow my message was deleted

What I wanted to say is that Elon's timing may be off but at the end he always delivers

What is astounding to me is that there are hundreds of people working at Tesla on FSD. Elon's time at Tesla is mostly devoted to FSD. And here you find remarks like:" Oh yeah, what about roundabouts or what about uncontrolled left turns?"

I mean guys really, do you really think that someone came into the office this morning and says: "Heh, I was reading this blog and it occurred to me we never thought of uncontrolled left turns. We better start working" or the lady at the coffee machine says: "what about roundabouts....shoot, better start programming". Come on!

People are skeptical because the promises have come and gone and we're still at a shaky L2. If I didn't think he'd deliver a decent product, I would not have invested the 7k. It's just that I'm not expecting L5, which is OK. For me, L3 with the current hardware is mission accomplished, but I'd settle for a solid L2 or L2+ and still not be disappointed. Money well spent.

 

[Dutchie]

NO! I did not say that. I said Elon might be naive in his description of L5. I never said that everybody working at Tesla is naive in how they do their jobs.

I would also point out that hundreds of the best minds also work at Waymo. I guess they are all wrong and wasting their time?

I never talked about Waymo, I am sure they are good at what they do. I just find the Tesla strategy with accumulation of real world data very convincing. That is why invested in FSD for my Model X. I cann't do that with Waymo. I do find a bit laughable to call Elon naive on L5 when this guy has been working with Karpathy et.al on this subject for years.

I know that SpcaeX and Tesla are the most popular employers for graduates, so they can probably attract the best minds.

 

[diplomat33]

I never talked about Waymo, I am sure they are good at what they do. I just find the Tesla strategy with accumulation of real world data very convincing. That is why invested in FSD for my Model X. I cann't do that with Waymo. I do find a bit laughable to call Elon naive on L5 when this guy has been working with Karpathy et.al on this subject for years.

I know that SpcaeX and Tesla are the most popular employers for graduates, so they can probably attract the best minds.

Fair enough. I certainly respect you for putting your money in a company that you believe in. I purchased my Model 3 and paid for FSD because I love Tesla's technology. AP is the #1 reason why I bought my Model 3. I might be skeptical about FSD but I love AP.

 

[linux-works]

I know that SpcaeX and Tesla are the most popular employers for graduates, so they can probably attract the best minds.

uhhh, tesla has a certain attraction, but as someone who works in the car field (very similar to tesla) - I know of no one that WANTS to go work for elon and get underpaid by tesla, overworked by tesla and have to endure the layoffs that are somewhat regular there. its chaotic (from the stories I've heard directly from ex-tesla employees) and unprofessional in so many ways.

I once considered applying there but after hearing enough about what tesla is like from the inside, I changed my mind pretty fast.

they have a good rep from the *outside* but its quite different from the inside. college kids want to work there, but they want a start. it used to be google (not sure how big google is at this point, as a first-job draw, since they have jumped the shark years ago).

but experienced engineers won't put up with elon's BS and the company does not treat people well.

 

[timewasted]

It's remarkable that you guys are accusing me of being naive and arrogant yet you all are going around declaring that all the experts working on FSD for years are wrong, that nobody will even come close to FSD, that only Elon/Tesla will achieve FSD and that it will be the best, safest and only viable solution to FSD. WOW!

A good chunk of the people on here are vying to SEC.

 

[ItsNotAboutTheMoney]

Picture your Model 3 without a steering wheel and pedals, or you sitting in the back seat. Then picture yourself relaxing or sleeping on the way to work and not worrying about anything (people running lights, cutting you off, merging into you, sudden stops etc. Even in bad weather). That's what L5 is supposed to accomplish. Do you think your Model 3 would be capable of that level of automation by the end of the year?

You missed my point. In order to be _really_ good at Level 2 or 3 a car has to be able to do what Level 5 does.
So Tesla working on improving Level 2 or achieving Level 3, _has_ be to working towards Level 5.

The difference is whether companies are working up or working down.

 

[diplomat33]

You missed my point. In order to be _really_ good at Level 2 or 3 a car has to be able to do what Level 5 does.
So Tesla working on improving Level 2 or achieving Level 3, _has_ be to working towards Level 5.

The difference is whether companies are working up or working down.

L2 and L3 are completely different. You cannot group them together.

No, you can do great L2 without doing any work on L5 because L2 and L5 are completely distinct systems. They are not related or connected to each other at all.

 

[Bladerskb]

Here is a full quote from Elon from WAIC in more context:

“I think there are no fundamental challenges remaining for Level 5 autonomy. There are many small problems. And then there’s the challenge of solving all those small problems and then putting the whole system together and just keep addressing the long tail of problems. So you’ll find that you’re able to handle the vast majority of situations. But then there will be something very odd. You have to have a system figure out a train to deal with these odd situations. This is why you need real world situations. Nothing is more complex and weird than the real world,”
Tesla's Elon Musk talks Autopilot and Level 5 Autonomy at China AI conference

I am sorry to sound "anti-Tesla" but that first sentence sounds incredibly naive or arrogant to me. There are really no fundamental challenges remaining for L5? It's all just small problems from here on out? Who knew!

"autonomous driving is a solved problem" - Elon Musk (2015)
Tesla’s Elon Musk Says Autonomous Driving Not All That Hard to Achieve

 

[Bladerskb]

Oh you don't have to worry about that we know you're anti Tesla.
And you've proven time and again that the one naive and arrogant is you, not Musk.

You mean the guy screaming that he will have Level 5 cars with no drivers roaming all across the country in 2017 isn't naive and arrogant?

 

[Bladerskb]

I’m not saying Elon is right here, but I’ve noticed that you seem to base a lot of your comments on what our current version of Autopilot can handle/understand. It’s been shown both in the talk Andrej gave in Feb 2020 and in his CVPR video that what they’re working on internally and what we have are two very different software branches. You can see this in how the birds-eye NN predictor sees the world and how much more accurate it is compared to how our current sees the world (falls apart when it tries to make sense of an entire scene like an intersection).

Not only that, but based on the CVPR video it shows their birds-eye view predictor handling multi-lane intersections and predicting their layouts incredibly well. This was not show in the Feb 2020 talk, so it seems they’ve made significant progress from Feb to June and it only stands to reason that they’re making good progress on other parts of FSD as well since they clearly aren’t standing still. Coupled with Elon’s recent comments about releasing the new Autopilot software stack in 2-4 months with “a lot of new functionality” and this just further reiterates that what we’re running is all based on an old stack that they are barely even working on anymore.

I know you’ve watched all these videos and understand that these things are being worked on, but it seems like you still think Tesla will continue to stick with their L2 Autopilot system. What makes you think that despite the progress they seem to be making?

There is no two seperate branch its a myth that has been going on for a long time to excuse Tesla's AP for not matching Elon's statements and proclamations. Welcome to 2016.

True, but what I’m saying is since they’ve shown they’re not standing still on solving perception, that gives me reason to believe they’re also not standing still regarding the other parts of self driving (driving policy, planning, etc).

My theory is I think they’re leading up to a big reveal of a brand new Autopilot in the next 2-4 months. I don’t think they’ve showed off those other aspects of self driving like Waymo has because then they’d have to reveal how their in development system is working and that would ruin the big reveal that’s in a few months.

To be clear I don’t think that what they reveal will be L5, that would be truly astonishing if it were the case. But I do think we’ll see a completely new Autopilot that will be much more hands off... L4 with supervision at first seems likely, but even that would be a massive leap over what we have now.

How much kool-aid have you drank? Those things ain't healthy for you.

But landing boosters on barges in the middle of the ocean?

Come on, guys! There are hundreds of people at Tesla working on FSD. Most of the time of Elon at Tesla is dedicated to FSD. I have yet to hear anyone who actually is working or worked at Tesla doubt the progress and achieveability of the end goal. Most of the R&D budged is going to FSD.
Do you really think he is making this stuf up and he doesn't know what he is talking about? Do you really think he risks his company due to breach of contract for people who bought cars with FSD? They even increased the price of FSD recently. Do you guys really think they would have done that if they would not have show something soon?

Yes. He's been doing it for years. He has over 100 failed self driving statements and proclamations. Wake the F up!
He uses his falsified claims for marketing purpose to sell cars and garner mind-shares because he knows the media will report on his recent statements and people only read headlines.

For Example:
(Forbes) Elon Musk: Tesla Will Have Level 5 Self-Driving Cars This Year
(VentureBeat)Elon Musk says Tesla is ‘very close’ to level 5 autonomous driving
(Futurism) Elon Musk Brags That Tesla Is "Very Close" to Full, Level 5 Self Driving
(ZDNet) Tesla's Elon Musk: Our fully autonomous cars are now 'very close'
(Fox News) Tesla "very close" to level 5 autonomous driving technology, Musk says

There are over 10,000 Elon articles similar to this about his 100+ proclamations on self driving and level 5. This is why 10/10 laymen i talk to fully believe you can actually fall asleep in a tesla and that its fully self driving. The same way the media created Trump, they created Elon. This time its actually worse. They just regurgitate everything he says, giving him FREE Advertisement to hundreds of millions of laymen people.

I've followed Waymo/Google self driving way longer than Tesla's Autopilot/FSD work.

And I reject them because they've delivered nothing. Their business case for commercializing it is terrible, not scalable and prone to the whims of each local municipality wherever they try to expand into.

In the 5 years of watching Tesla approach this problem, I've seen way more innovation and progress. It's like they took a blank sheet approach to the problem, what a concept!

I guess in a way I can compare it to Boeing versus SpaceX especially with the way the crewed capsules worked for both companies and their demo missions.

If it wasn't clear Waymo is Boeing and the example above.

Let's roll the tapes:

December 2015: "We're going to end up with complete autonomy, and I think we will have complete autonomy in approximately two years."

Elon Musk Says Tesla Vehicles Will Drive Themselves in Two Years

January 2016: "In ~2 years, summon should work anywhere connected by land & not blocked by borders, eg you're in LA and the car is in NY"

Elon Musk on Twitter

June 2016: "I think we are less than two years away from complete autonomy, safer than humans, but regulations should take at least another year," Musk said.

Two years until self-driving cars are on the road – is Elon Musk right?

March 2017: "I think that [you will be able to fall asleep in a tesla] is about two years" -

Transcript of "The future we're building -- and boring"

March 2018: "I think probably by end of next year [end of 2019] self-driving will encompass essentially all modes of driving and be at least 100% to 200% safer than a person."

SXSW 2018

Nov 15, 2018: "Probably technically be able to [self deliver Teslas to customers doors] in about a year then its up to the regulators"

Elon Musk on Twitter

Jan 30 2019: "We need to be at 99.9999..% We need to be extremely reliable. When do we think it is safe for FSD, probably towards the end of this year then its up to the regulators when they will decide to approve that."

Tesla Q4 Earnings Call

Feb 19 2019: "I think we will be feature complete full self driving this year. The car will be able to find you in a parking lot, pick you up, take you all the way to your destination without an intervention this year. This is not a question mark. I'm certain of that. It will be essentially safe to fall asleep and wake up at their destination towards the end of next year"

On the Road to Full Autonomy With Elon Musk — FYI Podcast

April 12th 2019 : "I think it will require detecting hands on wheel for at least six months.... I think this was all really going to be swept, I mean, the system is improving so much, so fast, that this is going to be a moot point very soon. No, in fact, I think it will become very, very quickly, maybe and towards the end this year, but I say, I'd be shocked if not next year, at the latest that having the person, having human intervene will decrease safety. DECREASE! (in response to human supervision and adding driver monitoring system)"

https://www.youtube.com/watch?v=dEv99vxKjVI&feature=emb_title

April 22nd 2019: "We expect to be feature complete in self driving this year, and we expect to be confident enough from our standpoint to say that we think people do not need to touch the wheel and can look out the window sometime probably around the second quarter of next year."

April 22nd 2019: "We expect to have the first operating robot taxi next year with no one in them! One million robot taxis!"

May 9th 2019: "We could have gamed an LA/NY Autopilot journey last year, but when we do it this year, everyone with Tesla Full Self-Driving will be able to do it too"

April 12th 2020: "Robotaxis release/deployment... Functionality still looking good for this year. Regulatory approval is the big unknown.

https://twitter.com/elonmusk/status/1249210220200550405

April 29th 2020: "So I think we could see robotaxis in operation with the network fleet next year, not in all markets but in some."

 

[Bladerskb]

Well, by that definition, my AP1 Model X is self-driving*

* on highways below 90mph.

It is useful to have concrete definitions, otherwise AP1 would satisfy 2 of the 3 criteria - anytime and anyone...

so there is no driver in your ap1 model x from start to finish?

 

[diplomat33]

Here is what I am expecting for maybe the end of this year:
- Beta "turning at intersections" with required driver confirmations.
- Improved FSD visualizations where we see better cross traffic and see intersections laid out on the screen.