Musk endeavours

[NotSure]

As things are no one knows if Ai will be capable of driving cars or not.

Quite.

I can recall being told by a senior Don in Oxford physics who had spend his career in computers and electronics for various space missions, that computers would never beat humans at chess.

Chess is really easy - it's a very closed problem. Solving that was largely down to Moore's Law, which a senior don should have been able to predict.

Then I was informed by similar educated experts that AlphaGo could never beat Lee Sedol, but it did.

While Go is a bit like chess - there are a closed an finite set of possible scenarios - but Moore's Law alone was insufficient. There are just too many possible scenarios. The success of AlphaGo took many by surprise and was a real game changer - China almost instantly devoted a good chunk of GDP to machine learning/'artificial intelligence' research.

Can you provide (a) reason(s) why AI will never be able to drive as well as humans?

No. I never said that, I just pointed out the challenges were very considerable. An analogy might be "what would AlphaGo do if someone suddenly inserted a red piece"? It would behave in an utterly unpredictable way, as it had never seen a situation like that before. Driving a vehicle on roads full of unpredictable vehicles and weather is a long way from any board game, where people always follow the rules, and take it in turn.

But saying it may be possible is a long way from saying ML/AI has 'no limitations'.

[odysseus2000]

Yes, Chess can be played via brute force to exceed human levels, but the most interesting demonstration was when Alphazero learned how to play chess in 4 hours and then became able to defeat computational approaches:

https://arstechnica.com/gaming/2017/12/ ... -overlord/

According to the analysis of one game, the software package Stockfish-8 looks at 70 million positions per second whereas Alphazero only looks at 80,000 per second:

https://www.youtube.com/watch?v=0g9SlVdv1PY

In the competition of 100 games, Alphazero had 28 wins and 72 draws.

All of this suggests to me that techniques of Alphazero are a substantial step up from anything possible before.

One can argue that any software can be thrown by the introduction of something it has never seen before which is a big supporter of the Tesla claim that they have the best data both in terms of quality and volume. At some point only some extraordinary unlikely event will phase the AI software, similar to how an experienced motorist can cope with almost anything as he or she has seen such outlier events before or will not cope if it is something never before experienced.

As I look at the field it seems to me that if optical self driving can not be made many times super human then there is some other aspect to the human mind that is not currently known. Without that caveat it looks to be just a matter of time and refinement of pattern recognition matrix, machine labelling and such before cars will drive better than humans at a level that will be recognised and approved for use. If that moment comes the pressure to remove human licenses to drive will begin to grow.

Regards,

[onthemove]

According to Wiki:

https://en.wikipedia.org/wiki/Motor_veh ... S._by_year

there were in 2019 36,120 deaths on US roads = 36120/365 = 98 deaths per day. Yet a Tesla accident where apparently the driver was not in control of the vehicle, in violation of Tesla recommendations, becomes from page news, the other on average 97 deaths do not.

The measurements from Tesla show an accident rate with human combined with autopilot of approximately 10x less accidents based on distance between accidents. If we equate accidents to deaths then if the entire US fleet was driving with autopilot the death rate could be reduced to about 10 per day. This would be a very significant reduction in human misery and economic costs.

Whether complete auto pilot (no human driver) happens we still have a technology that has been demonstrated to be better than human alone driving by close to one order of magnitude.

The neural net technology used by Tesla came into existence about 10 years ago and there is no data/theory that I am aware of which limits this technology. Unless some currently unknown limit to the technology emerges, we can expect the software to continually improve, especially as the labelling of images is now being done automatically.

It is easy to argue that the technology will never reach a mass market and be allowed by regulators based on previous experiences and prejudice, but that is not a scientific way to think about new technologies.

Regards,

But 'self driving' and 'full self driving' aren't promising 'driver assistance'.

They're the promise of driver replacement - have a nap or read a book while the car gets you home through the rush hour traffic jams.

While I don't strictly agree with the exact claim that 'there are no limits to the technology', I do firmly (as someone qualified in the field of AI) believe the current state of the art is perfectly capable of producing a self driving car (without the need for human in the driver seat) that can drive better than a human would.

I just don't believe that Tesla are anywhere as near perfecting it (or even just making it acceptable for a first generation) as quite a few people believe, and I wouldn't be surprised if Tesla find themselves wrong-footed on their choice of hardware.

i.e. I agree...

"The neural net technology used by Tesla came into existence about 10 years ago and there is no data/theory"

So on what basis can Tesla know that the hardware computing capabilities they are putting into their cars for which they seem to be promising a 'software only' upgrade to FSD, will actually be up to the job? I don't believe they can know it until they actually get a proven, working FSD using it. And they don't have that yet. So how can they know that their hardware will be adequate?

I think a lot of the Tesla fans misjudge the comparison with Waymo. A lot of Tesla fans seem to think that 'driver-assisted-but-not-geofenced' is somehow better than 'no-driver-but-geofenced'.

I think those people misunderstand the effort involved in those final few 'tweaks'. It's not unknown for development attempts to completely hit a brick wall (metaphorically speaking, no pun intended) when trying to go that final distance. It can be like when you're a kid trying to reach under a fence to retrieve something... you stretch your arm as much as you can and you can just about touch what you're reaching for ... but you can't quite get the extra distance you need to grab it. You have no option but to withdraw your arm and either go and ask your neighbour, or at least find an additional tool to get you the final distance.

It's all about the foundation. Is it robust enough. Is it expressive enough. Is it powerful enough. There is no data / theory, so how can Tesla know for sure?

Waymo, having a geofenced-but-no-driver service running, have proven that they can go all the way (to FSD), albeit in limited circumstances.

Telsa on the other hand have gone for a broad-based, lots of foundation, but haven't gone the whole way (to FSD) in anything - nowhere have they proven that the foundation is fit to support the ultimate aims.

Think of it like building a city of skyscrapers.

Waymo have built one skyscraper all the way to completion. Waymo have proven that skyscraper's foundation can support it. They done full stack beginning to end for at least one stack. It's true that the other stacks won't be exact replicas and may even be 'taller'. But don't underestimate the significance of proving that one stack all the way to real world implementation, developed far enough to completely remove the driver from the driving seat. That last leap from safety driver to no safety driver was no small step for mankind!

Telsa on the other hand have put multiple half built skyscrapers out. None of them have reached their final height. It's not yet proven whether Telsa's foundation will be able to support the final skyscrapers. There's only two things to driving - speed and direction, and just watch the Tesla videos, and you regularly see the driver dialing up or down the speed on autopilot.

Unlike real skyscrapers where the engineering and theory is well established, like you yourself say, with the neural net technology... "there is no data / theory". There's no data or theory that you can look to when you pour the cement of the foundations (or sell lots of cars with the supposed 'final' hardware already in place)

One final thought.

A lot of Tesla fans proclaim about how Tesla are getting huge amounts of real world data from their cars which Waymo with its limited fleet aren't able to match.

I saw a video a few months ago where a Telsa representative (giving the speech) put up a slide, proclaiming about how many millions Tesla had vs Waymo. If forget the exact numbers, but obviously being a Tesla presentation the Tesla miles driven was a couple of orders of magnitude higher than Waymo.

I couldn't help think to myself... yes, but ...

Waymo = "Keep hands off the steering wheel"
Tesla = "Keep hands on the steering wheel"

Waymo might have less data, but they seem to be getting far more out of the data that they do have.

If you're interested, there's a guy with a channel largely dedicated to riding in Waymo cars - he isn't paid by Waymo, and is paying to ride just like any other public riders.

https://www.youtube.com/c/JJRicksStudios/videos

But be warned.

Compared to the Tesla videos they can be quite boring, because, well, the Waymo cars tend to just, how can I put this, "work" without any feeling of will it or won't it. They don't quite have the feeling of jeopardy that you get with the Tesla videos.

But they can do some pretty smart stuff. Identifying and reacting appropriately to emergency vehicles, road works, etc.

I mean, put it this way... it's a huge leap from safety driver to no safety driver. To remove the safety driver and put public passengers in the car, and allow it to drive in supermarket car parks right up at the front where people are walking in and out of the store all the time, you really do have to believe it will acceptably handle all situations it's ever likely to encounter. And that's no mean feat! Don't underestimate the significance of that.

[odysseus2000]

Some excellent points by onthemove.

As I see things there is no fundamental reason why the Tesla system can't be a lot better driver than a human, but there are likely a lot of small incremental things that have to be found and smoothly integrated to have a reliable system.

I have no idea if Tesla will succeed, but betting against human ingenuity when there are very big prizes on offer has always not been an easy call.

Similarly knowing what is needed to make a technology practical has always been difficult to know until someone does it and then all manner of clones and simplifications are found, just because researchers now know that it can be done.

If Tesla do succeed the share price goes up a lot, if they fail it likely does not go down that much as few, save for Ark, believe that Robotaxis are anything but a very long term possibility.

We shall see.

Regards,

[odysseus2000]

This is an interesting episode with Munro interviewing Alex Guberman who has just bought an ID4.

There are some interesting snippets, including that Munro has been contacted by various German engineers who agreed with this criticism. It seems that within VW there are the ICE folk who argue they are paying the wages of the BEV guys and who make some requirements that the BEV team have to suffer.

A slightly rambling podcast but interesting for background and for Guberman's defence of the ID4 (run time 27.07 minutes):

https://youtu.be/m4zu46ojkbo

Regards,

[NotSure]

Yes, Chess can be played via brute force to exceed human levels, but the most interesting demonstration was when Alphazero learned how to play chess in 4 hours and then became able to defeat computational approaches:

AlphaGo was indeed a huge breakthrough. It particularly made the Chinese sit bolt upright and pay attention, since Go expertise is highly valued in their culture. The first version of Alpha was very clever, the second almost scarily so - as you say, it became a world expert in Go, from scratch, in hours, with no 'coaching' (the original chess programs were effectively just jammed full of expertise, but quick and never distracted - brute force). AlphaGo used 'deep learning', as I am sure does Tesla.

I am sure that self-driving will be solved one day, I just wouldn't like to say when and by whom. It would be much easier if all cars were self driving - that is more or less a solved problem - but accounting for other vehicles may be an issue for a while.

In addition to unforeseen circumstances, one also has to consider what is known as 'adversarial' situations - someone trying to 'break' the algorithm. For example, a strategically placed sticker on one's face can totally upset some facial recognition software.

Image some prankster with a head lamp mounted in the rear of their pick-up, facing back down the road. In front of the lamp is a variable aperture, like used in cameras. You switch an the lamp, then rapidly open the aperture. To a system largely relying on cameras (like Tesla, no lidars), this would just look like a motor bike rapidly accelerating on the wrong side of the road, an due to impact it in a fraction of a second. What would a human do? What would ML do? I'm sure you can dream up alternative scenarios.

I'm a great admirer of Musk in many ways (of course he is flawed, but so are the most disruptive individuals). I do think though that there is far too much optimism, much of it uninformed, built into the share price. I am a small holder, though not by choice - their market cap means it forms a decent chunk of even global trackers. There is a cycle with disruptive tech - first it is not much discussed, then hyped to the moon, then considered to be worthless, finally finding its true value. I just think we may still be in the second phase with Tesla, and it could turn out to be 'Blackberry' not 'Apple'. If a further breakthrough, like deep learning, emerges, Tesla are back to square one with everyone else. They really should be spending some of those Bitcoins on R&D IMHO (their current R&D budget is $1.5B, so it looks like they prize BTC as much as R&D).

[Howard]

Does Tesla have some quality issues in China?

A woman protested at an auto event on Monday.

"the protestor won widespread sympathy when videos of her spread online. Many users joined in to vent about their Tesla problems. Posts with the hashtag “Tesla stand turned into a stage for defending rights” garnered over 220 million views on Weibo within two days."

regards

Howard

"Tesla mulls cars tailored to China amid mounting criticism"

https://techcrunch.com/2021/04/20/tesla ... s62701VPJd

[odysseus2000]

Does Tesla have some quality issues in China?

A woman protested at an auto event on Monday.

"the protestor won widespread sympathy when videos of her spread online. Many users joined in to vent about their Tesla problems. Posts with the hashtag “Tesla stand turned into a stage for defending rights” garnered over 220 million views on Weibo within two days."

regards

Howard

"Tesla mulls cars tailored to China amid mounting criticism"

https://techcrunch.com/2021/04/20/tesla ... s62701VPJd

The protestor was focused on what she considers a brake fault that lead to her Father's death, but the official investigation blamed her Father's speeding.

She was put in the slammer for 5 days.

Full story in this Tesla daily as well as more info on the Texas crash and other stuff:

https://youtu.be/iX9qQ4HFwJA

Tesla have always said they will make cars for each market and have a Chinese design studio for Chinese variants.

Regards,

[BobbyD]

All hail the King!

Volkswagen ID.4 named World Car of the Year 2021

- https://www.autoexpress.co.uk/volkswage ... -year-2021

[BobbyD]

Bill Nye the Taycan Guy

- https://newsroom.porsche.com/en_US/2021 ... 24229.html

The films are very well made. Less than 2 minutes, use an American icon, accessible, dealing with basic EV concepts and leave you wanting to take a Taycan for a spin....

[BobbyD]

TESLA BLASTED BY CHINESE STATE MEDIA OVER ALLEGED SAFETY ISSUES

TESLA has been attacked by state-run Chinese media over safety issues, after a video of a protester standing on top of one of the electric car maker’s vehicles at the Shanghai Auto Show went viral.

The American company needs to improve its “quality and commitment to users” in order to gain the trust of motorists, according to an article published by Xinhua, China’s official state-run press agency.

A separate opinion piece published in the Xinhua Daily Telegraph criticised the “arrogant response” of Tesla executives to the incident.

- https://www.driving.co.uk/news/business ... ty-issues/

[Howard]

Reading UK Tesla driver/owner threads there is evidence that the Model 3 has a number of basic design flaws which irritate drivers. Erratic indicators, headlights and windscreen wipers are frequently quoted. And OTA updates which often cause more problems than they solve. It's also surprising how many posts there are about drivers purchasing sound deadening kits and finding ways of improving bodywork seals and flaws.

Range issues have also worried Standard Range drivers.

Maybe some of these problems are beginning to influence Chinese drivers.

Our KIA BEV has none of the above faults and, whilst I haven't driven it on a journey of over 100 miles so far, I've been pleasantly surprised by how little the range has dropped despite motorway journeys at 70 mph. Sadly, I know these are just real world experiences and pale into insignificance compared with the enthusiasm of Tesla fans who haven't yet owned/driven a PHEV or BEV.

regards

Howard

[BobbyD]

Reading UK Tesla driver/owner threads there is evidence that the Model 3 has a number of basic design flaws which irritate drivers. Erratic indicators, headlights and windscreen wipers are frequently quoted. And OTA updates which often cause more problems than they solve. It's also surprising how many posts there are about drivers purchasing sound deadening kits and finding ways of improving bodywork seals and flaws.

Range issues have also worried Standard Range drivers.

Maybe some of these problems are beginning to influence Chinese drivers.

Our KIA BEV has none of the above faults and, whilst I haven't driven it on a journey of over 100 miles so far, I've been pleasantly surprised by how little the range has dropped despite motorway journeys at 70 mph. Sadly, I know these are just real world experiences and pale into insignificance compared with the enthusiasm of Tesla fans who haven't yet owned/driven a PHEV or BEV.

regards

Howard

To be fair, whilst I've seen all those complaints and the rock solid riposte 'mine doesn't do that you are a lying troll' once you're facing off against the Chinese government, in China, the rights and wrongs of the situation become somewhat academic...

It's interesting watching cars regularly exceed their EPA at motorway speeds in real world tests... Obviously the laws of physics determine that you would go further were you travelling slower, but there is a growing selection of capable BEV's which won't scare away the Joneses making it to market.

[odysseus2000]

All hail the King!

Volkswagen ID.4 named World Car of the Year 2021

- https://www.autoexpress.co.uk/volkswage ... -year-2021

Have VW really thought through the ID4?

They have launched a 20th century car into the 21st century, full of teething and associated troubles with cheapo rear hub brakes and no doubt some punters will be encouraged to buy one via the huge marketing spend and brown paper packages of money left for the judges of the World Car of the year?

In the by and by many of these punters will realise they have been had and will be upset that they have a second rate cheap car, not the embodiment of 21st century driving they had been lead to believe that they were getting.

Sure VW will fool some of the folk, but they are setting themselves up for a lot of unhappy punters.

Regards,

[dealtn]

All hail the King!

Volkswagen ID.4 named World Car of the Year 2021

- https://www.autoexpress.co.uk/volkswage ... -year-2021

Have VW really thought through the ID4?

They have launched a 20th century car into the 21st century, full of teething and associated troubles with cheapo rear hub brakes and no doubt some punters will be encouraged to buy one via the huge marketing spend and brown paper packages of money left for the judges of the World Car of the year?

In the by and by many of these punters will realise they have been had and will be upset that they have a second rate cheap car, not the embodiment of 21st century driving they had been lead to believe that they were getting.

Sure VW will fool some of the folk, but they are setting themselves up for a lot of unhappy punters.

Regards,

Can you actually remember what cars were like 22 years ago? Your continuous reference to last century grows increasingly bizarre, or desperate.

[odysseus2000]

dealtn
Can you actually remember what cars were like 22 years ago? Your continuous reference to last century grows increasingly bizarre, or desperate.

Yes, and should I forget I can look at the ID4.

The ID4 has been deliberately made to resemble a 20th century car as came out clearly in the Munro investigations and he got emails from VW engineers noting how they were unhappy with they had been ordered to design, but that they were controlled by the ICE establishment in VW who do not want the ID4 to be a huge success as it would hurt their empire.

One can argue that VW are attempting to make a transition car for ICE drivers to become BEV drivers which is a recipe for trouble as someone splashing out for a new car or leasing, expects a new car taking advantage of the latest developments, not some Frankenstein mix of old and new.

I have no doubt that the VW advertising machine will sell ID4's but if I was a VW share holder I would be fretting about what kind of image this puts into the VW owners who are getting something that is far inferior to what is possible and currently offered by Tesla.

The clear message from VW is that they have not given up on ICE cars and are doing what ever is needed to make their own BEV only acceptable to a subset of their consumer base.

It is exactly what the UK manufactures did when faced with the technical superior models from Japan and we all know how that went.

Regards,

[murraypaul]

Current early adopters are buying electric cars because they are passionate about owning an electric car.
They will become mass market propositions when people buy them because they are just like any other car, but cheaper to run.

[NotSure]

Current early adopters are buying electric cars because they are passionate about owning an electric car.
They will become mass market propositions when people buy them because they are just like any other car, but cheaper to run.

Teslas are status symbols. There's a good market for status symbol cars, but it is inevitable that Tesla will soon be sharing this space. As such, to justify their market cap, they will surely also need a (good) share of the 'A to B and never goes wrong' market? That was the market that became completely dominated by Japanese cars 'back in the day'. Their success was largely due to very customer-focussed R&D and dramatically raising the bar on reliability, rather than selling to the types who like to drive Mercs and beamers.

[BobbyD]

The ID4 has been deliberately made to resemble a 20th century car as came out clearly in the Munro investigations and he got emails from VW engineers noting how they were unhappy with they had been ordered to design, but that they were controlled by the ICE establishment in VW who do not want the ID4 to be a huge success as it would hurt their empire.

It's deep-ICE man, the truth is out there.

[Howard]

All hail the King!



- https://www.autoexpress.co.uk/volkswage ... -year-2021

Have VW really thought through the ID4?

They have launched a 20th century car into the 21st century, full of teething and associated troubles with cheapo rear hub brakes and no doubt some punters will be encouraged to buy one via the huge marketing spend and brown paper packages of money left for the judges of the World Car of the year?

In the by and by many of these punters will realise they have been had and will be upset that they have a second rate cheap car, not the embodiment of 21st century driving they had been lead to believe that they were getting.

Sure VW will fool some of the folk, but they are setting themselves up for a lot of unhappy punters.

Regards,

Can you actually remember what cars were like 22 years ago? Your continuous reference to last century grows increasingly bizarre, or desperate.

The problem is that Ody is an expert on last century cars. He doesn't have any experience of driving a 21st century car and has to rely on Tesla shareholders' YouTube videos for information.

One day he'll get in a modern car, perhaps a VW, or maybe even a BEV and it will be a revelation.

regards

Howard