Donate to Remove ads

Got a credit card? use our Credit Card & Finance Calculators

Thanks to johnstevens77,Bhoddhisatva,scotia,Anonymous,Cornytiv34, for Donating to support the site

Musk endeavours

The Big Picture Place
odysseus2000
Lemon Half
Posts: 6366
Joined: November 8th, 2016, 11:33 pm
Has thanked: 1535 times
Been thanked: 959 times

Re: Musk endeavours

#453183

Postby odysseus2000 » October 26th, 2021, 12:15 pm

BobbyD
So please show them to us.


As you know these are privately held by Tesla.

The whole idea is to get the system working in a real world environment, just like any beta.

Regards,

odysseus2000
Lemon Half
Posts: 6366
Joined: November 8th, 2016, 11:33 pm
Has thanked: 1535 times
Been thanked: 959 times

Re: Musk endeavours

#453199

Postby odysseus2000 » October 26th, 2021, 1:01 pm

Hertz did not get a discount on the 100,000 cars it ordered:

https://twitter.com/elonmusk/status/145 ... 27625?s=20

Regards,

Adamski
Lemon Quarter
Posts: 1076
Joined: July 13th, 2020, 1:39 pm
Has thanked: 1465 times
Been thanked: 553 times

Re: Musk endeavours

#453237

Postby Adamski » October 26th, 2021, 2:36 pm

Tesla over $1 trillion market cap. Up 31% in past month.

I'd wager unless Tesla gets a cheap £20,000 car on the market, they will go the way of Cisco in 1999-2000 ie bubble followed by crash.

As the other car makers go cheap workable EVs. There's only so many millionaires around to buy 40-55k cars.

Lootman
The full Lemon
Posts: 18684
Joined: November 4th, 2016, 3:58 pm
Has thanked: 628 times
Been thanked: 6564 times

Re: Musk endeavours

#453243

Postby Lootman » October 26th, 2021, 2:45 pm

Adamski wrote:Tesla over $1 trillion market cap. Up 31% in past month.

I'd wager unless Tesla gets a cheap £20,000 car on the market, they will go the way of Cisco in 1999-2000 ie bubble followed by crash.

As the other car makers go cheap workable EVs. There's only so many millionaires around to buy 40-55k cars.

I thought about buying some a few months ago when it was about $600 a share. Given that it closed over $1,000 last night that would have been smart.

But then I realised that, between my holdings in an S&P 500 index fund, Scottish Mortgage IT and the ARKK ETF, I already have about £10,000 invested in Tesla. It's enough for a share that is impossible to value.

BobbyD
Lemon Half
Posts: 7814
Joined: January 22nd, 2017, 2:29 pm
Has thanked: 665 times
Been thanked: 1289 times

Re: Musk endeavours

#453268

Postby BobbyD » October 26th, 2021, 4:16 pm

odysseus2000 wrote:
BobbyD
So please show them to us.


As you know these are privately held by Tesla.


So you can't know that safety score has been properly validated as a method of assessing the fitness of untrained drivers to oversee experimental software on public roads can you?

BobbyD
Lemon Half
Posts: 7814
Joined: January 22nd, 2017, 2:29 pm
Has thanked: 665 times
Been thanked: 1289 times

Re: Musk endeavours

#453276

Postby BobbyD » October 26th, 2021, 4:41 pm

Adamski wrote:Tesla over $1 trillion market cap. Up 31% in past month.

I'd wager unless Tesla gets a cheap £20,000 car on the market, they will go the way of Cisco in 1999-2000 ie bubble followed by crash.

As the other car makers go cheap workable EVs. There's only so many millionaires around to buy 40-55k cars.


Tesla are still working on the 2014 Roadster 2 ands the 2019 Semi so I wouldn't hold your breath.

odysseus2000
Lemon Half
Posts: 6366
Joined: November 8th, 2016, 11:33 pm
Has thanked: 1535 times
Been thanked: 959 times

Re: Musk endeavours

#453301

Postby odysseus2000 » October 26th, 2021, 6:34 pm

BobbyD

So you can't know that safety score has been properly validated as a method of assessing the fitness of untrained drivers to oversee experimental software on public roads can you?


No I can't, but no one asks me what is allowed on UK roads, let alone US roads.

Tesla have shared their approach with US regulators and got the green light. What else does one need to know?

Regards,

BobbyD
Lemon Half
Posts: 7814
Joined: January 22nd, 2017, 2:29 pm
Has thanked: 665 times
Been thanked: 1289 times

Re: Musk endeavours

#453330

Postby BobbyD » October 26th, 2021, 7:36 pm

odysseus2000 wrote:Tesla have shared their approach with US regulators and got the green light. What else does one need to know?


In which case you can point us to US regulators statement about how they have assessed Safety score and found it to be a valid method of assessing whether or not untrained drivers are fit to oversee experimental software on public roads, along with a list of the road and automotive safety experts, the engineers programmers, psychologists and statisticians who were on the panel which validated it.

odysseus2000
Lemon Half
Posts: 6366
Joined: November 8th, 2016, 11:33 pm
Has thanked: 1535 times
Been thanked: 959 times

Re: Musk endeavours

#453352

Postby odysseus2000 » October 26th, 2021, 8:28 pm

BobbyD
In which case you can point us to US regulators statement about how they have assessed Safety score and found it to be a valid method of assessing whether or not untrained drivers are fit to oversee experimental software on public roads, along with a list of the road and automotive safety experts, the engineers programmers, psychologists and statisticians who were on the panel which validated it.


I have no connection with either Tesla or the US regulatory agencies.

How do you think I could have any information on the US regulatory process and private discussions with a manufacturer?

What I know is that Tesla are legally beta testing their hardware and software and from that I conclude that all appropriate procedures have been followed and that Tesla have been given permission for said tests.

No investor with outsider knowledge knows more than the publicly available statements in which Tesla say they are doing beta testing. If this was not leal then Tesla would be prevented from doing such tests.

Regards,

BobbyD
Lemon Half
Posts: 7814
Joined: January 22nd, 2017, 2:29 pm
Has thanked: 665 times
Been thanked: 1289 times

Re: Musk endeavours

#453353

Postby BobbyD » October 26th, 2021, 8:38 pm

odysseus2000 wrote:
BobbyD
In which case you can point us to US regulators statement about how they have assessed Safety score and found it to be a valid method of assessing whether or not untrained drivers are fit to oversee experimental software on public roads, along with a list of the road and automotive safety experts, the engineers programmers, psychologists and statisticians who were on the panel which validated it.


I have no connection with either Tesla or the US regulatory agencies.

How do you think I could have any information on the US regulatory process and private discussions with a manufacturer?


Because you just claimed that

odysseus2000 wrote:Tesla score is backed up by accident statistics.

odysseus2000
Lemon Half
Posts: 6366
Joined: November 8th, 2016, 11:33 pm
Has thanked: 1535 times
Been thanked: 959 times

Re: Musk endeavours

#453359

Postby odysseus2000 » October 26th, 2021, 8:55 pm

BobbyD
Because you just claimed that

odysseus2000 wrote:
Tesla score is backed up by accident statistics.


It is an inference from observed behaviour.

If the accident statistics were terrible, the beta trials would have been made illegal, but they haven't hence the accident statistics must be acceptable to the authorities.

In investment and trading one often does not have specific information and one has to deduce it from what one knows.

Regards,

onthemove
Lemon Slice
Posts: 540
Joined: June 24th, 2017, 4:03 pm
Has thanked: 722 times
Been thanked: 471 times

Re: Musk endeavours

#453423

Postby onthemove » October 26th, 2021, 11:35 pm

odysseus2000 wrote:
BobbyD
Because you just claimed that

odysseus2000 wrote:
Tesla score is backed up by accident statistics.


It is an inference from observed behaviour.

If the accident statistics were terrible, the beta trials would have been made illegal, but they haven't hence the accident statistics must be acceptable to the authorities.

In investment and trading one often does not have specific information and one has to deduce it from what one knows.

Regards,


Tesla seems to have taken the approach that their FSD is just a 'driving aid' and therefore requires the driver to still be in control of the vehicle at all times. So I suspect that has possibly negated much (perhaps not all) need to apply for licences for 'self driving' car testing on public roads. It's officially not a self driving car - not even when the FSD beta is engaged.

As for their accident statistics, it's getting too late in the evening to go searching for the references now, but I seem to recall the national transportation and safety board has quite a lengthy list of incidents involving Tesla's that they're investigating. From what I gather, the regulators already forced Tesla to put more warnings about the driver needing to be in control, and the software checking the hands are on the steering wheel and such like.

Youtube has just recommended this video to me...

https://www.youtube.com/watch?v=QngynvW ... DirtyTesla
"Tesla FSD Beta Emergency Bug Fix Update V 10.3.1 - What Happened?"


Needless to say, the Tesla owner who produced the video is praising how quickly Tesla were to push out a fix.

Personally, I think that speed of deployment is something of concern, not something to be praised.

The self driving stack is quite complex, and there is always the possibility that different layers might interact with each other in unexpected ways in response to even small tweaks.

So I would hope that every (software) build that they produce would undergo an awful lot of formal testing before being rolled out, even to beta testers.

Let's get real here - this software is being used in scenarios where people really could die if it doesn't work appropriately.

I know when I write software on relatively small scale applications in comparison, even the builds my team produce take a while to complete with all the unit tests etc.

So for Tesla to release that update "just 24hrs later" raises alarm bells for me.

Surely putting out a buggy release is of concern in itself, and calls into question the testing even then ... but then to respond to a buggy release by even more rapidly releasing another new version to supposedly 'fix' the issue... :roll: It really is beginning to feel like they're winging it.

Was this just a one off mistake?

Or a symptom that they could now feeling the pressure of a stack that isn't up the job? Are attempts to squeeze more out of the stack now resulting in existing functionality breaking because the hardware isn't up to handling it all? (The skyscraper foundation analogy that I've used before)

Or worse, is this perhaps an indication that they've lost architectural / engineering integrity in their stack development - are changes in some places breaking existing functionality in others because architectural integrity has been lost? If they're rushing to change the architecture - e.g. to introduce temporal tracking of things, etc - irrespective of whether the hardware foundation could in theory cope with it, are they now taking too many short cuts and rushing, and losing control as a result?

OK, so far it's only once this has happened, so perhaps I'm prematurely jumping to possibilities.

Though I seem to recall on some of the previous Tesla videos, the commentators have said they've felt some aspects have gone backwards at times when new FSD beta releases have come out.

BTW, if anyone's interested, it looks like the Waymo team have re-started putting out technical talks / videos about their self driving efforts again... (they used to do this early on and was very interesting, but then seemed to go all quiet .. but happily it looks like they're doing them again)...

https://youtu.be/oJ96bgmSaW0?list=PLcvM ... B6dJU&t=77

Very interesting video... comparing what I've seen in that video, to what I've seen from Tesla presentations, I still have the distinct impression that Waymo are aiming for a skyscraper orders of magnitude higher than Tesla.

The things that Waymo are considering still seem to me to go way beyond what Tesla seem to be considering.

In my view, Tesla really are taking a very big gamble that they've pitched their stack appropriately.

If Tesla are right, then the additional costs for Waymo will make it no contest - clearly Tesla would definitely get the cost advantage.

But only if they can make it work, but I really do have doubts (as I've bored people with enough already :D ) that Tesla's current stack is up to the job.

Just look at the level of detail that the Waymo lidar is picking up in objects! https://youtu.be/oJ96bgmSaW0?list=PLcvM ... 6dJU&t=490

Compare that to Tesla, with Tesla's fairly low resolution video being used to estimate depth!

That extra level of detail gives Waymo a huge lead in being able to predict behaviours.

I mean, Tesla's visualisations from what I've seen in the videos on youtube seem to just barely even be able to detect people half the time, but even when it does detect people, they seem to just displayed as almost fixed 'sprites' in terms of arm and leg positions, with the sprites jumping 90 degrees seemingly at random at times, and not always matching the direction the person is facing.

Whereas Waymo's detection of people includes detailed, stable data from the lidar as to the position and motion of the limbs, etc. If you want to predict if a person walking close to the edge of the road is likely to step out into the road, having that extra detail can make the difference between reliable prediction and guess work.

The waymo development goes into quite some depth on behaviour prediction of the things around the vehicle... https://youtu.be/oJ96bgmSaW0?list=PLcvM ... dJU&t=1096

I don't recall seeing anything similar from the Tesla technical videos, though if anyone has any references to any that do show it, I'd definitely be interested to see them.

odysseus2000
Lemon Half
Posts: 6366
Joined: November 8th, 2016, 11:33 pm
Has thanked: 1535 times
Been thanked: 959 times

Re: Musk endeavours

#453439

Postby odysseus2000 » October 27th, 2021, 12:29 am

The more I think about robotic driving, the more I return to the question: Is this just about computation, or do humans have something that is beyond computation?

If driving is just about computation and humans from a wide range of abilities are able to do it, then the Tesla system as now configured ought to be able to drive better than humans as it will be distraction free.

If there is something more than computation, then the technologies that allow super human mapping of the road way, like Lidar may be very important in compensating for the difference between a pure computation system like Tesla and a human.

The Tesla approach of going with the simplest system possible that has similar optical performance to humans ought to work if the robot can be trained to recognise things as well as a beginning driver can.

In China there are several companies claiming there systems are better than Tesla. Most use lidar but are otherwise low cost.

Dunno, my general impression from having watched a lot of Tesla self drive videos is that their system is getting better, but it can still be fooled by things that it does not recognise. Whether this will always be the case I am not so sure and for now I am minded to think that the Tesla system will before too long drive better than any human.

I could be completely wrong, but systems like this can improve exponentially and that makes this such a fascinating field.

For now the equity price seems untroubled by the failure to find a working fsd, but would be almost unbounded to the upside if a working fsd emerges. Whether this floor with potential large upside will hold for long is difficult to be sure. For earnings I thought they had not done enough, but I was wrong and then the equity price rocketed on the news of the 100k purchase of Tesla cars by Hertz.

I am also encouraged by the bears who are coming out of their shelter to dis Tesla, setting up a potential short squeeze.

Regards,

odysseus2000
Lemon Half
Posts: 6366
Joined: November 8th, 2016, 11:33 pm
Has thanked: 1535 times
Been thanked: 959 times

Re: Musk endeavours

#453571

Postby odysseus2000 » October 27th, 2021, 1:20 pm

Adamski wrote:Tesla over $1 trillion market cap. Up 31% in past month.

I'd wager unless Tesla gets a cheap £20,000 car on the market, they will go the way of Cisco in 1999-2000 ie bubble followed by crash.

As the other car makers go cheap workable EVs. There's only so many millionaires around to buy 40-55k cars.


If you follow the logic of this then one should short all the luxury brands. Look at the prices of Volvo and Mercedes:

https://www.volvocars.com/uk/v/cars/sedan

https://www.mercedes-benz.co.uk/passeng ... w=BODYTYPE

If everyone in the world is to drive £25k cars, then all luxury brands are doomed, but I doubt very much that everyone will want a £25k car. After all there are many £25k ICE cars and yet luxury brands have thrived.

The analogy with Cisco escapes me.

Regards,

onthemove
Lemon Slice
Posts: 540
Joined: June 24th, 2017, 4:03 pm
Has thanked: 722 times
Been thanked: 471 times

Re: Musk endeavours

#454199

Postby onthemove » October 29th, 2021, 5:09 pm

Youtube recommended another Tesla video to me...

imv this is showing the reality of living without lidar...

https://youtu.be/4ib5uSUobSA?t=572

If you look closely it seems to be seeing the traffic light reflected in the shiny brickwork.

Similarly if you rewind to @6:50 ..

https://youtu.be/4ib5uSUobSA?t=405

It stops 3 times, and each time the red light is visible reflecting in the brickwork.

As some people have pointed out in the comments below the video, the visualisation isn't showing a red light, even though @9:34 the car beeps with a red light warning (according to the driver).

I suspect this is because the cameras are seeing the red light in the reflection, but due to it being an imperfect reflection, it's really confusing the vision system, and causing it to struggle to estimate the position of the traffic light that it's seeing - remember Tesla is inferring depth from the cameras ... meaning I suspect it doesn't know where to show it on the visualisation, (or it's trying to show it but it's estimated position is not within the field of the visualisation).

When he says "no clue why it's stopping here" you can see the red light reflected to the left of the lamppost. And if you watch carefully just prior, you can see that it was even orange. So it does look like it is the reflection of a traffic light in the shiny wall. (As opposed to the big red sign higher up the building that some commenters suggest)

If you compare with Waymo...

https://youtu.be/oJ96bgmSaW0?list=PLcvM ... 6dJU&t=772

You can see that Waymo's '5th generation' lidar has the resolution to be able to easily 'see' the traffic light hardware.

Admittedly, shiny surfaces could pose a problem for lidar, so I'm not suggesting the lidar will positively recognise the shiny wall that I believe is causing the problem in the video, without some difficulty either.

But when trying to positively identify a traffic light, Waymo should be able to use the lidar to confirm the lidar signature matches what a traffic light lidar signature should look like. With the distortions in the reflections in the wall, I very much expect that even if the lidar is getting a lot of noise from the lidar itself reflecting off the shiny surface, I very much doubt it would resemble anything like a traffic light in the circumstances of the Tesla video above.

As for the rest of the Tesla video, it does struggle with bus lanes and such like, though admittedly, some of those streets I think as a human driver, I'd struggle with those as well!

But these are things it's going to need to be able to reliably cope with before FSD could be released, and it looks like it might have a long way to go yet. The challenge with some of these things, is going to be that Tesla will need to get a whole lot better at reading the words in the road, and on the signs.

I've seen a number of Tesla videos now on the latest FSD, which show that it can't even recognise "Road Closed" let alone more complex things like bus lane hours of operation, etc.

Waymo, with all the Google background behind it, probably has quite a head start in terms of language recognition and interpretation (google translate and many other projects, e.g. https://cloud.google.com/vision/docs/ocr and https://cloud.google.com/natural-language#section-6 etc ... they even show it reading a street sign in one of those pages).

onthemove
Lemon Slice
Posts: 540
Joined: June 24th, 2017, 4:03 pm
Has thanked: 722 times
Been thanked: 471 times

Re: Musk endeavours

#454209

Postby onthemove » October 29th, 2021, 5:37 pm

In the interests of balance, this is pretty impressive... Tesla driving without any intervention in some pretty horrible weather conditions...

https://www.youtube.com/watch?v=CAnyWcS ... el=AIDRIVR

Though I do wonder how much is down to 'luck' and how much the Tesla is actually adapting to the conditions.

I noticed on another video I just watched before this one, that the Tesla accelerated to 25mph when the road sign clearly indicated a 20mph limit. (In good weather conditions). And at another point in the previous video, the driver instructed the car to speed up because there was someone behind getting held up - but the video narrator didn't consider it an 'intervention'. Hmmm.. I'd disagree... driving is just about speed and direction... if the driver has to tell the car to change speed or limit speed or whatever, in my view, that should count as an intervention.

I've seen previous videos of the Tesla's being tested by the beta testers in the snow, and it was abundantly clear in those that the car wasn't making any attempt at all to adapt to the conditions - the car just carried on regardless. That's why I'm sceptical whether the video in the link above is really as good as it appears... is it actually adapting to the road conditions, or just getting away with driving normally and got lucky?

Also couldn't help notice in this video that he seemed to be following a waymo car for some of the journey. :D

All that said though, the video in the link above does look pretty impressive nonetheless.

(Still doesn't change my view that they've still got a long way to go before they can realistically think of release as final FSD).

BobbyD
Lemon Half
Posts: 7814
Joined: January 22nd, 2017, 2:29 pm
Has thanked: 665 times
Been thanked: 1289 times

Re: Musk endeavours

#454241

Postby BobbyD » October 29th, 2021, 9:15 pm

With 2 days left in the month the sales race in Norway is hotting up:

Audi Etron GT: 46

Tesla: 55

odysseus2000
Lemon Half
Posts: 6366
Joined: November 8th, 2016, 11:33 pm
Has thanked: 1535 times
Been thanked: 959 times

Re: Musk endeavours

#454259

Postby odysseus2000 » October 29th, 2021, 11:50 pm

Interesting commentary by Munroe on the paradigm shift that is currently beginning:

https://youtu.be/qnGMoO_46N8

Regards,

odysseus2000
Lemon Half
Posts: 6366
Joined: November 8th, 2016, 11:33 pm
Has thanked: 1535 times
Been thanked: 959 times

Re: Musk endeavours

#454261

Postby odysseus2000 » October 30th, 2021, 12:16 am

onthemove

In the interests of balance, this is pretty impressive... Tesla driving without any intervention in some pretty horrible weather conditions...

https://www.youtube.com/watch?v=CAnyWcS ... el=AIDRIVR

Though I do wonder how much is down to 'luck' and how much the Tesla is actually adapting to the conditions.


I wonder if human driving is about luck too, or to be more scientific, is about the roads and cars being designed to make the probability of not having an accident high and the probability of having an accident low.

Clearly with 10 deaths and/or serious injuries per day in the UK the "high" is not high enough and the "low" is not low enough.

However, if you remove human emotion and human distraction and train the car to behave sensibly you probably make "high" higher and "low" lower without having to have all the other sensors with their higher resolution and cost.

If this thinking is correct then it comes down to questions like, can the AI read "Road Closed" and act appropriately? Can the AI tell the difference between a reflection and a traffic light? These seem tractable problems for AI.

In these terms it seems likely to me that Tesla will soon become much safer than human drivers.

One can argue that if Tesla do become good enough, they will still fail due to other drivers being distracted, driving under the influence of prohibited chemicals etc. This will almost certainly happen, but humans are likely no better at dealing with a driver in such circumstances that is an AI. One then moves to the potential for politicians to outlaw human driving and/or restrict it significantly. There are huge human implications and financial loss from every accident and at some point it may become unacceptable to have human drivers when distraction and emotion free AI is available at low cost.

Regards,

tjh290633
Lemon Half
Posts: 8209
Joined: November 4th, 2016, 11:20 am
Has thanked: 913 times
Been thanked: 4097 times

Re: Musk endeavours

#454317

Postby tjh290633 » October 30th, 2021, 12:09 pm

odysseus2000 wrote:I wonder if human driving is about luck too, or to be more scientific, is about the roads and cars being designed to make the probability of not having an accident high and the probability of having an accident low.

Clearly with 10 deaths and/or serious injuries per day in the UK the "high" is not high enough and the "low" is not low enough.

However, if you remove human emotion and human distraction and train the car to behave sensibly you probably make "high" higher and "low" lower without having to have all the other sensors with their higher resolution and cost.

If this thinking is correct then it comes down to questions like, can the AI read "Road Closed" and act appropriately? Can the AI tell the difference between a reflection and a traffic light? These seem tractable problems for AI.

In these terms it seems likely to me that Tesla will soon become much safer than human drivers.

One can argue that if Tesla do become good enough, they will still fail due to other drivers being distracted, driving under the influence of prohibited chemicals etc. This will almost certainly happen, but humans are likely no better at dealing with a driver in such circumstances that is an AI. One then moves to the potential for politicians to outlaw human driving and/or restrict it significantly. There are huge human implications and financial loss from every accident and at some point it may become unacceptable to have human drivers when distraction and emotion free AI is available at low cost.

Regards,

From all that I have seen, Tesla are possibly safe on dual carriageway highways, but they will be hopeless in narrow country lanes. In our narrow lane, human drivers are able to take it in turns to get past parked cars, in the face of oncoming traffic. Two cars in FSD mode would never be able to sort the problem out. Some aspects of human judgements and politeness are probably beyond the capability of AI.

There was a fatal accident near us last week. A young occupant lost his life when his car hit a tree in a single vehicle accident. Obviously lost control at high speed in a 30-limit area. One assumes that this could not happen with a self-driving system. If he had encountered a car overtaking another as they came round the bend, with no escape route, nothing would have saved him. I don't think even lidar can look round bends.

TJH


Return to “Macro and Global Topics”

Who is online

Users browsing this forum: No registered users and 6 guests