Major issues with Tesla Model 3

[Cosmik]

The Tesla can't drift a little to the left (US Roadway) and take a peek at what's in front of the car ahead?

It probably could be developed that way. But I think it leads us to a really important question: should the car prioritize the protection of its passengers, or other motorists, or pedestrians? Or minimise total number of injuries/deaths? Or prioritize the young over the old? Etc.

At some point, someone is going to have to write the code that does this. I could imagine that they would most likely offload the decision to the user (dropdown menus and an 'I agree' before being allowed to drive) or get an official policy from the government that might vary from country to country. Would the driver or the government be exempt from being sued for the consequences of that decision? It's a heck of a decision to take, and not something that anyone normally has to do explicitly. Could be a bit of a downer having to select who to kill before setting off on each jolly outing.

 

[Blumlein 88]

They're not untouchable...

Wanna buy a wreck?

https://abetter.bid/en/car-finder/t...MIu-bfl6HS2wIVB9tkCh0qywOkEAAYASAAEgLwKvD_BwE

Ray,

You know those were probably run into by human drivers. And in cases where that is not the case surely it was the result of the human owner of the Tesla driving it themselves.

 

[Blumlein 88]

It probably could be developed that way. But I think it leads us to a really important question: should the car prioritize the protection of its passengers, or other motorists, or pedestrians? Or minimise total number of injuries/deaths? Or prioritize the young over the old? Etc.

At some point, someone is going to have to write the code that does this. I could imagine that they would most likely offload the decision to the user (dropdown menus and an 'I agree' before being allowed to drive) or get an official policy from the government that might vary from country to country. Would the driver or the government be exempt from being sued for the consequences of that decision? It's a heck of a decision to take, and not something that anyone normally has to do explicitly. Could be a bit of a downer having to select who to kill before setting off on each jolly outing.

I've seen this question before and think little of it.

Every human driver prioritizes himself. If they have time to decide at all, it is usually to avoid a crash. If that isn't possible, rarely will there be a chance to decide what to hit and how and whom it should be. If all the auto-piloting does is the best possible job of avoiding crashes it will be a boon to safety. What you describe might be some kind of possibility way down the road for software. It effectively is not an option at all for human drivers.

 

[Cosmik]

Every human driver prioritizes himself. If they have time to decide at all, it is usually to avoid a crash....It effectively is not an option at all for human drivers.

That's why I said

It's a heck of a decision to take, and not something that anyone normally has to do explicitly.

This forces us (or someone at any rate) to actually formalize what should be done.

Normally it is something that happens spontaneously on the spur of the moment; as you say, a normal human reaction may simply to be save oneself. A person who does that and ends up killing someone else at least has the mental crutch that it was all a blur and no one could have known what would happen. They probably won't be sued in the case of a true accident.

This is completely different from sitting down and in the cold light of day making a decision that our product should protect its occupants even if that means that blameless pedestrians get killed, or that it should always protect other drivers even if it means our own children are killed, or whatever. Ditto for the owner of the car being left with that decision. Who is going to want to press 'I agree' to any of that?

 

[Cosmik]

A rather good article on the subject:
https://motherboard.vice.com/en_us/article/j5a8d3/self-driving-car-policy-uber

 

[Blumlein 88]

I think the future is people not owning their own cars, but engaging auto driving services.

Always Safe transport always puts the safety of its passengers first above everything else.

ALL_WAYS safe Transport is a socially responsible company that makes dynamic decisions to make the best most socially good choice should an accident be unavoidable.

Which one do you think will do the most business?

I also think about good samaritan laws which protect those who decide to help someone in danger or need of assistance. The flip side however is that people are not required to give assistance. I would think it likely that just to cut what will always be a Gordian knot, laws will lean toward taking greatest care of the safety of who is in self driving cars. While that can sound like an every man for himself scenario, the other of factoring in all of this other data simply becomes impossible to define in ways that won't spawn second guessing limited only by the creativity of attorneys. I also think that other factors once such data could be run thru in real time will make self driving reliable enough you'll nearly never need to make the moral ethical decision about who gets harmed. That level of technology will have networked cars and highways reporting and planning on conditions well enough you'll nearly never have accidents occur. Making the whole crazy idea moot. Until then self interest will both push toward safer driving and make clear what the responsibilities are for people and the technology involved.

 

[Cosmik]

Tesla on fire: UK director Michael Morris' car bursts into flames
https://www.bbc.co.uk/news/world-us-canada-44511200

 

[palamudin]

TinFoil theory: Someone clipped the brake lines. Tesla software detected failure of brakes and locked the clamp. Person drived with locked brakes till the friction caused temperature spike that lit the cable isolation. Its actually the first time tesla prevented assassination.

 

[Cosmik]

TinFoil theory: Someone clipped the brake lines. Tesla software detected failure of brakes and locked the clamp. Person drived with locked brakes till the friction caused temperature spike that lit the cable isolation. Its actually the first time tesla prevented assassination.

Could be! I'm thinking more along the lines of that being what a battery fire looks like. Vast amounts of energy stored in highly reactive chemicals held between wafer-thin bits of metal and the like.

 

[palamudin]

Front wheel arch is not the place where batteries are stored. If the pack got lit he would go ablaze in a matter of seconds.

 

[FrantzM]

Could be! I'm thinking more along the lines of that being what a battery fire looks like. Vast amounts of energy stored in highly reactive chemicals held between wafer-thin bits of metal and the like.

As opposed to driving with a cauldron with at least 12 Gallons of highly flammable and energy dense liquid ! Plus an electrically alive system ?

EV technology is at its infancy and EV shall only improve. Whether by Tesla or others. All that I can say is that Tesla is a disruptor. Other companies will either follow suit (make electric cars) or disappear. The days of ICE engines are numbered. These simply don't make any sense.. what with their paltry always <30% efficiency.

Still in-line and eager to get mine. ... .... ? We'll see

 

[Cosmik]

As opposed to driving with a cauldron with at least 12 Gallons of highly flammable and energy dense liquid ! Plus an electrically alive system ?

I was careful not to say it was any worse than a petrol vehicle (diesel not so dangerous?). Nevertheless, anything like these batteries scares me a bit because it is not as mature a technology as petrol.

There is still a sense (in my mind) that the people promoting electric vehicles are very gung-ho about it and, even worse, that governments are giving them an easy time with approvals, etc. because of the virtue signalling opportunities they provide.

 

[Cosmik]

I've read about these accidents involving auto-pilot with a strange sense of detachment: it's mainly in the USA; there aren't many Teslas; no one would be stupid enough to rely on the system and watch videos, etc. But now that I come to think about it, it's entirely possible that I could be driving on a road at 60 mph with nothing between me and oncoming vehicles but some painted lines, and a 2-tonne car could just decide to swerve towards me because it has mistaken a reflection off a wet road for a cyclist and decided to spare their life instead of mine. There really is just their stupid, ridiculous, idiotic, naive pattern recognition software between us and death. Have we gone totally crazy!!?

 

[svart-hvitt]

I've read about these accidents involving auto-pilot with a strange sense of detachment: it's mainly in the USA; there aren't many Teslas; no one would be stupid enough to rely on the system and watch videos, etc. But now that I come to think about it, it's entirely possible that I could be driving on a road at 60 mph with nothing between me and oncoming vehicles but some painted lines, and a 2-tonne car could just decide to swerve towards me because it has mistaken a reflection off a wet road for a cyclist and decided to spare their life instead of mine. There really is just their stupid, ridiculous, idiotic, naive pattern recognition software between us and death. Have we gone totally crazy!!?

Autopilot has no place in ordinary traffic! Before letting autocars enter ordinary traffic, let them in all their varieties first take over restricted areas like harbours. Let’s see what happens then.

I guess we hear little about this because people see that the machines are not mature and would clutter harbours, make a mess where today’s clockwork is pretty good.

Letting autocars enter ordinary traffic in what is really a test, trial with ordinary humans’ lives at stake tells us a great deal about one’s view on the value of human life as compared to pecuniary issues.

 

[FrantzM]

The issue is not a simple one. No doubt lives are at stake. I am ambivalent about AI and autopilot is one of its application. We may find ourselves having to rely on a larger machine/brain/processing power to reduce the risks of Human lives loss. I shudder to think what happens when we are disconnected from the Cloud and have to rely on a smaller AI/brain inside the car ... Which takes us to square one...
On the other Hand ... Without autopilot, we have relied on failing humans... Our car's us the drivers... and those on the road.. Appealing to human reflexes, awareness, decency, aptitude , proper reflexes and adequate decision making ...IOW on damn sheer luck... a person being distracted for a second by the sound ofa phone is enough to remove another from this plane of existence... or an enraged person swerving .. just because ... or the Texter.... or the good citizen but poor driver with slow reflexes ...
Are AIs better drivers than humans? Yes! Now we have to weight the decisions processes, trees, criteria, weight and ultimately results... Who will ponder these us, Human, or the AIs? Are that much better in managing the contradictions than an AI?
I, for one, don't have any answers. AI driving a car seems to me more reliable than a Human being... I do not if it is safer at this point in time. It may become so in a hurry....

 

[Thomas savage]

Love to see a AI car navigate its way about Dartmoor with sheep sleeping on the road and cows marching up back lanes, then there’s all the irrational human behaviour on the road for the AI to deal with not to mention the vast array of diffrent road styles.

Can’t see it working personally, the human brain is probably the most incredible thing in the universe I’m not sure that’s really appreciated .

 

[svart-hvitt]

The issue is not a simple one. No doubt lives are at stake. I am ambivalent about AI and autopilot is one of its application. We may find ourselves having to rely on a larger machine/brain/processing power to reduce the risks of Human lives loss. I shudder to think what happens when we are disconnected from the Cloud and have to rely on a smaller AI/brain inside the car ... Which takes us to square one...
On the other Hand ... Without autopilot, we have relied on failing humans... Our car's us the drivers... and those on the road.. Appealing to human reflexes, awareness, decency, aptitude , proper reflexes and adequate decision making ...IOW on damn sheer luck... a person being distracted for a second by the sound ofa phone is enough to remove another from this plane of existence... or an enraged person swerving .. just because ... or the Texter.... or the good citizen but poor driver with slow reflexes ...
Are AIs better drivers than humans? Yes! Now we have to weight the decisions processes, trees, criteria, weight and ultimately results... Who will ponder these us, Human, or the AIs? Are that much better in managing the contradictions than an AI?
I, for one, don't have any answers. AI driving a car seems to me more reliable than a Human being... I do not if it is safer at this point in time. It may become so in a hurry....

What you write relates to a distant future. Producers of EV and AI take advantage of people’s fascination with science (fiction) to secure on-going financing.

As I wrote above: Why do you think EV and AI producers don’t succeed in replacing humans in harbours and other restricted areas where the safety margin is higher and economic value add is easily measurable.

I am growing tired of the AI story. To illustrate my point: Take a look at computer chess ratings. The growth in the number of transistors followed Moore’s power law; yet the rating growth was linear.

Very few AI (and EV) enthusiasts take the time to think through this empirical fact.

My point is: Progress is slow. It’s in the financial interest of AI and EV producers to pretend otherwise. Ignore empirical facts at one own’s peril, but be careful to use the political system to favour dreams and possibly dangerous innovations for the benefit of a handful at the cost of the many.

 

[Frank Dernie]

I am looking forward to AI cars. I am going to bully them. Pull up at a give way sign, see traffic coming, OK it has right of way but it is AI so I am pulling out. All I have to judge is if it is far enough away to slow before it hits me.
Just joking...

 

[Frank Dernie]

I am a fan of plug in hybrid. Much of the electric benefit but less range anxiety. Batteries have a good life if there is good control over charging, our 13 year old Prius which we are selling still has in spec traction battery and our 5 year old plug in seems to have the same range as when new. The new plug-in is better again in most ways (just less internal oddments storage).
The problem is "supercharging" the battery takes it to the performance limit so batteries won't last as long as those charged over night at home.
I see EVs as a bit of an unsustainable blip, with charging being the deal breaker in the end, both time and capacity. Also the electrical bit has high efficiency but the generation and charging does not so any efficiency benefit is illusory.

 

[Cosmik]

Tesla chief executive Elon Musk has accused an employee of carrying out "extensive and damaging sabotage" at the electric carmaker.

In an email to staff, Mr Musk said an unnamed employee made unspecified coding changes to its manufacturing operating system and sent sensitive data to unnamed third parties.

https://www.bbc.co.uk/news/business-44531777