Exactly. Any driver can put you at risk. Nothing to do with Tesla.
As far as we can tell from reading the article, this whistleblower is reporting on stuff that is entirely resulting from driver misbehavior, which is possible in any car. But since he worked at Tesla and has some kind of issue with Tesla, it seems like he’s trying to make it about FSD when it’s really just down to the driver. At least that’s my read of it.
>Exactly. Any driver can put you at risk. Nothing to do with Tesla.
Well, to the extent it is caused by a Tesla, it absolutely has something to do with Tesla. I don't get to smash into a car because a drunk driver would. Why should a Tesla? It's a bit disingenuous to "put it on the driver" when the driver isn't driving because he's using Tesla's self-driving software.
That's what Tesla says, but that's clearly just a way for them to avoid liability. A system that points you off a cliff and then gives you control after its already terminal is facially defective.
>When the non-beta exists, then we can talk about that, but we aren't there yet.
Sorry, but putting the label "test" on things is not actually a way to avoid legal liablity for your products. No less when you sell them without restriction.
>How do you propose any self driving system reach perfection without ever having a beta testing period?
A few thoughts here: (1) I didn't propose otherwise, I just pointed out that it's deficient (at least, legally) to say "I can be negligent because others are negligent", or to put it another way "other people's negligence should put you on warning with regards to my negligence", (2) I don't think it's my burden to satisfy this problem, regardless, even granting your point, it doesn't have to proceed the way Tesla has so far, it certainly hasn't with regards to other similar companies making similar products.
>And why not give drivers responsibility for maintaining safety?
Because the manufacturer of a self-driving product should be primarily responsible for how it functions. How exactly is that controversial to you? To whatever extent the driver is jointly liable might be a question, but, given how Tesla's self driving feature works, Tesla is clearly responsible for how it functions, including when it determines that the driver must take over. I've seen nothing that indicates that this function occurs in a way that always allows a driver to establish a safe response to whatever the self-driving couldn't handle. So, to be honest, your question seems a bit absurd to me.
>Shouldn’t they try to be safe, and carry responsibility when they are fully in control?
Yeah, but no one here is talking about when the driver is driving, we are talking abotu when the self-driving feature is driving. You are farcically insisting that the driver is driving even when self-driving is enabled. Can't have it both ways.
>Because the manufacturer of a self-driving product
Hold on right there. There is no self-driving product in the picture or in the conversation. Other than one in your imagination maybe. It is a beta which is not yet self driving.
>should be primarily responsible for how it functions.
By your logic, should Toyota pay for all accidents that happen in Toyotas? No, because Toyotas are not fully self driving? I have news for you. Neither are Teslas.
>How exactly is that controversial to you?
Your hypothetical sounds reasonable in theory, but it’s predicated on an imaginary situation that does not exist.
>Hold on right there. There is no self-driving product in the picture or in the conversation. Other than one in your imagination maybe. It is a beta which is not yet self driving.
They sell it, it's a product. You can call it a beta product. This is entirely pointless semantics.
>By your logic, should Toyota pay for all accidents that happen in Toyotas? No, because Toyotas are not fully self driving? I have news for you. Neither are Teslas.
You are going in a loop. That's not logic, and certainly not my logic. Toyota doesn't sell self-driving cars, so why would Toyota be responsible for how the cars drive. If there is a defect in their cars, they would be responsible for that. You are once again playing semantics. Tesla sells cars with a self driving product. You are trying to play a semantic game by saying it's not "full self driving" because some aspect of its function, but what words we call it is besides the point, what matters is how its functions. Tesla is liable for how the product functions because it's in control of how the product functions, not because its "full" or "partial" self driving.
>Your hypothetical sounds reasonable in theory, but it’s predicated on an imaginary situation that does not exist.
You literally just keep making up nonsense in order to justify your positions. Like saying "it's a beta" so it's not a product. You can buy it. It's a product. They are already facing lawsuits over it. Get real. Hands down one of the most pointless conversations I've had here when you are engaging like this.
It may be a little complex to talk about but the product you speak of is essentially a technology preview (the beta) wrapped together with a future functionality the buyer is prepaying for.
By paying, they get access to the preview now, and they will get the actual released software later when it is ready.
It’s all bundled up as, as you say correctly, a product.
What you’re missing is a bit of subtlety about the contents of the product, that’s all.
During the preview period users are responsible (blah blah I don’t need to repeat that part yet again, right?).
I think it’s not actually super hard to understand but it is admittedly not exactly like most other products. It’s a product the features of which are being rolled out progressively until it is fully ready. Meantime it is made crystal clear in many ways and repeatedly to buyers what their responsibilities are… before buying, after buying, before turning it on in settings, during driving, etc. It is very well communicated.
The problem is that humans are bad at sitting there paying attention while they're not actively engaged, which is what's required to use FSD safely. Also, FSD has essentially no benefits if you use it safely. If you listen to someone talk about what they want FSD for, they'll probably tell you they want to be paying attention to something other than driving or that they want to sit in a more comfortable position where their feet aren't on the pedals and/or their hands aren't on the steering wheel.
The current Tesla FSD beta has no purpose other than as a tech demo, but people absolutely treat it as if it's ready to be your chauffeur while you take a nap in the back seat.
Speak for yourself but I have no problem at all staying engaged when my life is on the line. And FSD beta with me supervising is I must admit a way safer driver than myself alone.
The biggest delta is probably follow distance in edge cases.
I generally try to keep good follow distance pretty religiously, but FSD does it better. And I’m not talking about the amount of distance. If anything it is closer on average then I would be. I’m talking about the edge case moments when overtaking or getting over to an exit; it is consistently, assiduously, safe at points where I might take a risk and it will not. It also sees way more than I can.
Also, it does not have an ego.
And it lets me detach my ego involvement with how another driver is behaving. Related, it handles stop and go absolutely amazingly and completely removes the stress of the experience.
> The current Tesla FSD beta has no purpose other than as a tech demo
This is a bizarre take. FSD beta is many things beyond a demo. To name a few, it’s gathering data, it’s testing the system and improving, and it’s training other drivers, pedestrians, and governments how to live in a world that has FSD cars on the road.
I have worked in SW for the automotive industry for 10 years.
The amount of times we where stopped from implementing features or had to roll them back and the number of delays because a test showed a decrease in safety was frustrating but I do feel safe driving my daughter around in the cars I worked on because off it.
With the cybertruck, just the transition to 48V electrical system means ALL electronic components had to re-engineered, and for an established manufacturer this would mean huge costs and time to retest everything.
18 comments
[ 0.21 ms ] story [ 52.3 ms ] threadLike any car, drivers have to keep their hands on the wheel and their attention on the road situation, and stay in control at all times.
Until FSD is not in beta, there’s no use pointing out that a bad driver can do bad things because that is true in any car.
Saying the system is not ready may be true, but it’s not anything we don’t already know.
If the point is that some drivers need better education, then sure, this is definitely a special case of that and worthy of close attention.
As far as we can tell from reading the article, this whistleblower is reporting on stuff that is entirely resulting from driver misbehavior, which is possible in any car. But since he worked at Tesla and has some kind of issue with Tesla, it seems like he’s trying to make it about FSD when it’s really just down to the driver. At least that’s my read of it.
Well, to the extent it is caused by a Tesla, it absolutely has something to do with Tesla. I don't get to smash into a car because a drunk driver would. Why should a Tesla? It's a bit disingenuous to "put it on the driver" when the driver isn't driving because he's using Tesla's self-driving software.
When the non-beta exists, then we can talk about that, but we aren't there yet.
>When the non-beta exists, then we can talk about that, but we aren't there yet.
Sorry, but putting the label "test" on things is not actually a way to avoid legal liablity for your products. No less when you sell them without restriction.
And why not give drivers responsibility for maintaining safety?
Shouldn’t they try to be safe, and carry responsibility when they are fully in control?
A few thoughts here: (1) I didn't propose otherwise, I just pointed out that it's deficient (at least, legally) to say "I can be negligent because others are negligent", or to put it another way "other people's negligence should put you on warning with regards to my negligence", (2) I don't think it's my burden to satisfy this problem, regardless, even granting your point, it doesn't have to proceed the way Tesla has so far, it certainly hasn't with regards to other similar companies making similar products.
>And why not give drivers responsibility for maintaining safety?
Because the manufacturer of a self-driving product should be primarily responsible for how it functions. How exactly is that controversial to you? To whatever extent the driver is jointly liable might be a question, but, given how Tesla's self driving feature works, Tesla is clearly responsible for how it functions, including when it determines that the driver must take over. I've seen nothing that indicates that this function occurs in a way that always allows a driver to establish a safe response to whatever the self-driving couldn't handle. So, to be honest, your question seems a bit absurd to me.
>Shouldn’t they try to be safe, and carry responsibility when they are fully in control?
Yeah, but no one here is talking about when the driver is driving, we are talking abotu when the self-driving feature is driving. You are farcically insisting that the driver is driving even when self-driving is enabled. Can't have it both ways.
Hold on right there. There is no self-driving product in the picture or in the conversation. Other than one in your imagination maybe. It is a beta which is not yet self driving.
>should be primarily responsible for how it functions.
By your logic, should Toyota pay for all accidents that happen in Toyotas? No, because Toyotas are not fully self driving? I have news for you. Neither are Teslas.
>How exactly is that controversial to you?
Your hypothetical sounds reasonable in theory, but it’s predicated on an imaginary situation that does not exist.
They sell it, it's a product. You can call it a beta product. This is entirely pointless semantics.
>By your logic, should Toyota pay for all accidents that happen in Toyotas? No, because Toyotas are not fully self driving? I have news for you. Neither are Teslas.
You are going in a loop. That's not logic, and certainly not my logic. Toyota doesn't sell self-driving cars, so why would Toyota be responsible for how the cars drive. If there is a defect in their cars, they would be responsible for that. You are once again playing semantics. Tesla sells cars with a self driving product. You are trying to play a semantic game by saying it's not "full self driving" because some aspect of its function, but what words we call it is besides the point, what matters is how its functions. Tesla is liable for how the product functions because it's in control of how the product functions, not because its "full" or "partial" self driving.
>Your hypothetical sounds reasonable in theory, but it’s predicated on an imaginary situation that does not exist.
You literally just keep making up nonsense in order to justify your positions. Like saying "it's a beta" so it's not a product. You can buy it. It's a product. They are already facing lawsuits over it. Get real. Hands down one of the most pointless conversations I've had here when you are engaging like this.
By paying, they get access to the preview now, and they will get the actual released software later when it is ready.
It’s all bundled up as, as you say correctly, a product.
What you’re missing is a bit of subtlety about the contents of the product, that’s all.
During the preview period users are responsible (blah blah I don’t need to repeat that part yet again, right?).
I think it’s not actually super hard to understand but it is admittedly not exactly like most other products. It’s a product the features of which are being rolled out progressively until it is fully ready. Meantime it is made crystal clear in many ways and repeatedly to buyers what their responsibilities are… before buying, after buying, before turning it on in settings, during driving, etc. It is very well communicated.
The current Tesla FSD beta has no purpose other than as a tech demo, but people absolutely treat it as if it's ready to be your chauffeur while you take a nap in the back seat.
The biggest delta is probably follow distance in edge cases.
I generally try to keep good follow distance pretty religiously, but FSD does it better. And I’m not talking about the amount of distance. If anything it is closer on average then I would be. I’m talking about the edge case moments when overtaking or getting over to an exit; it is consistently, assiduously, safe at points where I might take a risk and it will not. It also sees way more than I can.
Also, it does not have an ego.
And it lets me detach my ego involvement with how another driver is behaving. Related, it handles stop and go absolutely amazingly and completely removes the stress of the experience.
This is a bizarre take. FSD beta is many things beyond a demo. To name a few, it’s gathering data, it’s testing the system and improving, and it’s training other drivers, pedestrians, and governments how to live in a world that has FSD cars on the road.