On the ease of programming - you're probably right, a 'shouldn't be that much more complicated' was intended to be more of a cap - producing a part from a CNC machine shouldn't exceed 110% of the effort of doing it with a printer, and quite possilby be less.
Also, I've seen small CNC machines, and while you can indeed get a printer more the size one of those popcorn machines while the CNC I saw was about double it, both are still best left in a garage/machine shop. I saw an article recently how the melting plastic releases toxic fumes. Lovely.
It's also that the barrier to entry is lower in that there's a nice library of 3D things to print online and the slicing process for the printing is simpler the software to do the printing is more readily available.
Same deal with C++ vs VB. VB is easy to access by anybody with a windows box. Programming in C++ takes a greater initial investment, but after that you can do so much more...
As for producing weapons with each, keeping in mind that I put a liberator(effectively single shot) up against a M-3(full auto, service life in years), I'm sure that you could build nearly everything you need for a firearm on the CNC. So it might be the only non-hand tool you need, assuming the proper design. The M-3 was designed to be cheap to make in large quantities for the time it was made, when CNC wasn't available, but big sheet metal presses were. So it's very much NOT optimized to be made individually in a small shop using not much other than a CNC. You'd probably want to redesign the parts that are intended to be welded together, for example. It's an interesting though process....
First, I'm not going to intrude into grandpa's medical situation, and I'm certainly not going to recommend a currently illegal drug on the basis of an internet post.
Second, well, typical Yoga and Tai Chi are far beyond my grandfather's ability. Merely standing is something of a challange for him while wearing supports.
Maybe if this stuff had been available 40+ years ago when he was injured...
Limited liability has happened before. Plenty of people killed by company drivers, generally the company makes some payout, but not enough to kill the business.
To sum it up: I figure that the companies will fight for, and win, some sort of limited liability. Beyond that if the systems are sufficiently safer than human drivers it won't even cost that much. Consider how much liability insurance costs people now.
If the cars are 80% less likely to be in an accident, you could probably increase liability levels 10X over standard (currently ~$100k in most states) and still come out ahead.
Personally, I'd compare it more between C++ and visual basic scripting. While it does indeed take more knowledge to operate the lathes and such, currently that tool set can produce far more capable devices, and I'd imagine that at least the CNC cutter shouldn't be that much more complicated to program than the printer.
With the printer you can make a 'liberator' type firearm - a single shot weapon that you MIGHT be able to reuse the trigger group for. With the machine shop equipment you can churn out full auto M-3 'grease guns' for not much more money than the plastic in the liberator.
Why? In order to share traffic data, routing intentions, etc.... How? There's a lot of options - infrared, radio, directional, non-directional, etc... Hacking? Still possible, just means that you also need to write something to 'speak' in whatever protocol the cars are using. Though I wouldn't assign a huge amount of trust in the signals. Each car should still assume that the other cars are likely to behave erratically and move accordingly.
There is no autonomous car that doesn't require a person to be sitting in the driver's seat ready to take over in case something goes wrong.
That's due to legal requirements to be on the road. They're still in the research phase after all. Even a self-driving car that's limited to highway driving for 'no monitoring needed' would still be highly useful.
Personally, I think that they're less than 5 years off, so it's time to be really working on the legal aspects.
Well, yes. Not enough data, low hanging fruit, and all that.
The first chess programs couldn't beat a human familiar with the rules. Today they have machines that can beat the best chess minds in the world.
I predict first we'll see a widening of 'assisted' driving - monitoring for tailgating shifting towards 'automatic follow', automatic braking if something intrudes into the road, etc...
Then we'll have the equivalent of early drones - capable of traveling, from parking area to parking area. Parking(or at least selecting a spot)/landing depends on you. Eventually those situations will be less and less(they can now land on their own), and they'll start showing up in places like rich people's cars and highway trucks. The trickle down will continue, hitting the 'worst' drivers first - drunks, etc...
I think the insurance companies will gladly support the new cars.
I do as well, but there are still some extreme complexities.
The vast majority of auto accidents today are easily blamed on a driver, making the manufacturer not liable. However, consider the Toyota acceleration problem. What if somebody experiencing that ended up rear ending another car, pushing them into traffic where they are then struck again, increasing injuries?
That's a case where the manufacturer could end up being liable again. They're still a huge company, so it's generally not worth suing, but given enough auto-driving cars with accidents arguably blamable on the AI driver I could see a class-action starting up. As a result I could actually see 'liability insurance' baked into the cost of the system.
While most people carry insurance limited to around $100k at the top end per person, there is no such limit in a lawsuit. So that could drive costs for the manufacturer through the roof.
On the other hand, if accidents are truly 80% less likely than traditional drivers, you'd have to increase maximum liability by more than 5 in order for it to actually end up costing the consumer more, as each subsequent $ of insurance is cheaper than the last, because it's less likely to be used. Just because you might carry a $1M umbrella policy doesn't mean that somebody will get $1M if you cause them to break a finger(for example).
But from a business standpoint I could see car companies pushing to have their liability limited - in this day and age, with limits where they are now, the 80% thing, and such, I think $1M per person limit from congress would make the businesses looking to put out autonomous cars seriously consider it.
That's more than enough to drive a business out of business if it's auto-drive system is truly flawed, but more than enough to screw those HIT by bad drivers less than they currently are, while saving drivers/riders enough money to save them money in the saving of lives.
I can already predict crashes due to hacking/ buggy softwares and etc.
I can too, which is why I always assume that at least *some* accidents will still occur. For that matter, I figure that the profiles of accidents will change - moving away from things like 'ran through a red light' and towards 'drove into a not completely correctly marked construction zone'.
Still, the question becomes one of how the autonomous cars affect the average rate and severity of crashes. If it cuts off 50%, concentrated in the extreme, more commonly fatal crashes, it's a real benefit to society.
However, what I dislike is how it is being presented here.
If you follow the links you'll see a mention of cutting accident rates by 80%. I tend to view 50% as a good dividing line for starting adoption, 80% just makes the case so much tighter.
Seriously: no, I cannot* imagine that the autonomous cars are actually better than the average intoxicated driver, according to modern limits. Your average intoxicated driver gets home safe and doesn't put anybody at risk. Given what I know about the difficulty in training a computer to behave intelligently in unplanned situations, I cannot imagine that autonomous cars are significantly better than your average intoxicated driver.
Excepting your disclaimer, I'd say that you have a very limited imagination. Personally, I can easily imagine it. Indeed, with the Google cars and such I'd say that we're very very close. Especially if 'perfection' isn't your standard, merely 'median human driver', IE 'better than the median drunk driver'. Now, I do imagine that the accidents would be different. I picture our AI cars being very good at avoiding accidents via 'fast twitch' actions, trading that for 'duh' accidents that no human would get into. The sort of accidents we see when human drivers get too stupid about following GPS directions, essentially. Things like driving into a flooded ford, through construction, etc...
When I figured it out, I pegged a self-driving car with merely 'half' the accident rate of humans to be worth a couple grand a year to me, and vastly more to those who either drive more or are outright bad drivers. I figured that DUI convicts would be among the first adopters - forced to do so by the courts, as interlocks are both expensive and hard to use, and I'd imagine the insurance break would be substantial at that point.
That's already been known since like the '80s. My grandfather has spinal damage from both an accident and polio. He's been using a electrostimulation machine for decades, slowly increasing/restoring function in his legs.
The trick would be something that works about 10X faster so you could 'finish' in a couple decades rather than it being essentially for life....
What about purely online services? I haven't encountered this yet, and I'm sort of surprised. I'm using a public wifi outside of my home country, and that triggers me using my private VPN times two. I have a VPS I mess around with, set up VPN on it. I've used it to access things like Netflix, which isn't available in my current country, get the 'correct' steam pricing, etc... If anybody really wanted to they could track me down from that IP address, but it'd probably require a warrant.
But my VPS isn't even in the same time zone as my home, and I've recently changed addresses along with the move.
I guess that if anybody asks I might not respond with 'no' to whether I'm on vacation, but with 'close, I'm on a business trip'.
Uh, that's what I said? Italians and such are much closer to 'white' in skin tone than they are to the really black blacks. Of course, blacks themselves range a wide gamut.
Good point; I thought a 3M level UID was too high for somebody of that history; However while cap-I wolf isn't as distinguished as lower-l wolf, he still had a fairly normal posting history until recently. I did check.
It's AlphaWoIf_HK, check the hidden comments attached to my original post - he's either infected with malware or had his account hacked, what with posting 4 duplicate 'mycleanpc' spam posts. Given that he has a normal posting history until very recently(and I remember other posts of his). Since this is an IT Pro focused site, it's 'his fault' in either case.;)
You're lucky HK screwed up, or I'd have never seen this. I know that there's a whole rainbow of colors in Africa, which is why I put 'straight from Africa black' in quote marks. The idea is to give readers a vision of the color in question, not to quibble about the pigmentation range of the continent.
Sorry to see that your account has apparently been hacked, HK. Or some malware is messing with your posting ability. Suggest you change your passwords and whatnot...
Is this for real? Hollywood produced plenty of Italian American superstars, as well as Latinos. How did Bollywood manage?
There's a vaster difference in skin tone between Italians, Latinos, and Indians compared to 'Straight from Africa Black' than they are from European. Most Europeans can come close to those shades with sufficient tanning.
That's just looking at it from a technical perspective. I still believe that the 'problems' were overstated, but then, given the tendency for makeup departments to do extensive work-ups on movie stars so they look 'normal' on screen, for clothing designed to present a certain appearance on screen as opposed to in real life, and such I can believe that there were issues at some point. After that it becomes a continuing education problem - you need a makeup department that knows how to properly prepare black people for their role on screen, and you might not have that with all crews due to lingering effects from racism, inertia, etc...
The problem with your premise is this assumption. You have to remember that Cinematography wasn't about accuracy, it was about looking good. Remember all the shortcuts taken during the original Star Wars trilogy? A major function through history was fooling your eyes. 'Good Enough' was everywhere.
Considering that there are lighting changes you want to do when you're photographing a black object vs a white one - Say a black PC Case vs a white PC Case, I believe that they might indeed have a point, if you assume that the directors, photographers, lighting directors and all are incredibly stupid.
Thinking back, I remember massive numbers of black people in films, all through history. The 'washed out appearance' that the article complains about I only really remember from the really early B&W films. I think lighting would be a minor issue up against the inertia of racism - there were plenty of 'black' films starting from back in the '70s to develop any necessary technique changes. It has to do more with the population of the USA being only ~18% black*.
I don't get it... you go through all that trouble to purchase in a manner you feel is anon on silk road, but then you post about it on slashdot using your registered account?
Given the nature of the internet, it can easily be argued that he's lying his ass off. Even if he isn't, 'small hash order' indicates he's a user, not a dealer, thus *on average* incredibly unlikely to be a worthy target for the 3+ agencies you'd need to coordinate with in order to track him down.
Off the top of my head - you'd need to get a warrant to get slashdot to disclose Cito's account and IP address information. Then you'd need to figure out WHERE in the world he is(presumably the USA). You have to hope that he was using home or at least work for his slashdot postings rather than using the same anonymous internet cafe. Once you've figured out where he is, you have to contact the appropriate state police agency to coordinate with, along with the postmaster general(assuming USPS was used as opposed to UPS/Fedex).
On Average it's just not worth it. They want dealers, not users.
I'm quite surprized that firefighters threw water at a litium fire (but most powder extinguishers are almost as bad), and punched extra holes in it (to let the water out, maybe?), but not in a good way.
Yeah, I only ever made it into 'introductory' fire fighting and even I know that tossing water on a metal fire is generally a very bad idea.
the average user can still install everything on their C: drive without making any changes from the default installation.
\
This one is important. I purchased the largest SSD I could reasonably afford(fiscally conservative), and found that messing with manually shifting programs around(lots of games) was costing me more time than what the SSD saved in quicker response times.
I'd like to buy a combo with a larger SSD, or see an automatic program capable of managing the caching, but these drives address a potentially huge market - almost as cheap as a HD solution while retaining ~90% of the 'real world' performance of an SSD.
I'd like to see a 16-64GB cache solution so the HD doesn't have to be anywhere near as aggressive in pruning stuff from the SSD, but I'm seriously interested.
I'm almost surprised that they didn't go to this first thing - water isn't good for gasoline fires either, though the sheer amount a fire engine can put out will often put out fires that water would not otherwise be recommended for.
Slashdot Mirror
On the ease of programming - you're probably right, a 'shouldn't be that much more complicated' was intended to be more of a cap - producing a part from a CNC machine shouldn't exceed 110% of the effort of doing it with a printer, and quite possilby be less.
Also, I've seen small CNC machines, and while you can indeed get a printer more the size one of those popcorn machines while the CNC I saw was about double it, both are still best left in a garage/machine shop. I saw an article recently how the melting plastic releases toxic fumes. Lovely.
It's also that the barrier to entry is lower in that there's a nice library of 3D things to print online and the slicing process for the printing is simpler the software to do the printing is more readily available.
Same deal with C++ vs VB. VB is easy to access by anybody with a windows box. Programming in C++ takes a greater initial investment, but after that you can do so much more...
As for producing weapons with each, keeping in mind that I put a liberator(effectively single shot) up against a M-3(full auto, service life in years), I'm sure that you could build nearly everything you need for a firearm on the CNC. So it might be the only non-hand tool you need, assuming the proper design. The M-3 was designed to be cheap to make in large quantities for the time it was made, when CNC wasn't available, but big sheet metal presses were. So it's very much NOT optimized to be made individually in a small shop using not much other than a CNC. You'd probably want to redesign the parts that are intended to be welded together, for example. It's an interesting though process....
First, I'm not going to intrude into grandpa's medical situation, and I'm certainly not going to recommend a currently illegal drug on the basis of an internet post.
Second, well, typical Yoga and Tai Chi are far beyond my grandfather's ability. Merely standing is something of a challange for him while wearing supports.
Maybe if this stuff had been available 40+ years ago when he was injured...
Limited liability has happened before. Plenty of people killed by company drivers, generally the company makes some payout, but not enough to kill the business.
I posted some more of my thoughts on it here
To sum it up: I figure that the companies will fight for, and win, some sort of limited liability. Beyond that if the systems are sufficiently safer than human drivers it won't even cost that much. Consider how much liability insurance costs people now.
If the cars are 80% less likely to be in an accident, you could probably increase liability levels 10X over standard (currently ~$100k in most states) and still come out ahead.
Personally, I'd compare it more between C++ and visual basic scripting. While it does indeed take more knowledge to operate the lathes and such, currently that tool set can produce far more capable devices, and I'd imagine that at least the CNC cutter shouldn't be that much more complicated to program than the printer.
With the printer you can make a 'liberator' type firearm - a single shot weapon that you MIGHT be able to reuse the trigger group for.
With the machine shop equipment you can churn out full auto M-3 'grease guns' for not much more money than the plastic in the liberator.
Why? In order to share traffic data, routing intentions, etc....
How? There's a lot of options - infrared, radio, directional, non-directional, etc...
Hacking? Still possible, just means that you also need to write something to 'speak' in whatever protocol the cars are using. Though I wouldn't assign a huge amount of trust in the signals. Each car should still assume that the other cars are likely to behave erratically and move accordingly.
There is no autonomous car that doesn't require a person to be sitting in the driver's seat ready to take over in case something goes wrong.
That's due to legal requirements to be on the road. They're still in the research phase after all. Even a self-driving car that's limited to highway driving for 'no monitoring needed' would still be highly useful.
Personally, I think that they're less than 5 years off, so it's time to be really working on the legal aspects.
Well, yes. Not enough data, low hanging fruit, and all that.
The first chess programs couldn't beat a human familiar with the rules. Today they have machines that can beat the best chess minds in the world.
I predict first we'll see a widening of 'assisted' driving - monitoring for tailgating shifting towards 'automatic follow', automatic braking if something intrudes into the road, etc...
Then we'll have the equivalent of early drones - capable of traveling, from parking area to parking area. Parking(or at least selecting a spot)/landing depends on you. Eventually those situations will be less and less(they can now land on their own), and they'll start showing up in places like rich people's cars and highway trucks. The trickle down will continue, hitting the 'worst' drivers first - drunks, etc...
I think the insurance companies will gladly support the new cars.
I do as well, but there are still some extreme complexities.
The vast majority of auto accidents today are easily blamed on a driver, making the manufacturer not liable. However, consider the Toyota acceleration problem. What if somebody experiencing that ended up rear ending another car, pushing them into traffic where they are then struck again, increasing injuries?
That's a case where the manufacturer could end up being liable again. They're still a huge company, so it's generally not worth suing, but given enough auto-driving cars with accidents arguably blamable on the AI driver I could see a class-action starting up. As a result I could actually see 'liability insurance' baked into the cost of the system.
While most people carry insurance limited to around $100k at the top end per person, there is no such limit in a lawsuit. So that could drive costs for the manufacturer through the roof.
On the other hand, if accidents are truly 80% less likely than traditional drivers, you'd have to increase maximum liability by more than 5 in order for it to actually end up costing the consumer more, as each subsequent $ of insurance is cheaper than the last, because it's less likely to be used. Just because you might carry a $1M umbrella policy doesn't mean that somebody will get $1M if you cause them to break a finger(for example).
But from a business standpoint I could see car companies pushing to have their liability limited - in this day and age, with limits where they are now, the 80% thing, and such, I think $1M per person limit from congress would make the businesses looking to put out autonomous cars seriously consider it.
That's more than enough to drive a business out of business if it's auto-drive system is truly flawed, but more than enough to screw those HIT by bad drivers less than they currently are, while saving drivers/riders enough money to save them money in the saving of lives.
I can already predict crashes due to hacking/ buggy softwares and etc.
I can too, which is why I always assume that at least *some* accidents will still occur. For that matter, I figure that the profiles of accidents will change - moving away from things like 'ran through a red light' and towards 'drove into a not completely correctly marked construction zone'.
Still, the question becomes one of how the autonomous cars affect the average rate and severity of crashes. If it cuts off 50%, concentrated in the extreme, more commonly fatal crashes, it's a real benefit to society.
However, what I dislike is how it is being presented here.
If you follow the links you'll see a mention of cutting accident rates by 80%. I tend to view 50% as a good dividing line for starting adoption, 80% just makes the case so much tighter.
Seriously: no, I cannot* imagine that the autonomous cars are actually better than the average intoxicated driver, according to modern limits. Your average intoxicated driver gets home safe and doesn't put anybody at risk. Given what I know about the difficulty in training a computer to behave intelligently in unplanned situations, I cannot imagine that autonomous cars are significantly better than your average intoxicated driver.
Excepting your disclaimer, I'd say that you have a very limited imagination. Personally, I can easily imagine it. Indeed, with the Google cars and such I'd say that we're very very close. Especially if 'perfection' isn't your standard, merely 'median human driver', IE 'better than the median drunk driver'. Now, I do imagine that the accidents would be different. I picture our AI cars being very good at avoiding accidents via 'fast twitch' actions, trading that for 'duh' accidents that no human would get into. The sort of accidents we see when human drivers get too stupid about following GPS directions, essentially. Things like driving into a flooded ford, through construction, etc...
When I figured it out, I pegged a self-driving car with merely 'half' the accident rate of humans to be worth a couple grand a year to me, and vastly more to those who either drive more or are outright bad drivers. I figured that DUI convicts would be among the first adopters - forced to do so by the courts, as interlocks are both expensive and hard to use, and I'd imagine the insurance break would be substantial at that point.
That's already been known since like the '80s. My grandfather has spinal damage from both an accident and polio. He's been using a electrostimulation machine for decades, slowly increasing/restoring function in his legs.
The trick would be something that works about 10X faster so you could 'finish' in a couple decades rather than it being essentially for life....
What about purely online services? I haven't encountered this yet, and I'm sort of surprised. I'm using a public wifi outside of my home country, and that triggers me using my private VPN times two. I have a VPS I mess around with, set up VPN on it. I've used it to access things like Netflix, which isn't available in my current country, get the 'correct' steam pricing, etc... If anybody really wanted to they could track me down from that IP address, but it'd probably require a warrant.
But my VPS isn't even in the same time zone as my home, and I've recently changed addresses along with the move.
I guess that if anybody asks I might not respond with 'no' to whether I'm on vacation, but with 'close, I'm on a business trip'.
Wouldn't growing a fodder crop like wheat be the same as fallow?
Negative, you need a nitrogen fixing crop like Alfalfa, Soybeans, etc...
Wheat is pretty much the definition of a 'depleting crop'.
Uh, that's what I said? Italians and such are much closer to 'white' in skin tone than they are to the really black blacks. Of course, blacks themselves range a wide gamut.
Good point; I thought a 3M level UID was too high for somebody of that history; However while cap-I wolf isn't as distinguished as lower-l wolf, he still had a fairly normal posting history until recently. I did check.
It's AlphaWoIf_HK, check the hidden comments attached to my original post - he's either infected with malware or had his account hacked, what with posting 4 duplicate 'mycleanpc' spam posts. Given that he has a normal posting history until very recently(and I remember other posts of his). Since this is an IT Pro focused site, it's 'his fault' in either case. ;)
You're lucky HK screwed up, or I'd have never seen this. I know that there's a whole rainbow of colors in Africa, which is why I put 'straight from Africa black' in quote marks. The idea is to give readers a vision of the color in question, not to quibble about the pigmentation range of the continent.
Sorry to see that your account has apparently been hacked, HK. Or some malware is messing with your posting ability. Suggest you change your passwords and whatnot...
Just don't install 'mycleanpc'.
Is this for real? Hollywood produced plenty of Italian American superstars, as well as Latinos. How did Bollywood manage?
There's a vaster difference in skin tone between Italians, Latinos, and Indians compared to 'Straight from Africa Black' than they are from European. Most Europeans can come close to those shades with sufficient tanning.
That's just looking at it from a technical perspective. I still believe that the 'problems' were overstated, but then, given the tendency for makeup departments to do extensive work-ups on movie stars so they look 'normal' on screen, for clothing designed to present a certain appearance on screen as opposed to in real life, and such I can believe that there were issues at some point. After that it becomes a continuing education problem - you need a makeup department that knows how to properly prepare black people for their role on screen, and you might not have that with all crews due to lingering effects from racism, inertia, etc...
capture an accurate colour spectrum
The problem with your premise is this assumption. You have to remember that Cinematography wasn't about accuracy, it was about looking good. Remember all the shortcuts taken during the original Star Wars trilogy? A major function through history was fooling your eyes. 'Good Enough' was everywhere.
Considering that there are lighting changes you want to do when you're photographing a black object vs a white one - Say a black PC Case vs a white PC Case, I believe that they might indeed have a point, if you assume that the directors, photographers, lighting directors and all are incredibly stupid.
Thinking back, I remember massive numbers of black people in films, all through history. The 'washed out appearance' that the article complains about I only really remember from the really early B&W films. I think lighting would be a minor issue up against the inertia of racism - there were plenty of 'black' films starting from back in the '70s to develop any necessary technique changes. It has to do more with the population of the USA being only ~18% black*.
*Number pulled from memory
I don't get it... you go through all that trouble to purchase in a manner you feel is anon on silk road, but then you post about it on slashdot using your registered account?
Given the nature of the internet, it can easily be argued that he's lying his ass off. Even if he isn't, 'small hash order' indicates he's a user, not a dealer, thus *on average* incredibly unlikely to be a worthy target for the 3+ agencies you'd need to coordinate with in order to track him down.
Off the top of my head - you'd need to get a warrant to get slashdot to disclose Cito's account and IP address information. Then you'd need to figure out WHERE in the world he is(presumably the USA). You have to hope that he was using home or at least work for his slashdot postings rather than using the same anonymous internet cafe. Once you've figured out where he is, you have to contact the appropriate state police agency to coordinate with, along with the postmaster general(assuming USPS was used as opposed to UPS/Fedex).
On Average it's just not worth it. They want dealers, not users.
I'm quite surprized that firefighters threw water at a litium fire (but most powder extinguishers are almost as bad), and punched extra holes in it (to let the water out, maybe?), but not in a good way.
Yeah, I only ever made it into 'introductory' fire fighting and even I know that tossing water on a metal fire is generally a very bad idea.
the average user can still install everything on their C: drive without making any changes from the default installation.
\
This one is important. I purchased the largest SSD I could reasonably afford(fiscally conservative), and found that messing with manually shifting programs around(lots of games) was costing me more time than what the SSD saved in quicker response times.
I'd like to buy a combo with a larger SSD, or see an automatic program capable of managing the caching, but these drives address a potentially huge market - almost as cheap as a HD solution while retaining ~90% of the 'real world' performance of an SSD.
I'd like to see a 16-64GB cache solution so the HD doesn't have to be anywhere near as aggressive in pruning stuff from the SSD, but I'm seriously interested.
I'm almost surprised that they didn't go to this first thing - water isn't good for gasoline fires either, though the sheer amount a fire engine can put out will often put out fires that water would not otherwise be recommended for.
they most commonly occur when an electrical problem interacts with a dirty engine (usually because an oil seal started failing).
And how much oil does a Tesla need to have in it? How much heat does it produce to encourage any oil to catch fire?