Tesla Rolling Out Autopilot Software Updates to 1000 Cars
bloomberg.com
bloomberg.com
Whole point of gen2 hardware is it should allow fully autonomous cars, even in rain which others likely wont be able to do, once the software is fully figured out. Next year will be exciting to see how quickly they can push updates towards that goal, as well as how quickly regulation can catch up.
The driver still has complete control, authority and responsibility. There is really no reason to assume there will be more crashes.
http://99percentinvisible.org/episode/children-of-the-magent...
With self driving cars where we're relying on the car for moment-to-moment instantaneous reactions, I think it's very unlikely that human drivers will be able to pick up the slack in a system failure. If you're expecting the car to stop it self and it doesn't, you have what, a quarter second to jump in manually?
The recent video of a self driving uber going through a red light is probably a similar scenario. Even though uber insists it's a "driver assistance technology" and not a self driving car, the driver expected the car to stop itself, so they didn't bother applying the brakes. And that one was at low speeds, maybe 10 mph.
The saving grace here is that they're competing against human drivers. The technology doesn't have to be perfect, just better than us. But relying on human oversight to maintain 100% perfect attention for a six hour journey of doing literally nothing but watching a car drive itself in case you need to step in? I think that's a non-starter. For these to work, the car has to be able to manage on its own.
I imagine this has a similar effect.
The flip side is, I always have my foot resting lightly on the gas, so I'm not really at a disadvantage if I need to brake. Do Tesla drivers using autopilot have their hands hovering over the wheel typically? If not I imagine reaction time could be impacted.
In general, I hypothesise that the more a system is automated, the less awareness is demanded of the operator and thus the more likely the operator is to get into some kind of trouble, whether this stems from development of a lazy habit, rustiness or just not being able to recover the situation if the automation abruptly fails. In fact, we've seen several plane crashes attributed to pilot error related to increasing automation in aircraft.
https://www.tesla.com/blog/all-tesla-cars-being-produced-now...
Highlights:
* The manufacturer has in place and has provided the department with evidence of the manufacturer’s ability to respond to a judgment or judgments for damages for personal injury, death, or property damage arising from collisions or accidents caused by the autonomous vehicles produced by the manufacturer in the form of an instrument of insurance, or a surety bond... That's the key condition. If it crashes, the manufacturer is financially responsible and has to carry insurance or post a bond, just like human drivers.
* The manufacturer shall identify any commonly-occurring or restricted conditions, including but not limited to: snow, fog, black ice, wet road surfaces, construction zones, and geo-fencing by location or road type, under which the vehicles are incapable of operating in the autonomous mode and certify that the vehicles are incapable of operating in autonomous mode under those conditions. Reasonable enough.
* The manufacturer shall certify in the application that the subject autonomous vehicles are equipped with an autonomous technology data recorder that captures and stores autonomous technology sensor data for all vehicle functions that are controlled by the autonomous technology at least 30 seconds before and at least 5 seconds after or until the vehicle comes to a complete stop after a collision That's like the air bag data recorder cars have now, but more comprehensive.
* Certification that the vehicle complies with the “Vehicle Performance Guidance for Automated Vehicles” in the National Highway Traffic Safety Administration’s Federal Automated Vehicles Policy This is the national policy.[2] It's not very restrictive; it's mostly self-certification. This is where there's a big distinction between NHTSA levels 1 (basic auto-brake) and 2 (auto-brake plus lane keeping), and levels 3 and above. Basic concept: at levels 1 and 2, the driver must pay attention and systems must not let the driver stop paying attention. This is the "hands on the wheel" requirement. At level 3 and above, the driver doesn't need to pay attention in autonomous mode, and if anything goes wrong, the manufacturer is responsible.
* The national requirements require a "safety assessment letter" from the manufacturer to the NHTSA answering questions in a few key areas:
Data Recording and Sharing
Privacy
System Safety
Vehicle Cybersecurity
Human Machine Interface
Crashworthiness
Consumer Education and Training
Registration and Certification
Post-Crash Behavior
Federal, State and Local Laws
Ethical Considerations
Operational Design Domain
Object and Event Detection and Response
Fall Back (Minimal Risk Condition)
Validation Methods
* DMV reserves the right to revoke the self-driving license of a system that drives badly.
Passing certification gets a sticker for each car, like the stickers for low-emissions vehicles.
All this is far less restrictive than what's required for light aircraft.
Tesla needs to qualify for a license. Time to put up or shut up.
[1] https://www.dmv.ca.gov/portal/wcm/connect/211897ae-c58a-4f28... [2] https://one.nhtsa.gov/nhtsa/av/av-policy.html
It's interesting to see that the burden of insurance falls upon the manufacturer. I think that's fair during these pilots.
I still think there are a ton of legal questions that need to be addressed though. Should self driving car algorithms be standardized across the industry for safety? Because of the nature of saving lives, should self driving software be open for inspection and review in ways other software isn't? If people modify the software in the car, who is responsible? Can car "owners" legally modify code when safety and lives are of concern?
I had a post that addressed a few of these:
http://penguindreams.org/blog/self-driving-cars-will-not-sol...
That's what makes it all work without detailed regulation. Otherwise, there would have to be detailed technical regulations and pre-deployment testing. This way, the manufacturer is responsible. Period.
Volvo's CEO originally proposed that.[1] Volvo Car Group President and CEO Håkan Samuelsson said in 2015 that "the company will accept full liability whenever one of its cars is in autonomous mode." That's now US policy. Tesla and Uber probably hate that; both have tried to evade liability for bad autonomous driving. GM, Ford, and Chrysler seem to be OK with it; they've all survived big vehicle defect lawsuits and recalls. It's going to lead to much safer cars.
[1] http://fortune.com/2015/10/07/volvo-liability-self-driving-c...
>Should self driving car algorithms be standardized across the industry for safety?
Just like every other safety relevant feature of my car; Baseline requirements and tests standardised, implementation open to innovation and competition.
>Because of the nature of saving lives, should self driving software be open for inspection and review in ways other software isn't?
Same rules as the software of my ABS system, or of my airbag: it's up to the manufacturer, in practise nobody cares.
>If people modify the software in the car, who is responsible?
If I modify my brakes or my steering system I'm responsible for any damages caused by this modification. Why would software be different.
>Can car "owners" legally modify code when safety and lives are of concern?
Can I legally modify my breaking systems? Yes, but it has to fulfill the same road safety regulations, and frostin depending on what I do getting all required certifications might be impractical.
[0] https://www.tesla.com/blog/all-tesla-cars-being-produced-now...
It's not clear that LIDAR offers any real advantages over a camera suite combined with RADAR.
LIDAR has the same limitations as optical cameras with regards to rain, fog, etc.
As image processing capabilities continue to improve, the case for LIDAR gets weaker.
Actually, no. With appropriate signal processing, LIDAR units can see through rain and fog. The early noise returns from rain and fog can be gated out until the system gets a reflection from something solid. There are expensive LIDAR binoculars that can see through fog. You adjust the minimum range until you're seeing past the fog. So far, automotive LIDAR units don't have that feature, but they could.
Tesla has a level 4 autonomy demonstration video on their site. Maybe computer vision is superior to LIDAR; not a lot of human drivers with laser arrays on them driving around.
Yes, the software is the hard part. That's why it makes sense to use hardware that makes the job of the software easier. LIDAR tells you how far away objects are; if you're using cameras you have to do that with software instead. AFAIK no one yet knows how to do computer vision with cameras in the sorts of situations a self-driving car needs (though I'm not a vision researcher). That means that Tesla is betting on solving a research problem to get the cars they've already sold to start driving themselves. I always thought that was an insanely audacious business plan.
I don't think this term even means anything. The problem can be solved by various mixtures of hardware plus software to varying degrees of adequacy.
On the one hand, we know that it's possible to drive a vehicle with high reliability in a wide range of circumstances with just two relatively low resolution cameras... if those cameras are feeding into a trained human-level intellect. What combination of hardware and software is required to generate that intellect is still unknown.
On the other hand, the problem seems relatively well solved using 3D laser scanners and relatively straightforward software. It works pretty well and just needs to be shaken down in all of the possible circumstances it might encounter. It's also unreasonably expensive.
We seem to be converging on an "easiest" solution which uses a mixture of cameras, radars, ultrasound and maybe LIDAR, using deep learning for image processing. That's just one solution, though.
We probably can, but most of us drive with two cameras with approximately 120 megapixels of feature resolution with highly optimized non uniform pixel density and incredible dynamic range. Micromovements allow the software to boost the effective resolution by about a factor of four, to over 400 megapixels. Image stabilization works in multiple layers and executes stabilization movements within 10ms in the fastest layer.
The human eyes and vision system are certainly one of our more impressive parts, and we have a really hard time replicating their performance.
Lots of LIDARs, lots of cameras.
"Additional guidance data was provided by a video camera used to observe driving conditions out to eighty meters (beyond the range of the LIDAR) and to ensure room enough for acceleration."
"Stanley was characterized by a machine learning based approach to obstacle detection. Data from the LIDARs was fused with images from the vision system to perform more distant look-ahead. If a path of drivable terrain could not be detected for at least 40 meters in front of the vehicle, speed was decreased and the LIDARs were used to locate a safe passage."
This computer vision processing was performed with 1.6ghz Pentium M processors 11 years ago. Is it so far fetched Nvidia parallel processing hardware can't do better with CCDs than LIDAR?
The 100k car/driver thing is a red herring. Steering angle / brake / acceleration info is not the kind of data you need for this task. Big useless data is still useless data, even if it's big.
The right kind of data, if this is to be treated as an ML problem, is a 3D model of the environment (this would be the label), along with camera data (this would be the data). Ironically you need LIDAR to get those labels. It bears noting that the state of the art in visual SLAM is not ML based (it's really difficult to get the "ML" problem I just described to actually work - and it's not for lack of data, Google probably has petabytes of this data from their cars).
Those are labels for a different problem.
2) In theory almost everything is "just" a compute problem (just enumerate or search over all possibilities - of course this may not finish for millenia). But if Moore's law says the compute is 20 years away, I think it's reasonable to call it a research problem.
Not to say LIDAR isn't essential. I think LIDAR will get cheap enough that it'll just be silly not to have it.
IMO the most successful implementations are ones that can predict accidents that humans couldn't even see taking place (obscured by other cars, etc).
Such as Tesla'a forward facing radar, highlighted the other day on HN preventing an additional accident participant by detecting a vehicle slowing down ahead that the driver could not visually detect?
That being said, i would imagine said hardware would evolve over time with new releases. It would be nice, and probably safer, if cars could actually communicate with each other, rather than have to rely on their own cameras, to determine another vehicles position. Same with traffic lights, road signs, etc.
that's probably a long way off, although i wonder if Teslas have hardware that would allow them to communicate with one another.
This is being tested right now eg. 802.11p/ITS-G5/Vehicle-to-everything (V2X) communication https://en.wikipedia.org/wiki/Vehicle-to-everything
A video of some demonstrations by Ford Motors https://www.youtube.com/watch?v=PF1a-g9suR8
Tesla might already have the hardware needed; this implementation for Linux uses Qualcomm Atheros AR93xx chipset devices operating in the 5.9 GHz band https://rtime.felk.cvut.cz/publications/public/ieee80211p_li...
(https://events.ccc.de/congress/2015/Fahrplan/system/event_at... outlines the current security model - car manufacturers are a small set of organisations that could act as intermediary Certificate Authorities and manage key exchange so messages via these protocols can be "trusted")
I wonder how useful it would actually be though. I don't think you'd want to involve such a complex a distributed system in a hard-realtime domain like deciding whether you're about to hit the other vehicle, the CAP theorem seems pretty unavoidable there.
The software clearly didn't follow the flow of traffic; it was rigidly locked in going the speed limit. That may be OK, but it was a very empty weekend and everyone was going 10-15 miles over the speed limit. So had quite a chain of cars backed up behind him and like a half-mile in front of him before the next car.
It was clear the Tesla software wasn't smart enough to close the gap by accelerating above the posted limits, or move him over to the right-hand lane where he wouldn't block traffic -- we all ended up having to pass him on the right.
I couldn't help but think, "Great... just what we need more of... simulated old people driving slow in the left-hand lanes and cars that encourage jackasses on their phones to take pictures instead of paying attention."
But... to be clear, I still want one. =P
You guys need to figure out whether the speed limit is wrong, or breaking it is wrong, because you're not being consistent.
If traffic is flowing at 5MPH over the posted limit, it would be reasonable and prudent to match speed with surrounding traffic.
I'm not sure where to stand, but I do believe that the safe speed to drive it is more than just a function of "do what the sign says."
In Texas, at least Travis County, you can go up to 25 MPH over the posted limit and be in the gray area of "safe and reasonable" -- that gray area is determined by weather conditions, other vehicles on the road, and condition of your vehicle. Technically 1 MPH over the posted limit could get you a ticket, but anything under 10 MPH over, assuming clear weather, is easy to claim that you were just going with the flow of traffic. (If you ever get hit by the speed trap on I35 in Kyle I've used this to get out of a ticket, going 80 in a 65.)
The less you cause other drivers to have to change lanes, the better. The less traffic congestion you cause by driving slower than the flow of traffic, the better. This is compassionate driving. If Tesla makes cars that drive selfishly it's bad for Tesla -- it's not just about the person in the car, it's about the safety of everyone on the road.
(But be careful, you'll get downvoted on here just for implying that you have ever been somewhere other than SF.)
That said, the driver can still push the accelerator and go above the limit with AP still engaged.
Is that acceptable or are you breaking the law?
Meaning... if someone wants to cut in front of me... I don't have to choose to let them. But the Tesla seemed to welcome people cutting in front of it. I didn't get a clear look for very long, but a few times someone passed it and pulled in front of it they cut in too soon for the software's comfort so it applied the brakes. A human may do that as well, but realistically if I got passed on the right by a jerk who cut in front of me to show his anger for making him pass me... I probably would see he was going much faster than me and wouldn't slow down just because he was driving aggressive.
But yeah, I still want one. Will be interesting to see how all this stuff plays out. I personally would like to see the Tesla do 3 things to avoid pissing off other drivers on the road (especially for the foreseeable future while we have mixed vehicles on the roads)...
1) Close the gap in front of it. If there aren't any cars in front of you, do like humans do and accelerate a bit above the posted limits. 10 MPH above posted seems reasonable, but I'm sure someone can come up with an algorithm for what is appropriate in a 35 vs an 85 zone.
2) Don't drive like an American. Drive smart, and only use the left-hand lane for passing. People in other countries tend to be better about this, I think Tesla needs to incorporate some Autobahn training into their software.
3) Automatically repot to what it sees. We'd all be thankful if Teslas automatically reported to Waze, debris on the road, stopped cars, police ahead, etc. And took pictures and sent it in to the appropriate entity responsible for road maintenance. How cool of a service would that be for everyone if Tesla reported potholes and helped cities get more data around what needed to be fixed.
Might other cars arrive a little ahead of you? Sure. Is that a problem that warrants significantly different and more dangerous driving behaviour? No.
But how should a machine behave while there is still ego present? Tesla will share the road with humans for the foreseeable future.
For starters, don't be in the passing lane aggravating people I think would be sufficient. Or, to MVP it... just an alarm that beeps at drivers who have auto-pilot on if they are in the left-hand lane and a car is behind them but not in front of them. Tesla has all this sort of sensors already so it should be able to recognize that situation that warrants an internal cabin alarm to remind the human driver to use the turn signal and get over.
Tesla has to be mindful of human behavior here. Build software that makes their cars perceived as good drivers by human motorists. What I saw yesterday... the Tesla was driving like an asshole and slowing down traffic. Seems like an easy problem to fix. Too bad we can't have our Dad's ride around yelling at Tesla for a few hundred miles while it learns how to drive "right." Ha.
You should not be worried about this.
The Tesla should just drive safely, even if other people can get ahead.
Texas (28M people over 696,241 km2) is very different than states like New Jersey (9M people over 22,591 km2).
For giggles... Just one city in Texas, Houston, has a metro-area footprint of 26,060 km2 -- bigger than a 8 US States in terms of land area, and bigger than 33 states in terms of population -- again, that's just one city. Population density varies widely between our 254 counties as well (one county had about 4.5M people, another had 112). We aren't a one-size-fits-all sort of place, Texas or America.
* List of U.S. states by population density - Wikipedia || https://en.wikipedia.org/wiki/List_of_U.S._states_by_populat...
* List of countries and territories by population density - Wikipedia || https://en.wikipedia.org/wiki/List_of_countries_and_territor...
* Texas County Population Estimates, 2010-2015 Arranged in Descending Order | TSLAC || https://www.tsl.texas.gov/ref/abouttx/popcnty201011.html
France makes an excellent case study for this. They only implemented speed cameras and related safety measures around 2002/2003 but managed a remarkable drop in fatalities and average speeds [1][2].
Speed is insidious; we live in a three-dimensional world so drag increases quadratically with velocity, or going at 40mph requires four times as much power as at 20mph, and there is of course a corresponding increase in the kinetic energy in an impact. This is particularly important for pedestrians, where the difference between 20mph and 30mph is likely injury vs likely death.
I have no doubt that speeding cameras make the road safer, but they are being abused in the Netherlands, because they are a very good source of income.
For example, driving 4 kmh (2.5 mph) over the limit causes a €37 fine. The threshhold for fining has been lowered from 10 kmh to 4 kmh. Fines have quadrupled over the last 3 decades.
This all gets worse because we have 4 (80, 100, 120, 130) + dynamic as common limits on the highways, which change every few kilometres on some roads with only minimal signs. A large amounts of fines was actually declared invalid by court because of this reason.
Sorry... I'm not even from Texas, but I've lived here for 12 years. I do love it. And I find so much of it hilarious. The government here does everything it can to put itself out of business.
Speed cameras, for example, are banned in Texas.
* Texas Transportation Code - TRANSP § 542.2035 | FindLaw || http://codes.findlaw.com/tx/transportation-code/transp-sect-...
Please pick the right-most lane that does not back up traffic behind you. If traffic is backing up, it increases the likelihood of accidents. Vehicles that are close together for long periods of time are more likely to hit each other. Of course people should maintain proper following distance, but we need to take into account how people do drive, not how they should drive. The left-most lane should be kept free-flowing unless traffic is very heavy across all lanes.
Speeding is like smoking and littering. Sure, many people ignore the signs telling them not to and police tend to ignore violations, but that doesn't mean I must respect the violator. As Paul Graham said, the normal action is perversely abnormal.
This means that the lane to my right is generally crowded. Hitting the brakes in order to shift is probably worse than continuing at a constant pace. I'll let the tailgater take the risk of changing lanes.
That may be the case. I've tried to make very clear that I'm talking about the case of a multi-lane road that is not congested (not just locally congested). In that case, if you're impeding traffic from being able to pass you on the left, you should move right, at least temporarily, even if it means you need to slow down. I haven't seen you acknowledge this case. I have you repeat phrases like right-most lane that flows at the speed limit, which sounds to me like the following example:
Four-lane divided highway (two lanes in each direction). Right lane has traffic exiting and entering and slower moving traffic. You want to drive without needing to adjust speed for these vehicles, so you drive in the left lane. Vehicles attempting to travel faster than you will approach and stack up behind you. Because you want to maintain a constant pace, you won't temporarily move into the right lane to allow this traffic to pass.
I'll admit that sometimes I'll go 60 if the limit is 55
As others have pointed out, the flow of traffic is important regardless of posted speeds. The nearest interstate is posted 55 MPH. If you're going 65 MPH, you're moving slow relative to the flow of traffic, with speeds of 75–80 MPH being not unusual. I mention this only because you continue to bring up the posted limit. In my opinion, the flow of traffic is more important from a practical standpoint because that's the environment I'm operating in.
This means that the lane to my right is generally crowded. Hitting the brakes in order to shift is probably worse than continuing at a constant pace. I'll let the tailgater take the risk of changing lanes.
If you're asking drivers to pass you on the right into generally crowded traffic, this is very dangerous. The distance and time a vehicle needs to pass you is a lot more than the room needed for you to pull into the right lane, and for a shorter period of time. This type of behavior causes knots of congestion on multi-lane roads which increases the likelihood of accidents. Please do not do this.
Also, you shouldn't need to "hit the brakes". You may need to apply the brakes, but it shouldn't be abrupt. Paying attention to what's happening behind you as well as in front of you should give you plenty of time to signal and move right in a measured manner.
I've tried to be very clear here. If I'm still misinterpreting what you're saying, please do correct me, because that's not my intent. That said, please extend the same courtesy to me and be more explicit about what you would do in the specific cases I've presented.
Let's imagine rush hour on a 6-lane freeway (3 lanes in each direction). It's not the major congestion area, so traffic is dense but flowing at 60 on average. The right couple lanes are slower, with occasional clumps going at 45 or worse due to enter/exit merging.
To stay at 60ish and avoid issues with merging cars, in this scenario I'd drive in the left lane. Unfortunately, because my drive on this day happens to be about an hour long, I'll eventually find a speeder tailgating me. I saw the car approaching quite fast in my rearview, about 90 I'd guess. Before I can react, the car is following dangerously close, wiggling a bit, and flashing high beams. The lane to my right is significantly slower and crowded, making a shift non-trivial. My expectation is that the car behind me will attempt to pass on the right at earliest opportunity.
Should I slow down to merge right safely, antagonizing the car behind me further? I generally choose to maintain speed and stay in my lane. Back when I was a reckless jerk going 90, I enjoyed the thrill of weaving through traffic, thinking I was smarter than all the sheep plodding along. I'll let the car behind me enjoy that same thrill.
You might say that I should never have been in the left lane to begin with. After all, someone might want to drive a reasonable 70 mph and I'd be forcing that person to my slower 60. In that case, I'd need to go slower myself, around 50 or 55, and frustratingly modulate my speed as people shift in and out of my lane. I'm not clear on why the person who wants to drive 70 should be given priority. It seems to me that we are in quite similar situations. How about that person when there's someone who wants to drive 75? Or 80? What's the threshold?
When the car behind me doesn't appear so crazy and is following at a safe distance, I do shift right temporarily to let them pass. However, if someone is dangerously close, I feel I owe them no courtesy.
By all means, yes! You should also signal your intent to merge right. If the traffic to your right is as congested as you suggest, you'll be able to merge right before the following vehicle can pass you on the right, as they require more distance and time to do so than you do just changing lanes.
You might say that I should never have been in the left lane to begin with.
I've said no such thing, nor would I. I've been very clear, repeatedly, that you should move right if you're impeding traffic from passing you on the left. If that means that you need to move repeatedly between two lanes of traffic and you don't wish to do so, I suggest you modulate your speed to flow with the traffic in one of those two lanes. You need to take into account the environment you're in — the flow of traffic — not necessarily the ideal you may want.
As I've mentioned, if you impede traffic from passing you on the left, you create congestion which increases the risk of collision. There's no numerical threshold. If the vehicles are allowed to pass, the congestion is alleviated.
Allowing faster traffic to pass is not the same as entitling you to drive at a constant, slower speed while preventing them from passing. If you don't allow them to pass, you create congestion, regardless of their following distance.
If you expect them to pass on the right, you also increase risk. If you prevent someone from passing out of spite ("I feel I owe them no courtesy"), you create an even more dangerous situation, as you create tension within yourself which affects your driving behavior in addition to creating congestion. Indeed, if they're tailgating dangerously close, that should be even more motivation to merge right, not because they're driving in a responsible manner (they're not), but because you do what you can to ensure your own safety. On single-lane roads and someone is tailgating, I'll sometimes move onto the shoulder and slow down to let them pass for exactly this reason.
I've now repeated myself a number of times and can't think of anything else I want to add. If I've written something you don't understand or disagree with, please feel free to point it out and say so, and I'll do my best to express myself better.
What I'm trying to get at is that it's possible the faster driver ought to slow down rather than the slower driver speed up. Practicalities aside, just thinking of the ethics.
In nearly all cases I've argued that the slower vehicle should merge right, not speed up. The only case I've suggested increasing speed is when the driver of the slower vehicle doesn't want to modulate speed moving between two lanes to allow faster-moving traffic to pass. You're now concerned with the slower driver needing to speed up.
Given you're no longer contesting the original supposition, should I assume your agreement? You've now moved on to specifically construct a situation — significantly different from the original — where "it's possible the faster driver ought to slow down rather than the slower driver speed up." You're asking me to argue and potentially agree to this new situation, while not once having conceded or even shown understanding (or asked for clarification) of any of the points I've made so far. That said, I'll continue to do my best to argue in good faith and respond to your new situation.
On a single-lane road, a driver should drive the speed they're comfortable with the conditions, including traffic. A following vehicle will need to slow down to avoid rear-ending the lead vehicle. Absolute speeds and number of vehicles above 2 are irrelevant.
If the following vehicle is tailgating, the lead vehicle can maintain their speed, speed up, or move out of the way depending on opportunity. Those are the things they can control. They're not under any obligation to chose one over the others; it depends on their assessment of the situation.
Practicalities aside, just thinking of the ethics.
The rules of behavior really don't change. If everyone were behaving legally and safely (within the min and max speed limits, safe following distances), you still should move right to allow faster vehicles to pass on the left, and for the same reasons: prevent passing on the right, prevent congestion, both of which decrease risk of collision.
At this point, I can't think of anything else to add, nor am I interested in discussing new tangents. I sincerely hope you consider changing your driving behavior, as it will improve traffic flow and be safer all around. Be well, and drive safely.
That's how a debate works, no?
I wanted to look at a different scenario, because I think there wasn't really a difference in opinion to begin with. A good way to conclude an argument is to find a point both people can agree on and work back to the original topic.
In this case, it seems you misunderstood my initial comment as that I refuse to change lanes. In turn, I misunderstood your initial reply as telling me to avoid the left lane.
Perhaps I also exaggerated my refusal to change lanes in response to a tailgater. I didn't mention that I use my turn signal to let them know I'm looking for a chance to shift. If the opportunity does not immediately present itself, I begin to gradually slow down until I can change lanes. However, I find about half the time they're too impatient and don't give me the chance to shift safely.
Also, consider why do they drive that fast - maybe it's an emergency. Sure, in most cases it isn't, but what if this one is?
I do try to avoid the left-most lane because of the incidence of reckless jerks. However, if the other lanes are too slow, I will pass on the left.
The thought of an emergency is interesting. I struggle to imagine an emergency that does not warrant police or ambulance, but does warrant taking on the hazard of extreme speeds.
I don't know about the US, but in my country (central Europe), it's better (and legal, almost encouraged - police will escort you if you tell them the reason) to break all speed limits in case of transporting an injured person (in some places, the ambulance would be just too slow), a pregnant woman about to give birth, etc.
Reaction time of an ambulance is around 2 minutes until they deploy, 15 minutes until they get to you, 15 minutes until they get back to the hospital - meanwhile you can drive the person yourself in 20 minutes.
This sounds like correct behavior, passing on the left. What I am reacting to (as are others, I believe) is a desire to maintain a constant speed in the left lane, impeding faster-moving traffic from passing you on the left.
Indeed, I feel unsafe when someone is tailgating. However, I find their behavior is often unpredictable and they might try to pass on my right at any moment without signaling.
They will not pass on the right if you move right before they get close enough tailgate.
It gives a new meaning to the word "crash"