And even using flawless multiwalled carbon nanotubes (even higher strength-to-weight ratio), it would only be *barely* strong enough to support its own weight - and no engineer worth their certification would sign off on a structure with a safety margin probably measured in single-digit percentages - especially not considering the devastation it would cause if it fell to Earth (some designs suggest designing it to vaporize in the atmosphere, but I'm not sure that would actually be a dramatic improvement.)
That only applies to "beanstalks" though. Personally, I'm a fan of tumbling cable / orbital wheel space elevators - takes a bit more coordination to get on, and you have to first get at least most of the way out of the atmosphere on your own, but they can operate with no moving parts, and serve as 100% efficient "angular momentum banks" to transfer angular velocity between up- and down-bound payloads. Plus they can service an entire Great Circle around the planet, instead of just a single ground station.
It is. Think of it not a tower rising up from the Earth, but a "rope" being dropped from geostationary orbit, all the weight is supported from above - you couldn't hope to keep it from crumbling under its own weight as a tower. And as you climb the rope, it has to keep getting thicker. At the bottom, it only has to support your weight. A mile up it has to support your weight, plus the weight of a mile of rope. Let's say it's ultralight, so that that mile of rope only weighs as much as you - then at that point you actually need TWO ropes that size - one to support you, and one to support the first mile of rope. At two miles up it takes 4 ropes - to support you, a mile of 1x rope, and a mile of 2x rope. At three miles that becomes 8, to support you+1x+2x+4x. By the time you hit Geostationary, at 22,236 miles (35,786km) elevation that rope is going to get insanely thick.
NOTE: that math is wrong, but captures the general concept. Actually the rope will have to get thicker even faster at first, at least until it get high enough that Earth's gravity begins to fade noticeably, and angular velocity begins to be an appreciable fraction of the amount required to maintain obit, but you'll be thousands of miles up by then.
After you reach Geostationary the rope starts starts thinning, since beyond that point additional rope is now moving too fast for it's orbital height, and will be trying to "fall" away from the Earth. At that point you can either attach to a counterweight (usually imagined as an asteroid), or just keep making more cable to do the same job. More cable is obviously more difficult, but comes with the advantage that you could essentially do a "firepole slide" away from the Earth to be launched onto an interplanetary trajectory.
The problem is the taper rate is actually exponential rather than linear, and if your strength-to-weight ratio isn't high enough, like steel for instance, then before you reach geostationary orbit your cable has to be wide enough to completely encase the Earth in a giant steel shell just to be able to support it's own weight.
Which, admittedly would then be much more capable of supporting its own weight, so you COULD do it if you were able to magic up several times the Earth's mass in steel. But I would have some pretty serious objections to the idea of completely blocking off the Earth from the sun, moon, and stars.
Always? It hasn't even been an issue but for a couple centuries. Centuries during which the myth of perpetual growth was also in ascendancy thanks to effectively unlimited resources, which are now also reaching their limit.
And in case you haven't noticed, the quality of replacement jobs has been declining for some time now.
The biggest threat UBI addresses is automation, and nobody has yet proposed an effective way to tax that. There's a persistent myth that automation will always simply open new opportunities, new kinds of jobs, while eliminating the old ones, but that ignores exactly what is being automated.
Consider: the industrial revolution with its rapid advancement in mechanized labor largely eliminated those whose place in the workforce was providing brute strength. Humans have a lot more to offer than that, but the labor market for horses and mules never recovered, contributing heavily to the US population crashing from around 22 million in 1900 to only 3 million by 1960.
Today, increasingly dexterous robots are pushing humans into ever-smaller roles on the assembly line, it won't be long before their dexterity exceeds our own, and the only role for humans there will be in roles exercising judgment. Meanwhile AI is rapidly catching up with us in terms of domain specific judgement - we already have AI beginning to outperform lawyers, pathologists, etc. in specific contexts. And there's just not really a whole lot of demand for broad-spectrum good judgement within the economy - and frankly, there's not a lot of humans that possess it anyway. That pretty much leaves what, art and receptionists?
And everyone working those service and luxury jobs will still need to buy all the staples - food, housing, durable goods, so the flow of wealth from the population to the production industries will continue unabated despite automation, but there will no longer be a reciprocal flow of wealth from the production industries to the general population (the former blue-collar employees), which is completely unsustainable. It only takes a handful of people to maintain and manage a fully automated factory or farm, and they can't hope to buy enough service and entertainment to transfer enough wealth to the general population to allow them to purchase all the goods being produced.
When was your youth? A quick google suggests the first "video" (interactive CRT display) game was Space War in 1962, running on a PDP-1 the size of a large car. Computer games went back even further - "Bertie the Brain" played tic-tac-toe in 1950. When pong came out in 1972 it was already an extremely crude "retro" game, it's claim to fame was that you could play it on the tiny little Atari entertainment system, a computer affordable by middle class households.
Meanwhile sprites, hardware windows, etc, weren't exactly giant leaps of innovation - yes, they allowed impressive new capabilities, but the technology itself was relatively simple - the "low hanging fruit" of graphical interface technology. Which was the point I was trying to convey in the second half of my post - early advancement seems very rapid, because those early advancements may be large conceptual leaps, but don't require large technological leaps to implement, so you can get a lot of them for a relatively small amount of effort. Once the low-hanging fruit has all been picked the same amount of invested effort yields far less obvious benefits.
I think there's more to it than that - not only were you more excited in your youth, but you had far less personal perspective. The progress of your youth was no less incremental, but you hadn't already spent decades watching the precursors.
There's also a legitimate external component though, if you're discussing computer technology specifically - 80s and 90s were sort of the golden age of computing: impressive computing power had just becoming accessible to the public, and its performance was accelerating rapidly. Meanwhile the entire field of computing and information technology was still very much in its infancy - the enabling technology had only just become affordable enough to be widely explored, and there was a resulting explosion in surrounding innovation as people figured out just what could be done with the new tools at their disposal. Absolute progress is probably no slower today, but we tend to see things in term of relative progress, and by those lights there's a much bigger difference between 2 and 3 than between 19 and 20.
As a matter of fact, anonymity is an extremely challenging feature to add. Certainly Bitcoin never really attempted it, despite all the early hype to the effect that it did.
Quite. In fact I recall a study done analyzing violent crime rates around the world that found in every country violent crime started climbing ~20 years after the introduction of leaded gasoline, and began falling ~20 years after it was eliminated. Basically, if you grew up breathing lead fumes, you were more likely to commit violent crimes as an adult. The fact that every country introduced and banned leaded gasoline at different times helps to eliminate most other confounding factors that might have been responsible.
Not at all surprising as a social observation, considering we know that on an individual level lead exposure in childhood tends to boost aggressiveness while reducing impulse control.
I'm not - which is why I leave location services off on my phone, as would any non-stupid criminal.
My objection is that a warrant for information about "every person who was in a 7 city block area in a 2.5 hour window" is ridiculously over-broad, and will almost certainly put dozens if not hundreds of innocent people under suspicion, while not giving any clue whatsoever about the actual criminal unless they were bone-headedly stupid.
It's only a stone's throw from outright government mass surveillance (which I hope you understand the dangers of) - in any city there's almost certainly several crimes within a 10th of a mile of you on any given day. Only it's even worse because it's completely blind to any even marginally intelligent criminal, and thus can only be used against the innocent and the idiot criminals - and if your criminal is an idiot then it should be easy enough to catch them through real police work.
So if location data of every person using google on their phone in a 7 city block area over a period of 2.5 hours is "limited", where exactly would you draw the line?
Yes, crashing in an aircraft almost always has much worse consequences. That's the nature of the beast, and doesn't much matter who's driving. It's offset by the fact that there are very few unpredictable problems. Mechanical failure is all but eliminated by proper maintenance, and collision all but eliminated by proper vigilance. Flying is already far safer than driving, and almost all of the flight-related risk is concentrated in the moments surrounding take-off and landing. Moments which I suspect will generally be FAR safer in a VTOL multicopter - it's almost stationary as it approaches the ground, and lacks the high angular momentum of a helicopter's large engine and rotors.
Autonomous flying is a far simpler nut to crack than autonomous driving, especially if you give it the option to simply refuse to fly in adverse conditions.
Consider that in the air you have no curbs, buildings, trees, pedestrians, or intersecting traffic to deal with. Go straight up until clear of all obstacles, fly to destination in straightest-line manner possible, land vertically, avoiding all obstacles. If you restrict it to open fields and pre-authorized landing pads (as you would almost certainly have to without special regulatory consideration) there's very little opportunity for anything to go really wrong. I'm pretty sure in-air accidents are virtually always caused by either mechanical failure (i.e. improper maintenance) or collisions (inattentive pilots). And if nothing else a computer offers unwavering attention, and can relatively easily be given a full sphere of vision with no compromises.
I'm curious, how do you get from "bought software for use on 38 machines" to installing it on 100,000 while "the number of active licenses doesn't exceed the ones it bought"
I'm really hoping one day we get our collective heads out of our asses and start demanding standardized components for consumer vehicles. I mean sure, you could have a million different seat designs, but they should all attach to the same mounting holes and power connector (if applicable). For ICE's, is there really a need for more than maybe 4 or 5 different models of alternators, air filters, etc.,etc.,etc.? EVs potentially make that a much simpler proposition as well, electrical connections are far more flexible, you just have to agree on voltage and current specs.
Mileage taxes are being considered in many places, even implemented in a few.
But the fact of the matter is that road damage is done almost exclusively by commercial freight vehicles and weather. Damage goes up with the cube, or 5th power, or some such of the weight of the vehicle. Something ridiculous that means it takes many thousands or millions of cars to do as much road damage as one semi.
So basically most of your gas taxes are already going to subsidize commercial vehicles that are causing the road damage. Mileage taxes would have the advantage that they could be distributed much more fairly based on vehicle weight, though doing so would of course result in higher shipping costs. Of course that might possibly drive more long-distance shipping back to the far more efficient rail system, as well as being an additional small force towards "buy local" campaigns.
The level of inconvenience largely depends on the specific habits user - if you can easily charge at home and mostly just drive around town within range of a single charge, then the "inconvenience" is only plugging your car in when you get home at night, and in exchange you never have to stop at a gas station, get an oil change, replace brake pads, etc. In which case an EV is actually considerably less inconvenient.
If you have a really long commute (and can't reliably charge at work) , or go on road trips on a regular basis, then yeah, there is a much larger inconvenience factor, and maybe an E.V. isn't right for you. At least not as your only vehicle.
>But Teslas, especially out of warranty, are rather expensive to own.
Could you expand on that? My understanding was that EVs in general have very low maintenance requirements - at least until the batteries need to be replaced. Do Teslas suffer from high failure rates in other components as well?
This site https://www.fueleconomy.gov/fe... says that the Tesla Model S AWD gets about 98MPGe, or if you want less "equivalents"and more hard numbers - 35kWh/100miles.
So cost per mile: ICE: $3/gallon * (1 gallon / 25 miles) = $0.120/mile Tesla: $0.24/kWh * (35kWh/100miles) = $0.084/mile Ignoring purchase and maintenance costs of course.
And of course the superchargers are intended for occasional, rushed charging with the assumption that most of the people most of the time will use home/work trickle chargers paying market rates of closer to $0.12/kWh. or about $0.042/mile.
Another thing to consider is that for most of the time the Falcon Heavy was being designed and built, the plan was seemingly for the BFR to be an interplanetary vessel, too large to satisfy normal launch demands. It's only in the last year or so I think that the plan was scaled down to something that could service existing Earth launch needs as well, when they realized that there was a sweet spot in size that could be cheaper to operate than an F9, while still being able to make it to Mars with a respectable payload. Thus solving the single biggest problem with the original BFR design - how to pay for them.
The Heavy may prove to have been a waste of time and money, but I don't know that I'd call it a mistake - it was simply rendered obsolete by a change in strategic vision. Strategic revisions almost always carry a cost like that, it's just part of the price of not being omniscient. It may even be that lessons learned while developing the Heavy were instrumental in revising the BFR plan, making the Heavy development a valuable investment regardless of the value of the rocket itself..
For all the other qualifiers you added, I'm not sure you actually added much difficulty. Refueling in orbit in preparation for a trip to Mars will be a new challenge. Refueling on Mars though will not be - making the fuel maybe, but that has nothing to do with the rocket. The only difference between refueling on Mars rather than Earth will be dealing with ambient vacuum, and considering the pressure the fuel is already under, I'm not sure that a one-atmosphere difference in ambient pressure is particularly relevant.
As for the heavy reusability though, that is indeed the major challenge, far beyond merely scaling up existing technology. In fact it's arguably a large portion of the reason behind the BFR in its current incarnation: small enough to profitably service Earth-orbital launch needs, and just big enough to get to Mars, and to reduce the relative mass of reentry systems to still allow for respectable payloads. It's a redesign made with business considerations front and center, and its success will ride on them being effective.
I'm not sure it would actually be all that substantially cheaper, but you might be on to something. Fewer engines at least, and the tanks are smaller. Even if it's only 30% cheaper... well that much less loss if it crashes or explodes.
There's also the fact that the booster is completely useless without the Ship, while the Ship has independent market possibilities. The suborbital transportation he's mentioned - for passengers perhaps, but potentially also for cargo to remote locations. If it's supposed to land on Mars or the Moon without infrastructure, then it should be able to land on any solid patch of clear ground here on Earth. If nothing else it could be a good way to start racking up flight hours and landing counts before the main booster is ready.
It'd be an expensive novelty, but maybe not ridiculously so - methane is cheap, and the goal at least is to radically reduce refurbishment costs. And the real win could just be getting the potential into popular consciousness. Give the idea a few extra years to "ferment" before the Ship fleet begins to grow large and reliable (aka cheap) enough to be effectively put to work.
And of course there's the fact that the second stage is in many ways actually the more complicated system - it's got a lot of aerodynamic reentry considerations to worry about, something SpaceX hasn't really dealt with before. They'd probably love to get the kinks worked out of that as much as possible at comparatively low speed before trying a reentry from orbital speeds. And of course there's the fact that they'll be running completely new methane-burning engines, which may have their own kinks to work out that don't show up in static firing tests.
Slashdot Mirror
And even using flawless multiwalled carbon nanotubes (even higher strength-to-weight ratio), it would only be *barely* strong enough to support its own weight - and no engineer worth their certification would sign off on a structure with a safety margin probably measured in single-digit percentages - especially not considering the devastation it would cause if it fell to Earth (some designs suggest designing it to vaporize in the atmosphere, but I'm not sure that would actually be a dramatic improvement.)
That only applies to "beanstalks" though. Personally, I'm a fan of tumbling cable / orbital wheel space elevators - takes a bit more coordination to get on, and you have to first get at least most of the way out of the atmosphere on your own, but they can operate with no moving parts, and serve as 100% efficient "angular momentum banks" to transfer angular velocity between up- and down-bound payloads. Plus they can service an entire Great Circle around the planet, instead of just a single ground station.
It is. Think of it not a tower rising up from the Earth, but a "rope" being dropped from geostationary orbit, all the weight is supported from above - you couldn't hope to keep it from crumbling under its own weight as a tower. And as you climb the rope, it has to keep getting thicker. At the bottom, it only has to support your weight. A mile up it has to support your weight, plus the weight of a mile of rope. Let's say it's ultralight, so that that mile of rope only weighs as much as you - then at that point you actually need TWO ropes that size - one to support you, and one to support the first mile of rope. At two miles up it takes 4 ropes - to support you, a mile of 1x rope, and a mile of 2x rope. At three miles that becomes 8, to support you+1x+2x+4x. By the time you hit Geostationary, at 22,236 miles (35,786km) elevation that rope is going to get insanely thick.
NOTE: that math is wrong, but captures the general concept. Actually the rope will have to get thicker even faster at first, at least until it get high enough that Earth's gravity begins to fade noticeably, and angular velocity begins to be an appreciable fraction of the amount required to maintain obit, but you'll be thousands of miles up by then.
After you reach Geostationary the rope starts starts thinning, since beyond that point additional rope is now moving too fast for it's orbital height, and will be trying to "fall" away from the Earth. At that point you can either attach to a counterweight (usually imagined as an asteroid), or just keep making more cable to do the same job. More cable is obviously more difficult, but comes with the advantage that you could essentially do a "firepole slide" away from the Earth to be launched onto an interplanetary trajectory.
The problem is the taper rate is actually exponential rather than linear, and if your strength-to-weight ratio isn't high enough, like steel for instance, then before you reach geostationary orbit your cable has to be wide enough to completely encase the Earth in a giant steel shell just to be able to support it's own weight.
Which, admittedly would then be much more capable of supporting its own weight, so you COULD do it if you were able to magic up several times the Earth's mass in steel. But I would have some pretty serious objections to the idea of completely blocking off the Earth from the sun, moon, and stars.
I quite agree.
Always? It hasn't even been an issue but for a couple centuries. Centuries during which the myth of perpetual growth was also in ascendancy thanks to effectively unlimited resources, which are now also reaching their limit.
And in case you haven't noticed, the quality of replacement jobs has been declining for some time now.
The biggest threat UBI addresses is automation, and nobody has yet proposed an effective way to tax that. There's a persistent myth that automation will always simply open new opportunities, new kinds of jobs, while eliminating the old ones, but that ignores exactly what is being automated.
Consider: the industrial revolution with its rapid advancement in mechanized labor largely eliminated those whose place in the workforce was providing brute strength. Humans have a lot more to offer than that, but the labor market for horses and mules never recovered, contributing heavily to the US population crashing from around 22 million in 1900 to only 3 million by 1960.
Today, increasingly dexterous robots are pushing humans into ever-smaller roles on the assembly line, it won't be long before their dexterity exceeds our own, and the only role for humans there will be in roles exercising judgment. Meanwhile AI is rapidly catching up with us in terms of domain specific judgement - we already have AI beginning to outperform lawyers, pathologists, etc. in specific contexts. And there's just not really a whole lot of demand for broad-spectrum good judgement within the economy - and frankly, there's not a lot of humans that possess it anyway. That pretty much leaves what, art and receptionists?
And everyone working those service and luxury jobs will still need to buy all the staples - food, housing, durable goods, so the flow of wealth from the population to the production industries will continue unabated despite automation, but there will no longer be a reciprocal flow of wealth from the production industries to the general population (the former blue-collar employees), which is completely unsustainable. It only takes a handful of people to maintain and manage a fully automated factory or farm, and they can't hope to buy enough service and entertainment to transfer enough wealth to the general population to allow them to purchase all the goods being produced.
Sorry, Pong was made by Atari, but for a dedicated console, the cartridge-based entertainment system wouldn't come out until 1977.
When was your youth? A quick google suggests the first "video" (interactive CRT display) game was Space War in 1962, running on a PDP-1 the size of a large car. Computer games went back even further - "Bertie the Brain" played tic-tac-toe in 1950. When pong came out in 1972 it was already an extremely crude "retro" game, it's claim to fame was that you could play it on the tiny little Atari entertainment system, a computer affordable by middle class households.
Meanwhile sprites, hardware windows, etc, weren't exactly giant leaps of innovation - yes, they allowed impressive new capabilities, but the technology itself was relatively simple - the "low hanging fruit" of graphical interface technology. Which was the point I was trying to convey in the second half of my post - early advancement seems very rapid, because those early advancements may be large conceptual leaps, but don't require large technological leaps to implement, so you can get a lot of them for a relatively small amount of effort. Once the low-hanging fruit has all been picked the same amount of invested effort yields far less obvious benefits.
I think there's more to it than that - not only were you more excited in your youth, but you had far less personal perspective. The progress of your youth was no less incremental, but you hadn't already spent decades watching the precursors.
There's also a legitimate external component though, if you're discussing computer technology specifically - 80s and 90s were sort of the golden age of computing: impressive computing power had just becoming accessible to the public, and its performance was accelerating rapidly. Meanwhile the entire field of computing and information technology was still very much in its infancy - the enabling technology had only just become affordable enough to be widely explored, and there was a resulting explosion in surrounding innovation as people figured out just what could be done with the new tools at their disposal. Absolute progress is probably no slower today, but we tend to see things in term of relative progress, and by those lights there's a much bigger difference between 2 and 3 than between 19 and 20.
As a matter of fact, anonymity is an extremely challenging feature to add. Certainly Bitcoin never really attempted it, despite all the early hype to the effect that it did.
Quite. In fact I recall a study done analyzing violent crime rates around the world that found in every country violent crime started climbing ~20 years after the introduction of leaded gasoline, and began falling ~20 years after it was eliminated. Basically, if you grew up breathing lead fumes, you were more likely to commit violent crimes as an adult. The fact that every country introduced and banned leaded gasoline at different times helps to eliminate most other confounding factors that might have been responsible.
Not at all surprising as a social observation, considering we know that on an individual level lead exposure in childhood tends to boost aggressiveness while reducing impulse control.
They didn't start removing the lead from gasoline, they just stopped intentionally adding it.
I'm not - which is why I leave location services off on my phone, as would any non-stupid criminal.
My objection is that a warrant for information about "every person who was in a 7 city block area in a 2.5 hour window" is ridiculously over-broad, and will almost certainly put dozens if not hundreds of innocent people under suspicion, while not giving any clue whatsoever about the actual criminal unless they were bone-headedly stupid.
It's only a stone's throw from outright government mass surveillance (which I hope you understand the dangers of) - in any city there's almost certainly several crimes within a 10th of a mile of you on any given day. Only it's even worse because it's completely blind to any even marginally intelligent criminal, and thus can only be used against the innocent and the idiot criminals - and if your criminal is an idiot then it should be easy enough to catch them through real police work.
So if location data of every person using google on their phone in a 7 city block area over a period of 2.5 hours is "limited", where exactly would you draw the line?
Yes, crashing in an aircraft almost always has much worse consequences. That's the nature of the beast, and doesn't much matter who's driving. It's offset by the fact that there are very few unpredictable problems. Mechanical failure is all but eliminated by proper maintenance, and collision all but eliminated by proper vigilance. Flying is already far safer than driving, and almost all of the flight-related risk is concentrated in the moments surrounding take-off and landing. Moments which I suspect will generally be FAR safer in a VTOL multicopter - it's almost stationary as it approaches the ground, and lacks the high angular momentum of a helicopter's large engine and rotors.
Autonomous flying is a far simpler nut to crack than autonomous driving, especially if you give it the option to simply refuse to fly in adverse conditions.
Consider that in the air you have no curbs, buildings, trees, pedestrians, or intersecting traffic to deal with. Go straight up until clear of all obstacles, fly to destination in straightest-line manner possible, land vertically, avoiding all obstacles. If you restrict it to open fields and pre-authorized landing pads (as you would almost certainly have to without special regulatory consideration) there's very little opportunity for anything to go really wrong. I'm pretty sure in-air accidents are virtually always caused by either mechanical failure (i.e. improper maintenance) or collisions (inattentive pilots). And if nothing else a computer offers unwavering attention, and can relatively easily be given a full sphere of vision with no compromises.
I'm curious, how do you get from "bought software for use on 38 machines" to installing it on 100,000 while "the number of active licenses doesn't exceed the ones it bought"
Sadly that makes perfect sense.
I'm really hoping one day we get our collective heads out of our asses and start demanding standardized components for consumer vehicles. I mean sure, you could have a million different seat designs, but they should all attach to the same mounting holes and power connector (if applicable). For ICE's, is there really a need for more than maybe 4 or 5 different models of alternators, air filters, etc.,etc.,etc.? EVs potentially make that a much simpler proposition as well, electrical connections are far more flexible, you just have to agree on voltage and current specs.
Mileage taxes are being considered in many places, even implemented in a few.
But the fact of the matter is that road damage is done almost exclusively by commercial freight vehicles and weather. Damage goes up with the cube, or 5th power, or some such of the weight of the vehicle. Something ridiculous that means it takes many thousands or millions of cars to do as much road damage as one semi.
So basically most of your gas taxes are already going to subsidize commercial vehicles that are causing the road damage. Mileage taxes would have the advantage that they could be distributed much more fairly based on vehicle weight, though doing so would of course result in higher shipping costs. Of course that might possibly drive more long-distance shipping back to the far more efficient rail system, as well as being an additional small force towards "buy local" campaigns.
The level of inconvenience largely depends on the specific habits user - if you can easily charge at home and mostly just drive around town within range of a single charge, then the "inconvenience" is only plugging your car in when you get home at night, and in exchange you never have to stop at a gas station, get an oil change, replace brake pads, etc. In which case an EV is actually considerably less inconvenient.
If you have a really long commute (and can't reliably charge at work) , or go on road trips on a regular basis, then yeah, there is a much larger inconvenience factor, and maybe an E.V. isn't right for you. At least not as your only vehicle.
>But Teslas, especially out of warranty, are rather expensive to own.
Could you expand on that? My understanding was that EVs in general have very low maintenance requirements - at least until the batteries need to be replaced. Do Teslas suffer from high failure rates in other components as well?
This site https://www.fueleconomy.gov/fe...
says that the Tesla Model S AWD gets about 98MPGe, or if you want less "equivalents"and more hard numbers - 35kWh/100miles.
So cost per mile:
ICE: $3/gallon * (1 gallon / 25 miles) = $0.120/mile
Tesla: $0.24/kWh * (35kWh/100miles) = $0.084/mile
Ignoring purchase and maintenance costs of course.
And of course the superchargers are intended for occasional, rushed charging with the assumption that most of the people most of the time will use home/work trickle chargers paying market rates of closer to $0.12/kWh. or about $0.042/mile.
Another thing to consider is that for most of the time the Falcon Heavy was being designed and built, the plan was seemingly for the BFR to be an interplanetary vessel, too large to satisfy normal launch demands. It's only in the last year or so I think that the plan was scaled down to something that could service existing Earth launch needs as well, when they realized that there was a sweet spot in size that could be cheaper to operate than an F9, while still being able to make it to Mars with a respectable payload. Thus solving the single biggest problem with the original BFR design - how to pay for them.
The Heavy may prove to have been a waste of time and money, but I don't know that I'd call it a mistake - it was simply rendered obsolete by a change in strategic vision. Strategic revisions almost always carry a cost like that, it's just part of the price of not being omniscient. It may even be that lessons learned while developing the Heavy were instrumental in revising the BFR plan, making the Heavy development a valuable investment regardless of the value of the rocket itself..
For all the other qualifiers you added, I'm not sure you actually added much difficulty. Refueling in orbit in preparation for a trip to Mars will be a new challenge. Refueling on Mars though will not be - making the fuel maybe, but that has nothing to do with the rocket. The only difference between refueling on Mars rather than Earth will be dealing with ambient vacuum, and considering the pressure the fuel is already under, I'm not sure that a one-atmosphere difference in ambient pressure is particularly relevant.
As for the heavy reusability though, that is indeed the major challenge, far beyond merely scaling up existing technology. In fact it's arguably a large portion of the reason behind the BFR in its current incarnation: small enough to profitably service Earth-orbital launch needs, and just big enough to get to Mars, and to reduce the relative mass of reentry systems to still allow for respectable payloads. It's a redesign made with business considerations front and center, and its success will ride on them being effective.
I'm not sure it would actually be all that substantially cheaper, but you might be on to something. Fewer engines at least, and the tanks are smaller. Even if it's only 30% cheaper... well that much less loss if it crashes or explodes.
There's also the fact that the booster is completely useless without the Ship, while the Ship has independent market possibilities. The suborbital transportation he's mentioned - for passengers perhaps, but potentially also for cargo to remote locations. If it's supposed to land on Mars or the Moon without infrastructure, then it should be able to land on any solid patch of clear ground here on Earth. If nothing else it could be a good way to start racking up flight hours and landing counts before the main booster is ready.
It'd be an expensive novelty, but maybe not ridiculously so - methane is cheap, and the goal at least is to radically reduce refurbishment costs. And the real win could just be getting the potential into popular consciousness. Give the idea a few extra years to "ferment" before the Ship fleet begins to grow large and reliable (aka cheap) enough to be effectively put to work.
And of course there's the fact that the second stage is in many ways actually the more complicated system - it's got a lot of aerodynamic reentry considerations to worry about, something SpaceX hasn't really dealt with before. They'd probably love to get the kinks worked out of that as much as possible at comparatively low speed before trying a reentry from orbital speeds. And of course there's the fact that they'll be running completely new methane-burning engines, which may have their own kinks to work out that don't show up in static firing tests.