J Storrs Hall’s book Where Is My Flying Car?: A Memoir of Future Past, told me new things I didn’t know about flying cars. The book is long, and says many things about tech and the future, including some with which I disagree. But his main thesis is a contrarian one that I’ve heard many times from engineers over my lifetime. Which is good, because by putting it all in one place, I can now tell you about it, and tell you that I agree:
«In each wave energy available eventually reached a plateau»
In the graph above the "Henry Adams curve" seems to me an artifact of optics: first what matters is energy *per person*, else population growth adds noise, and then I see three different parts of the graph: the before-coal era, reached a plateau, the coal era where it reached a plateau in the interwar period, and then the post-coal, oil era where it also reached a plateau.
Looking at the same graphs for England would be more useful, because on a wider scale diffusion adds quite a bit of noise.
«The main reason nuclear power has tapered off as a power supply is because it heavily favors scale and size»
Not really in technical terms: most large nuclear power plant sites actually have several smaller reactors rather than a single huge one. The cost advantage of large reactors are fairly small. Indeed at some point RR, that manufactures a rather tiny number of "small" nuclear power plants for submarines, proposed building hundreds of them in a factory to put them on "trailers" (no user serviceable parts inside) so that each small town/large village can have one. Considering that nuclear fuel (and its by-products) can be very poisonous, you can imagine that not many small towns or large villages would want one of them.
«and building large-scale nuclear power plants is an expensive, time-consuming, complex process compared to other sources of energy.»
Not significantly more, intrinsically, as compared to say an oil refinery, which are complex, dangerous, large plants, but for the fuel being potentially so poisonous. Perhaps regulations are a bit too strict, but since we don't really know how bad things could become, the precautionary principle is being followed. A fire at a refinery or a dyke failing are pretty bad outcomes, but we can estimate upper bounds a lot more easily than when nuclear fuel is involved.
It is very sad that nuclear power advocates are fixated on nuclear fuels and designs that are so particularly poisonous, for both historical and bad reasons (usually related to which fuels are optimal for military uses).
«why China [...] instead of really pushing the limits in stuff like mass nuclear energy»
Because they are not mad: they are already suffering a lot from coal pollution, and coal is far less poisonous than nuclear fuel. The biggest problem is that for nuclear power to work well and be acceptable *everything must be done just right* (and France and south Korea for now still can afford to do that), while the potential amount and area of damage for a poorly managed coal or oil or gas plant, or even a hydroelectric dam (except possibly for Assuan or Three Gorges), are far smaller.
Conceivably there are nuclear fuel cycles that use less poisonous fuels than uranium or plutonium (an important detail is the half-life of the fuel or its by-products), but there is nothing obviously viable yet.
«Didn't you once write a post that said, energy is only a small part of the world economy»
My estimate is that the benefits of oil are around 90-95% of "western" GDP (remove oil and 90-95% of GDP disappears), and our "western" high standards of living are entirely dependent on those benefits costing just 3-5% of "western" GDP.
«and so there isn't that much to gain from cheap energy?»
My impression is that our blogger is so blinded by the "transactional" (buy/sell) approach of neoclassical Economics that he cannot easily see the political economy as a physical system.
My view is that the "regulation" argument is completely wrong for two reasons:
* There are unregulated places in the world and they don't have flying cars either (no flying cars in Russia...).
* To have flying cars you don't need cheap electricity, but *very* cheap fuels that are also very energy dense, and that are easily fractionable.
The latter point is the critical one: there have been roughly two waves of improvement in energy usage, with coal and oil. Each wave had two sides: efficiency improvements and wider adoption. In each wave energy available eventually reached a plateau, where efficiency improvement become much smaller and the fuel has been adopted nearly everywhere.
What happened in the 1970s in the USA and most "developed" nations is just that oil as a fuel pretty much reached a plateau, just like coal had reached it before it, and no new fuel cheaper and more energy dense and more easily fractionable than oil has been found, unlike when coal was largely replaced by oil. Even worse, the cheapest oilfields have been used up, and extracting oil is becoming constantly more expensive.
Nuclear fuel is arguably cheaper and more energy dense than oil, but it has two huge problems:
* It is far, far more poisonous than coal or oil, which are pretty poisonous too (see deadly pollution during the coal and oil ages).
* It cannot be realistically used in small power plants, because it is not easily fractionable, and it can mostly be used to generate electricity, which is not a fuel, but a transmission medium, and that is very difficult to store cheaply and densely (in terms of weight).
Basically oil/petrol is not cheap enough and energy dense enough for flying cars, and nothing else is. It is perhaps cheap enough and dense enough for "delta wing" microlights, but those are the aerial equivalent of mopeds.
PS: there another big issue with flying cars: a "proper" flying car does not require a runway, it must be able to take off and land vertically, and that is really hard, see helicopters.
" The crazy regulation that killed nuclear energy is quite unjustified, not only because according to standard estimates nuclear causes far fewer deaths"
If the Teller-Ulam design is truly scalable without additional fissionable material beyond a small baseline (I believe it is still classified whether it is?), then almost any significant amount of proliferation is a bigger risk than anything else we know about (climate change, etc.).
Flying is hard because it's harder to keep track of things in three dimensions. Avoiding midair collisions in crowded skies is a much harder problem for humans than avoiding car crashes in heavy traffic, which is why airports need air traffic controllers but highways don't.
Is there any way to get a bound paper copy of that book? It seems to be Kindle only. I want to get my father a copy but he won't read it unless it's on paper.
I suppose I shouldn't say "ridiculous". My opinion though is that flying is very simple relative to driving, or even just walking down the street, because it doesn't require complex reflexes (tens of DOF for control at most) or much awareness (am I about to hit the ground, Y/N?). The skill seems complicated to humans because of unfamiliarity.
It's like Morevac's paradox: Chess is easy, lightning calculation is easy, things a two year old can do are hard.
One part of the desire behind the drive for self-driving cars is just the desire for the convenience. But another part is the desire to take control away from other people. Same with the "we need AI before we'll allow mass-personal-flying machines". Same with the regulating nuclear power to death. Will we, as a civilization, allow our neighbors to operate complex and possibly dangerous devices? If no, then the alternative is never hyper-competent New Soviet Man operating these devices, or some infallible magic robot operating these devices (we always discover that the infallible robot is far simpler and dumber than a notional operator): It's these devices never being built at all.
The traffic problem for 3d flight is much less fraught than the traffic problem for cars crammed onto roadways in the city. There is vastly more space in which you can arrange things like heading-altitude lanes. You just need to be aware of where everyone else is - there are no stop signs, deer crossings, mis-painted road markers, trucks shedding bricks, bicyclists, etc.
PS: I bought the book. Still in opening chapters but it's fascinating so far.
We don't need machine intelligence to run a flying car: That's ridiculous. Simple nonlinear controllers (linear PID controllers near equilibrium) can keep any number of rotors, thrusters, control surfaces and weight distributions balanced.One of the reasons why we have autopilot for jetliners and don't for cars is that keeping any sort of air vehicle in straight and level flight is dirt-simple. Understanding the world at ground level may well require general intelligence.
The reason we have cars and don't have flying cars is because no one needed permission to sell you a car back when they were being invented and the industry was born: It was a fait accompli before anyone thought to come up with a federal car administration.
It's pretty common to see people running down their fellow human's driving ability: And yet 90% of humanity managed to learn the skill, and we still don't have even the palest imitation of that depth of awareness in software. Flying isn't more complicated, it's just different, and people could have learned the skill at just as high a rate.
What makes you think Russia has no regulation?
«In each wave energy available eventually reached a plateau»
In the graph above the "Henry Adams curve" seems to me an artifact of optics: first what matters is energy *per person*, else population growth adds noise, and then I see three different parts of the graph: the before-coal era, reached a plateau, the coal era where it reached a plateau in the interwar period, and then the post-coal, oil era where it also reached a plateau.
Looking at the same graphs for England would be more useful, because on a wider scale diffusion adds quite a bit of noise.
«The main reason nuclear power has tapered off as a power supply is because it heavily favors scale and size»
Not really in technical terms: most large nuclear power plant sites actually have several smaller reactors rather than a single huge one. The cost advantage of large reactors are fairly small. Indeed at some point RR, that manufactures a rather tiny number of "small" nuclear power plants for submarines, proposed building hundreds of them in a factory to put them on "trailers" (no user serviceable parts inside) so that each small town/large village can have one. Considering that nuclear fuel (and its by-products) can be very poisonous, you can imagine that not many small towns or large villages would want one of them.
«and building large-scale nuclear power plants is an expensive, time-consuming, complex process compared to other sources of energy.»
Not significantly more, intrinsically, as compared to say an oil refinery, which are complex, dangerous, large plants, but for the fuel being potentially so poisonous. Perhaps regulations are a bit too strict, but since we don't really know how bad things could become, the precautionary principle is being followed. A fire at a refinery or a dyke failing are pretty bad outcomes, but we can estimate upper bounds a lot more easily than when nuclear fuel is involved.
It is very sad that nuclear power advocates are fixated on nuclear fuels and designs that are so particularly poisonous, for both historical and bad reasons (usually related to which fuels are optimal for military uses).
«why China [...] instead of really pushing the limits in stuff like mass nuclear energy»
Because they are not mad: they are already suffering a lot from coal pollution, and coal is far less poisonous than nuclear fuel. The biggest problem is that for nuclear power to work well and be acceptable *everything must be done just right* (and France and south Korea for now still can afford to do that), while the potential amount and area of damage for a poorly managed coal or oil or gas plant, or even a hydroelectric dam (except possibly for Assuan or Three Gorges), are far smaller.
Conceivably there are nuclear fuel cycles that use less poisonous fuels than uranium or plutonium (an important detail is the half-life of the fuel or its by-products), but there is nothing obviously viable yet.
«Didn't you once write a post that said, energy is only a small part of the world economy»
My estimate is that the benefits of oil are around 90-95% of "western" GDP (remove oil and 90-95% of GDP disappears), and our "western" high standards of living are entirely dependent on those benefits costing just 3-5% of "western" GDP.
«and so there isn't that much to gain from cheap energy?»
My impression is that our blogger is so blinded by the "transactional" (buy/sell) approach of neoclassical Economics that he cannot easily see the political economy as a physical system.
Consider instead:
https://voxeu.org/article/i...https://royalsocietypublish...https://www.cambridge.org/c...https://oll.libertyfund.org...
My view is that the "regulation" argument is completely wrong for two reasons:
* There are unregulated places in the world and they don't have flying cars either (no flying cars in Russia...).
* To have flying cars you don't need cheap electricity, but *very* cheap fuels that are also very energy dense, and that are easily fractionable.
The latter point is the critical one: there have been roughly two waves of improvement in energy usage, with coal and oil. Each wave had two sides: efficiency improvements and wider adoption. In each wave energy available eventually reached a plateau, where efficiency improvement become much smaller and the fuel has been adopted nearly everywhere.
What happened in the 1970s in the USA and most "developed" nations is just that oil as a fuel pretty much reached a plateau, just like coal had reached it before it, and no new fuel cheaper and more energy dense and more easily fractionable than oil has been found, unlike when coal was largely replaced by oil. Even worse, the cheapest oilfields have been used up, and extracting oil is becoming constantly more expensive.
Nuclear fuel is arguably cheaper and more energy dense than oil, but it has two huge problems:
* It is far, far more poisonous than coal or oil, which are pretty poisonous too (see deadly pollution during the coal and oil ages).
* It cannot be realistically used in small power plants, because it is not easily fractionable, and it can mostly be used to generate electricity, which is not a fuel, but a transmission medium, and that is very difficult to store cheaply and densely (in terms of weight).
Basically oil/petrol is not cheap enough and energy dense enough for flying cars, and nothing else is. It is perhaps cheap enough and dense enough for "delta wing" microlights, but those are the aerial equivalent of mopeds.
PS: there another big issue with flying cars: a "proper" flying car does not require a runway, it must be able to take off and land vertically, and that is really hard, see helicopters.
Didn't you once write a post that said, energy is only a small part of the world economy and so there isn't that much to gain from cheap energy?
https://twitter.com/robinha...
" The crazy regulation that killed nuclear energy is quite unjustified, not only because according to standard estimates nuclear causes far fewer deaths"
If the Teller-Ulam design is truly scalable without additional fissionable material beyond a small baseline (I believe it is still classified whether it is?), then almost any significant amount of proliferation is a bigger risk than anything else we know about (climate change, etc.).
"If car companies innovated at the same rate as computer companies, we'd be driving cars that cost $25 and got 1000 miles per gallon."
"Yes, but would you really want your car to crash twice a day?"
Flying is hard because it's harder to keep track of things in three dimensions. Avoiding midair collisions in crowded skies is a much harder problem for humans than avoiding car crashes in heavy traffic, which is why airports need air traffic controllers but highways don't.
Is there any way to get a bound paper copy of that book? It seems to be Kindle only. I want to get my father a copy but he won't read it unless it's on paper.
Chinese culture has always been extremely conservative, conformist, and anti-change, or so I hear...
I suppose I shouldn't say "ridiculous". My opinion though is that flying is very simple relative to driving, or even just walking down the street, because it doesn't require complex reflexes (tens of DOF for control at most) or much awareness (am I about to hit the ground, Y/N?). The skill seems complicated to humans because of unfamiliarity.
It's like Morevac's paradox: Chess is easy, lightning calculation is easy, things a two year old can do are hard.
One part of the desire behind the drive for self-driving cars is just the desire for the convenience. But another part is the desire to take control away from other people. Same with the "we need AI before we'll allow mass-personal-flying machines". Same with the regulating nuclear power to death. Will we, as a civilization, allow our neighbors to operate complex and possibly dangerous devices? If no, then the alternative is never hyper-competent New Soviet Man operating these devices, or some infallible magic robot operating these devices (we always discover that the infallible robot is far simpler and dumber than a notional operator): It's these devices never being built at all.
The traffic problem for 3d flight is much less fraught than the traffic problem for cars crammed onto roadways in the city. There is vastly more space in which you can arrange things like heading-altitude lanes. You just need to be aware of where everyone else is - there are no stop signs, deer crossings, mis-painted road markers, trucks shedding bricks, bicyclists, etc.
PS: I bought the book. Still in opening chapters but it's fascinating so far.
We don't need machine intelligence to run a flying car: That's ridiculous. Simple nonlinear controllers (linear PID controllers near equilibrium) can keep any number of rotors, thrusters, control surfaces and weight distributions balanced.One of the reasons why we have autopilot for jetliners and don't for cars is that keeping any sort of air vehicle in straight and level flight is dirt-simple. Understanding the world at ground level may well require general intelligence.
The reason we have cars and don't have flying cars is because no one needed permission to sell you a car back when they were being invented and the industry was born: It was a fait accompli before anyone thought to come up with a federal car administration.
It's pretty common to see people running down their fellow human's driving ability: And yet 90% of humanity managed to learn the skill, and we still don't have even the palest imitation of that depth of awareness in software. Flying isn't more complicated, it's just different, and people could have learned the skill at just as high a rate.