5 ms·

One interesting tidbit toward the end is that the author thinks building this ship is feasible within 200 years of the writing, with a cost estimate of $10^11 i

One interesting tidbit toward the end is that the author thinks building this ship is feasible within 200 years of the writing, with a cost estimate of $10^11 in 1968 dollars, assuming a 4% GNP growth rate. I haven't calculated what $10^11 would become in 2021 dollars, but growth hasn't been that high since the 90s, and it wasn't sustained growth, even then.

TrainedMonkey 5y ago
He gives a comparison estimate that each Saturn V cost about 10^8 each. Per quick Google search actual cost in 1970s dollars was $185m which is $1.15b today: https://en.wikipedia.org/wiki/Saturn_V#Cost https://en.wikipedia.org/wiki/Saturn_V#Cost (important, but unrelated to rest of computation fact is that this cost was amortized over 35 flights) 10^11 is 1000 * 10^8, applying same inflation to 10^11 we get $1.15t. To put this number in a context, Iraq war is estimated to have long term cost of $1.1tn to $1.92tn: https://en.wikipedia.org/wiki/Financial_cost_of_the_Iraq_War https://en.wikipedia.org/wiki/Financial_cost_of_the_Iraq_War
- m4rtink 5y ago
  If SpaceX Starship can get going as planned (eq. fully & rapidly reusable orbital transport) it should be possible to slash the costs of launch for big proposals like this quite significantly.
  - andyxor 5y ago
    exactly, all these old estimates need to be revised in SpaceX era. We are living in unprecedented times when a private company does something that was previously only possible with global superpower resources.
  - maxerickson 5y ago
    The proposal is a 5 million ton, 10 km shell of copper fueled by 30 million hydrogen bombs. Cost to orbit isn't particularly relevant (it would easily be optimized as part of the decades long project).
    - nomoreplease 5y ago
      I’ve heard of nuclear bomb propulsion before but I had no idea that 30 million hydrogen bombs we’re proposed. That’s insane!
      - m4rtink 5y ago
        And do you know what's more insane (and even more cool!) - Nuclear Salt Water Rocket! https://en.wikipedia.org/wiki/Nuclear_salt-water_rocket https://en.wikipedia.org/wiki/Nuclear_salt-water_rocket Riding a continuous nuclear explosion for fun and profit! :D
MichaelZuo 5y ago
For a government project I would at least double the numbers.
- 8bitsrule 5y ago
  Reminds me of a project I heard about yesterday. Decades ago, a bridge for a private railway company was built in such a way that it could also carry an interurban rail and be an automobile highway. After decades, only the railroad portion had ever been used. I suspect that was the intended outcome, and I suspect that the project was, in part at least, taxpayer-funded. From history, I also suspect that many such projects serve as a way to get public dollars into -the right- private pockets. Sometimes they're sold as 'job creation' programs. Uh-huh.
credit_guy 5y ago
The way the GDP and the GDP growth are calculated is a bit funny. In that if you start with one year and compound the official GDP growth for a number of years, you end up with a number that is much lower than the actual GDP at the end of that period. In this particular case, Dyson mentions a GNP of about $600 BN. Since I don't know exactly the difference between the GDP and the GNP, I'll use the official GDP as recorded by the Fed [1]. It was $968 BN in 1968 and $21.5 TN in 2020. That's a (compounded) growth rate of more than 6% per year. [1] https://fred.stlouisfed.org/series/GDP https://fred.stlouisfed.org/series/GDP
- dragonwriter 5y ago
  The data series you are looking at is nominal, not real, GDP [0], so the annual growth multiplier (1 + growth rate) is the real GDP growth multiplier times the inflation multiplier. In current dollars, 1968’s real GDP is something over $4 trillion. [0] The St. Louis Fed has both series, see https://fred.stlouisfed.org/series/GDPC1 https://fred.stlouisfed.org/series/GDPC1

TrainedMonkey 5y ago

He gives a comparison estimate that each Saturn V cost about 10^8 each. Per quick Google search actual cost in 1970s dollars was $185m which is $1.15b today: https://en.wikipedia.org/wiki/Saturn_V#Cost https://en.wikipedia.org/wiki/Saturn_V#Cost (important, but unrelated to rest of computation fact is that this cost was amortized over 35 flights) 10^11 is 1000 * 10^8, applying same inflation to 10^11 we get $1.15t. To put this number in a context, Iraq war is estimated to have long term cost of $1.1tn to $1.92tn: https://en.wikipedia.org/wiki/Financial_cost_of_the_Iraq_War https://en.wikipedia.org/wiki/Financial_cost_of_the_Iraq_War

m4rtink 5y ago
If SpaceX Starship can get going as planned (eq. fully & rapidly reusable orbital transport) it should be possible to slash the costs of launch for big proposals like this quite significantly.
- andyxor 5y ago
  exactly, all these old estimates need to be revised in SpaceX era. We are living in unprecedented times when a private company does something that was previously only possible with global superpower resources.
- maxerickson 5y ago
  The proposal is a 5 million ton, 10 km shell of copper fueled by 30 million hydrogen bombs. Cost to orbit isn't particularly relevant (it would easily be optimized as part of the decades long project).
  - nomoreplease 5y ago
    I’ve heard of nuclear bomb propulsion before but I had no idea that 30 million hydrogen bombs we’re proposed. That’s insane!
    - m4rtink 5y ago
      And do you know what's more insane (and even more cool!) - Nuclear Salt Water Rocket! https://en.wikipedia.org/wiki/Nuclear_salt-water_rocket https://en.wikipedia.org/wiki/Nuclear_salt-water_rocket Riding a continuous nuclear explosion for fun and profit! :D

MichaelZuo 5y ago

For a government project I would at least double the numbers.

8bitsrule 5y ago
Reminds me of a project I heard about yesterday. Decades ago, a bridge for a private railway company was built in such a way that it could also carry an interurban rail and be an automobile highway. After decades, only the railroad portion had ever been used. I suspect that was the intended outcome, and I suspect that the project was, in part at least, taxpayer-funded. From history, I also suspect that many such projects serve as a way to get public dollars into -the right- private pockets. Sometimes they're sold as 'job creation' programs. Uh-huh.

credit_guy 5y ago

The way the GDP and the GDP growth are calculated is a bit funny. In that if you start with one year and compound the official GDP growth for a number of years, you end up with a number that is much lower than the actual GDP at the end of that period. In this particular case, Dyson mentions a GNP of about $600 BN. Since I don't know exactly the difference between the GDP and the GNP, I'll use the official GDP as recorded by the Fed [1]. It was $968 BN in 1968 and $21.5 TN in 2020. That's a (compounded) growth rate of more than 6% per year. [1] https://fred.stlouisfed.org/series/GDP https://fred.stlouisfed.org/series/GDP

dragonwriter 5y ago
The data series you are looking at is nominal, not real, GDP [0], so the annual growth multiplier (1 + growth rate) is the real GDP growth multiplier times the inflation multiplier. In current dollars, 1968’s real GDP is something over $4 trillion. [0] The St. Louis Fed has both series, see https://fred.stlouisfed.org/series/GDPC1 https://fred.stlouisfed.org/series/GDPC1