Tuesday, January 7, 2020
Sunday, January 5, 2020
Terraforming Venus; Venus-forming Earth
Terraforming Venus would require taking almost all of the CO2 out of its atmosphere, so it becomes breathable, doesn't crush us (currently 90 atm pressure) and cools the planet down. Using our current mechanical carbon scrubber technology may seem simplistic and unimaginative, but the other options that have been discussed feature similar science-fiction-level ideas (crashing outer solar system ice moons into it, locking carbon into the crust down to a kilometer deep, or getting theoretically present hydrogen out of the mantle.)
This is less likely to happen than being able to move moons around the Solar System.
Image from reddit.com/r/mapporn.
Let's make many optimistic assumptions:
That we can build self-replicating independent carbon sequestration plants; this minimizes transport costs and covers the planet.
That they can build and fuel themselves from materials available on the surface of Venus.
That they can withstand conditions on Venus (when the longest any machine we've put down has lasted is on the order of an hour.)
Current carbon sequestration plants are the size of a cargo container, and sequester 900 tons of carbon per year. Assume that this is the rate at which they operate on Venus, and that self-replicating carbon sequesterers are 100x bigger than the real, non-self-replicating ones we have.
Assuming near 100% working replicas, and a one-year self-replication cycle, it would take 40 years to cover the entire surface of Venus with these - after which they would take 2000 years to clean the atmosphere of CO2. (This would still leave a nitrogen atmosphere several times higher pressure than Earth's.)
Venus is not the best candidate for terraforming or habitation, and humans will not settle its surface for thousands of years at least. We should concentrate on terraforming planets in our solar system, building self-replicating technologies, and having humans in isolation from Earth in case of some sort of collapse (most easily, on the Moon.)
On the other hand, here on Earth, just to keep even with carbon emissions at the 2017 level, we would need 40 million of the scrubbers we currently have. That means no matter where you went on Earth, there would be one within less than two and a half miles of you.
We do have machines that are Venus-forming Earth, by making more CO2. They aren't self-replicating, but they seem to have a relationship with one species (unclear if parasitic or symbiotic) and in places they cover the surface just the same.
This is less likely to happen than being able to move moons around the Solar System.
Image from reddit.com/r/mapporn.
Let's make many optimistic assumptions:
That we can build self-replicating independent carbon sequestration plants; this minimizes transport costs and covers the planet.
That they can build and fuel themselves from materials available on the surface of Venus.
That they can withstand conditions on Venus (when the longest any machine we've put down has lasted is on the order of an hour.)
Current carbon sequestration plants are the size of a cargo container, and sequester 900 tons of carbon per year. Assume that this is the rate at which they operate on Venus, and that self-replicating carbon sequesterers are 100x bigger than the real, non-self-replicating ones we have.
Assuming near 100% working replicas, and a one-year self-replication cycle, it would take 40 years to cover the entire surface of Venus with these - after which they would take 2000 years to clean the atmosphere of CO2. (This would still leave a nitrogen atmosphere several times higher pressure than Earth's.)
Venus is not the best candidate for terraforming or habitation, and humans will not settle its surface for thousands of years at least. We should concentrate on terraforming planets in our solar system, building self-replicating technologies, and having humans in isolation from Earth in case of some sort of collapse (most easily, on the Moon.)
On the other hand, here on Earth, just to keep even with carbon emissions at the 2017 level, we would need 40 million of the scrubbers we currently have. That means no matter where you went on Earth, there would be one within less than two and a half miles of you.
We do have machines that are Venus-forming Earth, by making more CO2. They aren't self-replicating, but they seem to have a relationship with one species (unclear if parasitic or symbiotic) and in places they cover the surface just the same. Star Wars Episode IX Review (CONTAINS SPOILERS BUT MAY NOT MATTER SINCE I'M THE LAST ONE TO SEE IT)
It's the best of the three new ones, and I think the second best in the entire franchise after Empire. I'm lazy, so some of these things would be easy to look up but I didn't. In no particular order:
- Overall, a fitting end to the saga. Much better than Episode VIII. Glad JJ Abrams is back; he's the master of playing well in others' universes. You walk out feeling that you saw a good Star Wars movie, much like Episode VII.
- As a Star Wars movie, at some point there will be: a desert planet, a cantina, a scummy trading city with a motley assortment of criminals, and people rescuing their friends and admitting how zany they are because they’re making it up as they go.
. But it's not dwelt upon, and there were so many types of planets in this one that the desert planet doesn't stick out. Also, I hope I'm not turning into a social justice warrior, but the colorful festival on the planet with "third world" aliens did bug me a little bit, just because it seems so obvious.
- In that spirit: Star Wars movies are fantasy movies in space: knights, sword-fighting, wizards. A horseback attack on a starcruiser is therefore not only cool and original, it completely fits. Also, it stuck out in Episodes I through III that a) there was no improvement in technology across 30 years and b) that late 90s cell phones were about as good as Obi Wan's communicator. Some of the tech in this one looks positively 1980s - wired headsets, colored wires in C3PO's head that look like a Commodore 64 console - and that's great. It fits Star Wars perfectly. It's a fantasy movie, and these are set pieces to maintain the tone - it's not about technology at all.
- I noticed in the end credits for animation, there was a group of 8-10 names together that looked Thai or Lao. I tried to look up whether a studio in one of those places was used but couldn't find any mention. Always good to see more talent in the game, it benefits film-goers.
- I couldn't remember if Carrie Fisher died before they made this (she did), and couldn't tell if she was live, computer-generated, etc., and wasn't particularly worried about it either way. It's that last part that says the most I think.
- The introduction of matter transmission through force connections (and its continuing use) is quite a good cinematic trick. It's creepy as hell when Ren grabs Rey's necklace, and then this trick is used in the final battle with the Emperor (unlike some inventions in Episode VIII like hyperspace missiles.)
- The action starts immediately at the beginning and doesn't let up for quite a while. It's also full of plot twists. Nothing like the (pointless, non-plot-advancing) dead space in Episode VIII.
- Plot holes, discontinuities, and other jarring moments must be taken in context with the franchise. I'm a fan of hard science fiction (i.e., real physics) to the core but if I ever directly overhear someone complaining that "it's unrealistic because there's hyperspace" I will punch them. That said, no one came to help in the second one, but Chewie and Lando can magically raise a rebellion in a few hours by flying around in person? Doesn't seem required to advance the plot, though it does feel good and advance a nice moral ("they win by making you feel alone") so it didn't bother me so much. Also, where has Lando been?
- I'm going to come out and say it. Grown-up Daisy Ridley is much, much more attractive than at the start of this trilogy (and she wasn't bad then either.) She's also come into her own as an actor.
- Good performances also from Adam Driver (duh) and Oscar Isaac. I always wonder how hard it is for Boyega to stay in-accent when he's working with other English actors. As time goes on, he gets by more on his innate charisma, which is fine because he's just likable anyway (honestly, it's the very rare actor who actually masters another character and doesn't just enhance his innate personality. I like watching Robert Downey Jr. and Anthony Hopkins, but come on, how different are they on-screen from their actual personalities? Gary Oldman and Christian Bale actually get outside their personalities for their characters, which carries a penalty of not being as "follow-able" as they might otherwise be.)
- Lots of nice parallelisms in this and tying up of themes. The Empreror as master manipulator tries to rise above the paradox of the Sith by having his death be the key to the rise of the Sith (and if Rey doesn't do it, her friends die and the fleet takes over anyway.) The balance brought back to the Force requires the dyad, with Ren and Rey. Ren, again Ben with the same words he said when he killed his father. Leia's death as she kills her son by distracting him (Ren killed Solo where Vader couldn't bring himself to kill Luke, twice.)
- Kylo Ren has to announce that Han Solo is a memory so we don't get confused. Luke has sparkly blue outlines so we know the difference between a memory and a Force ghost.
- There's a little bit of babbling of the sort that annoyed Alec Guinness, which annoys me mostly because you don't know where things are going and you can talk your way (in an unsatisfyingly unpredictable way) out of any plot knot. Then again, the Force is if nothing else a script writer's dream to resolve what would otherwise be massive plot holes by appealing effectively to magic.
- Much better psychology in this one, thinking about the characters' motivations (e.g., Hux being a spy because he hates Ren.)
- At the end where there's a crowd watching Palpatine and Rey: where did all these mystery Sith come from? And why have we never seen other Knights of Ren before? What is their connection to the Sith?
- The creator of Darth Maul said that Lucas came to him and said "I want you to use your worst childhood nightmare as inspiration." He described to Lucas a pale face slowly revealed by flashes of lightning. Sound familiar? I was happy that they used this.
- I didn't notice any lens flare, but there is a shot of Kylo Ren standing in front of the heaving ocean after Rey leaves, a bit fuzzy and shaky as if shot from far away, that gives it more immediacy. Adam Driver also looked genuinely cold, and I wouldn't be surprised to learn that he insisted on being cold and wet in reality for these shots.
- Rey being a Palpatine is somewhat of an obvious plot twist, just because if she IS the daughter of someone significant, the only decent move left is for it to be Palpatine (with intervening parents; why did they turn out so lame?)
- I was annoyed that a non-Jedi (Leia) could magically continue Rey's training. ("Hey, I'm a surgery resident. My attending just died. His sister is going to come continue to teach me surgery.") But they closed that loophole.
- Once Rey explained she gave a little bit of the force away to heal the giant snake critter, you could see a self-sacrifice (from someone) coming down Fifth Avenue in a cab. Same with the accidental force lightning, which immediately gave away what I suspected about Rey being a Palpatine. Good on the writers, I actually thought Chewie was dead. I'm sure that there are religious fundies somewhere all bent out of shape that Jedi can heal people.
Above: Jedi Christ, who can heal the sick; only Sith can raise the dead. I bet Golgotha would have been a whole lot different with a light saber.
- Good coordination with trailers and writers. They know going in that everyone knows Palpatine is in it. I'm always amazed when trailers give away things in the movie and the film-makers expect you to be surprised (often this is no doubt marketing that's not in their control.)
- I thought C3PO was added to the list of main character deaths - a robot is its memory in a way that humans are not - but he was saved too.
Saturday, January 4, 2020
Timeline of Manned Interstellar Travel, Based on Simple Economics: No Humans on Alpha Centauri Planets Until 2613
It has been estimated that a manned Mars mission would cost $100 billion. Compare this to the most recent unmanned lander, Insight, at $830 million; putting people on Mars then comes with a cost multiplier of 120.
The Initiative for Interstellar Studies estimates that an unmanned interstellar mission would cost at least "in the trillions"; Centauri Dreams cites Odenwald at $174 trillion. Assuming the same scaling, the lower and upper bounds on that then suggest that a manned mission would cost from $240 trillion to $21 quadrillion.
If on the other hand we take the projected cost of a manned mission to Mars, and assume it scales linearly with distance, a manned Mission to Alpha Centauri would cost $55 quadrillion.
It's worth pointing out here that world GDP is $80 trillion. Let's assume an annual economic growth rate over time of 2%. Let's also assume that starting tomorrow we put ALL of GDP toward such a mission - that is, every last human is working this mission and just barely otherwise just barely surviving as peasants eating crumbs.
Assuming an annual economic growth rate over time of 2%, then at earliest, we can launch a manned interstellar mission at the earliest by 2227; at latest, by 2501.
But forget about that. Because neither you, nor any other human on this planet will sign up tomorrow for their descendants being reduced to slavery for centuries for a space mission, which is what those numbers assume. So let's assume we continue to spend money on space exploration at the same rate that we in the US currently are - about 0.11% of GDP. This is already quite a generous assumption, given that most countries can't afford to dedicate such a fraction of wealth to endeavors that don't quickly return on investment. If you're more optimistic and want to set the relative rate of expenditure (over centuries) to the highest it has ever been (in a democracy - you said you were optimistic right?) that's 1966 USA, which is about twice what it is today, and only makes it happen 35 years earlier. (This is more dependent on economic growth than space program expenditure.) So let's stick with current NASA budget fraction, and assume that the future space program is ONLY working on this one mission.
By these assumptions, we can launch the mission at earliest by 2570; for the upper bound estimate, by 2845.
Our fastest spacecraft so far would take another 30,000 years after launch to get there. Let's be more optimistic and assume that the light sail technology we're talking about for unmanned probes also applies to manned craft, and can get the ship up to 10% of the speed of light. Therefore, taking into account travel time and speed-of-light delays, we wiill get the interstellar "Eagle has landed message" at an absolute cheapest earliest date of 2618.
Of course this is still unrealistic, because we're still assuming mission development starts in earnest tomorrow, assuming every government on Earth will let us use a NASA-sized fraction of their GDP for this, and that they will continue to cooperate for at least 550 years building the mission. Think of this in reverse: it's as if in 1470, the middle of the War of the Roses, and the Russians and Poles and Lithuanians still throwing off the Mongol yolk, everyone started spending money and cooperating on a project and continued to cooperate on it until this year.
I think it is unlikely, barring unforeseeable scientific revolutions, that human beings will leave the Solar System this millennium. I think it is likely that there will be civilization or species-threatening or destroying events in this millennium. This discussion of colonizing other planets to mitigate existential risks has a scatter plot listing a probability of event happening within 200 years/risk of civilizational collapse for nuclear war, coronal mass event, rogue AI, and nuclear war as 90%/20%, 70%/90%, and 95%/70%.
Using those same numbers, in the time period until launch there's a greater than a 96.6% chance of a rogue AI, and a greater than 99% chance of coronal mass event or nuclear war.
But fully automated probes could get out more quickly, particularly if we design self-reproducing von Neumann probes. We should start terraforming Mars now, as practice for remotely terraforming planets with von Neumann probes for when we eventually get there. We have time to terraform them, because if physical human bodies ever do get there, it will be in the distant future. But we do not have that much time to get the launch the hardware, which suggests we should at least colonize the Moon as insurance. Cryonics and hibernation technology at this point is still basically science fiction. These numbers are depressing given our previous dreams, but we calibrated on going from powered flight to standing on the moon in 2/3 of a century.
The Initiative for Interstellar Studies estimates that an unmanned interstellar mission would cost at least "in the trillions"; Centauri Dreams cites Odenwald at $174 trillion. Assuming the same scaling, the lower and upper bounds on that then suggest that a manned mission would cost from $240 trillion to $21 quadrillion.
If on the other hand we take the projected cost of a manned mission to Mars, and assume it scales linearly with distance, a manned Mission to Alpha Centauri would cost $55 quadrillion.
It's worth pointing out here that world GDP is $80 trillion. Let's assume an annual economic growth rate over time of 2%. Let's also assume that starting tomorrow we put ALL of GDP toward such a mission - that is, every last human is working this mission and just barely otherwise just barely surviving as peasants eating crumbs.
Assuming an annual economic growth rate over time of 2%, then at earliest, we can launch a manned interstellar mission at the earliest by 2227; at latest, by 2501.
But forget about that. Because neither you, nor any other human on this planet will sign up tomorrow for their descendants being reduced to slavery for centuries for a space mission, which is what those numbers assume. So let's assume we continue to spend money on space exploration at the same rate that we in the US currently are - about 0.11% of GDP. This is already quite a generous assumption, given that most countries can't afford to dedicate such a fraction of wealth to endeavors that don't quickly return on investment. If you're more optimistic and want to set the relative rate of expenditure (over centuries) to the highest it has ever been (in a democracy - you said you were optimistic right?) that's 1966 USA, which is about twice what it is today, and only makes it happen 35 years earlier. (This is more dependent on economic growth than space program expenditure.) So let's stick with current NASA budget fraction, and assume that the future space program is ONLY working on this one mission.
By these assumptions, we can launch the mission at earliest by 2570; for the upper bound estimate, by 2845.
Our fastest spacecraft so far would take another 30,000 years after launch to get there. Let's be more optimistic and assume that the light sail technology we're talking about for unmanned probes also applies to manned craft, and can get the ship up to 10% of the speed of light. Therefore, taking into account travel time and speed-of-light delays, we wiill get the interstellar "Eagle has landed message" at an absolute cheapest earliest date of 2618.
Of course this is still unrealistic, because we're still assuming mission development starts in earnest tomorrow, assuming every government on Earth will let us use a NASA-sized fraction of their GDP for this, and that they will continue to cooperate for at least 550 years building the mission. Think of this in reverse: it's as if in 1470, the middle of the War of the Roses, and the Russians and Poles and Lithuanians still throwing off the Mongol yolk, everyone started spending money and cooperating on a project and continued to cooperate on it until this year.
I think it is unlikely, barring unforeseeable scientific revolutions, that human beings will leave the Solar System this millennium. I think it is likely that there will be civilization or species-threatening or destroying events in this millennium. This discussion of colonizing other planets to mitigate existential risks has a scatter plot listing a probability of event happening within 200 years/risk of civilizational collapse for nuclear war, coronal mass event, rogue AI, and nuclear war as 90%/20%, 70%/90%, and 95%/70%.
Using those same numbers, in the time period until launch there's a greater than a 96.6% chance of a rogue AI, and a greater than 99% chance of coronal mass event or nuclear war.
But fully automated probes could get out more quickly, particularly if we design self-reproducing von Neumann probes. We should start terraforming Mars now, as practice for remotely terraforming planets with von Neumann probes for when we eventually get there. We have time to terraform them, because if physical human bodies ever do get there, it will be in the distant future. But we do not have that much time to get the launch the hardware, which suggests we should at least colonize the Moon as insurance. Cryonics and hibernation technology at this point is still basically science fiction. These numbers are depressing given our previous dreams, but we calibrated on going from powered flight to standing on the moon in 2/3 of a century.