Exemplary Futurism

Back in May there was a Starship Century Symposium in San Diego. I didn’t attend, but I later watched videos of most of the talks (here, here, book coming here). Many were about attempts by engineers and scientists to sketch out feasible designs for functioning starships. They’ve been at this for many decades, and have made some progress.

Most of us have seen many starships depicted in movies, and you might figure that since the future is so uncertain, fictional starships are our best guess to real future starships. Since no one can know anything, no one can beat fictional imagination. But that seems very wrong to me; we get a lot better insight into real starships from serious attempts to design them.

It is worth noting that these folks do futurism the way I say it should be done: making best combos.

Form best estimates on each variable one at a time, and then adjust each best estimate to take into account the others, until one has a reasonably coherent baseline combination: a set of variable values that each seem reasonable given the others. (more)

Even though this is the standard approach of historians, schedulers, and puzzle solvers, many express strong disapproval about doing futurism this way. Well at least when predicting social consequences. Starship designers don’t seem to get much flack. Why? I’d guess it is because they are high status. Starships and physicists are sexy enough that we forget to be politically correct, and just let experts do what seems best to them.

Some might say this is okay for engineering, because we know lots of engineering, but not ok for social things, because we know little there. But that is just wrong. Not only do we know lots about social things, this is still the right approach in areas of history and engineering where we know a lot less.

It is also worth noting that the usual starship vision mainly seem interesting if one expects familiar growth rates to continue for a while. A starship carrying humans would take about a decade or two in flight time, and a thousand times as much energy as the Apollo moon rockets. And today our economy doubles in about 15 years. So if energy capacity doubled with the economy, it would take about 150 years to get that capacity. Or since energy has doubled about every 25 years lately, it might take 250 years. But 150-250 years still seems culturally accessible to us; we feel we can relate to folks 200 years ago. Much more at least than to people 20,000 years ago.

But if growth rates either slow down or sleep up a lot, this doesn’t work. For example, if the economy doubled every thousand years, as it did during the farming era, then it would take ten thousand years to get enough capacity. And we feel much less related to people who will live ten thousand years in the future. We expect their culture to change so much that we are much less interested in stories about them, or in thinking about what they will do.

If growth rates instead speed up by the same factor, the economy would then double every three months. And then a decade long flight to another star would encompass forty doublings, or a factor of a trillion. At growth rates like that, a journey that  long just seems crazy. Before you’ve hardly left our system another much better ship is likely to whiz past you. And even if your ship gets there first, the civilization back home by then is likely to be culturally unrecognizable. A trillion size bigger economy is likely to be a very different place.

Of course fast growth can’t go on forever. So a fast growing economy will slow down eventually. And that is when it would make sense to take a decade long journey, when a decade doesn’t encompass that much cultural change back home. But such a post-fast-growth society will likely be so different from ours as to deflate most of our interest in thinking about their starships.

If our industry era growth rates continue on for several centuries, then we may have descendants capable of starlight, and culturally similar enough to us that we care a lot about them. But if growth rates either slow down or speed up a lot, the descendants who are finally willing and able to fly to the stars are likely to be so different from us that we are much less interested in them.

GD Star Rating
loading...
Tagged as: ,
Trackback URL:
  • Shaun Griffith

    Nice discussion. But I think the time scale is wrong. While it takes a decade to get to Pluto, unless we invent super-luminal technology, the nearest star is hundreds or thousands of years away in anything we can imagine now for carrying people. Those two orders of magnitude need priority over the other in between.

    • Robert Koslover

      Yes, but we don’t need super-luminal capabilities to visit distant stars within the lifetimes of the travelers themselves. Just a few years at constant g acceleration yields speeds close enough to c that time-dilation helps considerably. You could potentially travel 1000 light-year distances or more. But, if you ever came back, you couldn’t tell your friends about the trip (unless they lived to thousands of years old). There’s an old song from the movie “Dark Star” that laments a somewhat shorter relativistic journey — see http://www.youtube.com/watch?v=PbbYn7d_W8s and note the line, “I’m young and now you’re old.”

      • VV

        Just a few years at constant g acceleration yields speeds close enough to c that time-dilation helps considerably.

        You can’t use Newtonian mechanics at relativistc speeds.

        The amount of energy you need to accelerate to a relativistic speed is linear in the time dilatation you get:

        To get a 2X time dilatation you need to travel at 0.87c. The energy you require to accelerate a 1000 kg (the mass of a small car) object to that speed is about 9 * 10^19 Joules, about the current electical energy generation of the world in one year.

        To get a 10X time dilatation you would need about 8 * 10^20 Joules (the estimated energy of the world total uranium resources).

        Moreover, once you get to your target star, you have to pay about the same energy in order to stop.

      • Robert Koslover

        Thank you for the comment. Please note that I didn’t suggest it would be an easy engineering challenge, nor did I endeavor to apply Newtonian mechanics. I suppose I should have mentioned that I am referring to accelerating at g, as experienced in the (at any snapshot in time) reference frame of the passengers on the moving ship (since that is something that the passengers could be fairly comfortable with, long-term). For the pre-relativistic part of the trip, that also corresponds to approximately g in Earth’s frame. Later, it corresponds to a lesser acceleration as viewed in the Earth’s frame, so that (of course) the spaceship can never reach c. And yes, you have to spend a great deal of time slowing down too, so that you don’t fly past your target. And then to come back, you have to do it all over again. I suggest that if you do the math, you will see that (if I recall correctly) all of this is still theoretically possible within one human lifetime (for a passenger only, not an outside observer), to cover 1000 light years or more. And yes, the engineering is far, far beyond our present abilities, the energy requirements are indeed astoundingly enormous, and there are many great technological challenges involved, such as dealing with the intense radiation that you encounter in “empty” space. So, no one is likely to be doing this any time soon! Best regards.

      • VV

        I suggest that if you do the math, you will see that (if I recall
        correctly) all of this is still theoretically possible within one human
        lifetime (for a passenger only, not an outside observer), to cover 1000
        light years or more.

        I suppose that if the spaceship has to carry at least enough energy to stop, this probably requires matter-antimatter annihilation.

        such as dealing with the intense radiation that you encounter in “empty” space.

        Also taking into account blueshift of EM radiation. Go fast enough and even the cosmic background radiation could exert enough pressure to slow you down or just fry you.

  • VV

    So if energy capacity doubled with the economy, it would take about 150 years to get that capacity. Or since energy has doubled about every 25 years lately, it might take 250 years.

    In order to reach that energy level you would have to plaster the entire dry land of earth with 20% efficient solar panels.

    http://physics.ucsd.edu/do-the-math/2011/07/galactic-scale-energy/

  • Jess Riedel

    > you might figure that since the future is so uncertain, fictional starships are our best guess to real future starships. Since no one can know anything, no one can beat fictional imagination. But that seems very wrong to me; we get a *lot better insight into real starships from serious attempts to design them.

    Who argues this? I’d guess you’re referring to the kind of folks who criticize detailed investigations of scenarios in the future with massive uncertainty, such as the criticism you’ve gotten for thinking about the em economy.

    I believe you are badly straw-manning the argument when you characterize it as claiming that “no one can gain insights from best-combo thinking”. The better argument, which I’ve advanced in a previous comment, compares predictions taken from your best-combo scenarios to the (necessarily more modest) predictions based only on hard constraints (e.g. the laws of physics). This better argument concludes that any specific prediction provided by a best-combo scenario has negligible chance of being realized (unless it can be derived only from hard constraints). Ultimately, this is based on the exponential behavior of chaining conditional probabilities together; if your scenario requires several uncertain things to happen, the chance that they all happen vanishes very quickly as you chain them together.

    The possible payoff of best-combo thinking is the discovery of previously-unconsidered hard constraints, or perhaps unlikely spin-off ideas for planning in the near future (when best-combo thinking is *not* necessarily doomed). But neither seem particularly likely to me. That doesn’t mean such thinking is useless, but it does mean you should acknowledge the negligible probabilities of anything you discuss. And it means that any justification for far-future best-combo thinking should involve appeals to its great importance to outweight the small chance of gaining useful insight.

    One other angle: have you tried finding a case where theorizers in the far-past (e.g., ancient greeks) used best-combo thinking to make predictions about today? Are you comfortable being compared to them? If not, what advantage over them do you have?

    • http://juridicalcoherence.blogspot.com/ Stephen Diamond

      Ultimately, this is based on the exponential behavior of chaining conditional probabilities together; if your scenario requires several uncertain things to happen, the chance that they all happen vanishes very quickly as you chain them together.

      Why doesn’t this rule out best combo in the situations where it’s useful? Don’t they typically involve chaining conditional probabilities?

      • Jess Riedel

        Events in the near-future require shorter chains. A 2% chance per year of a disruptive black swan yields 90% confidence after 5 years but 13% after a century.

        Plans on near future are also less fragile (i.e. less likely to be strictly conditional). When the Big Dig plans on a certain contractor being available by a certain date, the effects of an unavailability of that contractor can be ameliorated by finding another contractor. If you’re designing starships and we discover that dark matter introduces unexpected requirements, all bets are off.

      • http://juridicalcoherence.blogspot.com/ Stephen Diamond

        When the Big Dig plans on a certain contractor being available by a certain date, the effects of an unavailability of that contractor can be ameliorated by finding another contractor.

        If wonder if the absence of strict conditionality isn’t of the essence. You would naturally look for robustness before you apply best combos.

  • Jess Riedel

    (I hope people will forgive me for two long comments in this thread.)

    > Even though this is the standard approach of historians, schedulers, and puzzle solvers, many express strong disapproval about doing futurism this way

    Many *do* criticize historians for theorizing detailed models of past scenarios for which we have only scraps of evidence. And it’s silly to compare the use of this best-combo method in large construction projects (which you’ve done previously, and which is what I assume “schedulers” refers to) to designing starships; the difference in degree of uncertainty is huge.

    > Starship designers don’t seem to get much flack. Why? I’d guess it is because they are high status.

    Huh? Going to conventions about designing interstellar spaceships is about as low status an activity as exists for people with the necessary intellect. It’s in the same status bucket as LARPing.

    You’ve been published in glossy magazines about your far-future best-combo economic thinking! You shouldn’t feel aggrieved :)

    > If growth rates instead speed up by the same factor, the economy would then double every three months…

    Note that, regardless of whether they’re right, most people do not expect such a dramatic shift. Any criticism you have of most other people which is based on them accepting this prediction is misaimed.

    Also, your wordpress theme now makes it very annoying to copy and paste text from your post. (It adds unnecessary formatting, and gallingly inserts an advertisement: “- See more at: http://www.overcomingbias.com/….”).

  • free_agent

    In regard to predicting social things, one problem is that we care about the results a lot more than we do about engineering and such (which people are generally willing to leave to the experts). This is complicated by the fact that when talking about social issues, all speech is politics, not science — we constantly conflate and confuse what is, what ought to be, what I want to have happen, etc. because what we *say* about social issues affects what other people *do*, and we have enormous interests in getting other people to do certain things. (Since this sort of politics goes back to the origin of speech, we can assume that this is an evolved, instinctual skill and so we aren’t conscious of how much we do it, lest we give the game away.)

    In regard to starships, one interesting question is how close we are to *unmanned* interstellar flight. Technologies develop a lot faster once we start using them.

  • IMASBA

    “Well at least when predicting social consequences. Starship designers don’t seem to get much flack. Why? I’d guess it is because they are high status.”

    The status difference comes from one group being real scientists and the other being a modern equivalent of soothsayers (making predictions so vague they can mean anything, customizing predictions to what the client wants to hear, hindsight blaming of everything but themselves to explain why they didn”t predict an important black swan).

    “A trillion size bigger economy is likely to be a very different place.”

    Where exactly would the energy for such an economy come from for a civilization that’s not yet interstellar or only barely so? Does the Sun even produce that much energy?

    “you might figure that since the future is so uncertain, fictional starships are our best guess to real future starships. Since no one can know anything, no one can beat fictional imagination. But that seems very wrong to me; we get a *lot better insight into real starships from serious attempts to design them.”

    Probably… I mean who would put the bridge on top and have large windows all along the thin hull (Star Trek), or allow cigarettes, candles and non-electric shaving aboard (the trying to be realistic, but not entirely there yet, Battlestar Galactica 2003)?

    • AnotherScaryRobot

      “Where exactly would the energy for such an economy come from for a civilization that’s not yet interstellar or only barely so? Does the Sun even produce that much energy?”

      Some quick Googling reveals human civilization presently consumes about 10^13 watts, and the Sun produces about 4×10^26 watts, 40 trillion times as much. A trillion times current consumption is close enough to that limit that guessing at feasibility devolves into a discussion over the details of Dyson swarm engineering, I think.

      • IMASBA

        So not gonna happen anytime soon…

  • VV

    Watched the first video (Starships and Fates of Humankind by Peter Schwartz) and I was negatively impressed:

    He talks about faster than light travel, which is already a huge warning sign, and he makes the argument that since we don’t know everything in physics, and there might be a paradigm shift, then we might discover a new physics that allows us to invent FTL.

    That’s wishful thinking at its worst.

    Also most of his scenarios for space colonization, even those that don’t involve FTL seem unrealistic.

    A world dominated by high fertility religious fanatics is unlikely (contrary to his claim, religion is indeed on decline) and if it were to happen, it would be a Malthusian hellhole where people wouldn’t be able to afford interstellar travel even if it was in theory technologically possible.

    The same argument applies to the “escape from a dying earth” scenario.

    The “trillionaires in space” scenario relies on more exotic technology and anyway historically exploration has been always funded by governments.

    Ultimately his probabilistic claims are unsubstantiated.

    I didn’t watch the other videos, but if that was the typical level of the presentations, I suppose we can write that event off as a convention of crackpots.

    • Marc Geddes

      Well, FTL might be possible, but I think it would have to be subject to some pretty serious limitations. For one thing, by the theory of relativity, if FTL is possible, so is backward time-travel, so both are equally puzzling. For another, SAI (Super Intelligence) should have been able to get our region of multiverse by now if FTL/backward time travel were possible. SAI hasn’t been able to do it, showing that FTL/time travel is indeed subject to some pretty serious limitations.

      The most plausible theory I’ve heard is FTL/backward time travel is possible, but you need a device at both ends. Essentially, you have to create a pair of worm-holes, and you need a device at both ends.

      So for backward time travel, you can’t travel back to a time before the first machine is created. And for FTL, someone would have to lug equipment at slower-than-light speed first in order to set up the other worm-hole at the destination you want.

      Interesting fact though: As soon as we open the first worm-hole, we should expect SAI (Super-Intelligence) to pop through straight away. So this is an alternative route to Singularity.

      • VV

        As soon as we open the first worm-hole, we should expect SAI
        (Super-Intelligence) to pop through straight away. So this is an
        alternative route to Singularity.

        Which would be an instance of the grandfather “paradox”.

      • Marc Geddes

        There’s no paradox, its all explained how time travel can work in David Deutsch’s ‘Fabric of Reality’, with the Many-Worlds-Interpretation of quantum mechanics. In short, its the same as information exchange between different Everett branches – what comes through
        the wormholes are the results of different possible outcomes.

      • VV

        I can’t really make sense of this. In the MWI there is no significat information exchange between Everett branches once they have diverged macroscopically, and anyway, quantum mechanics and general relativty don’t play well together.

    • http://overcomingbias.com RobinHanson

      I recommended the talks on starships, not the one by Peter Schwartz.