Live Long Or Wide?

One of my favorite science fiction novels is Kiln People, by David Brin in 2003. Not so much for its characters or plot, but because it takes an interesting future/tech scenario seriously. Most fiction with artificial intelligence describes a world with only a few of them, yet one of AI’s most important features is its easy of copying.

In Kiln People, Brin takes seriously this idea of cheaply copying intelligent agents. The key assumption is that in a few minutes and for a modest cost one can copy a person’s mind into a new clay body that lasts about a day. That copy’s memories of its day can also be added back into the original at the day’s end. Brin imagines many details of how this would change society. While he gets some things wrong, and an economist would get more right, Brin does far better than most science fiction.

Assume for the sake of argument that you came to accept that such clay copies really were “you.” So that if on Monday you made six copies and merged them all back in at the end of Monday, and then you slept the rest of the week, you would have lived just as much as an ordinary person in a normal week. You’d remember having lived for seven days that week.

Now imagine that this copy technology is improved let copies last ten years. Then compare two ways to stretch your life:

  • Time Stretched Life: You are able to live for another 110 years before dying.
  • Space Stretched Life: You make nine copies now, and the ten of you live for ten years. Then you merge the memories of all these copies back together, and live for another ten years before dying.

I suspect most people would admire the life stretched across time more than the life stretched across space, similar to the way most people admire a time stretched civilization more than a space stretched one, and to the way they accept time genocide more than space genocide. I again attribute this to the future seeming more far:

The far future seems more far … than situations far away in space, or in the far past. The near/far distinction was first noticed in how people treated the future differently, and our knowing especially little detail about the future makes it especially easy to slip into abstract thought about the future. … We are less practical, more idealistic, and more uncompromising in far mode.

Added 8a:  The time stretched life lets you see more of human history, but the space stretched life lets you help yourself more (e.g., the ten of you could start a business together), is better able to prevent your death, and trades later for earlier decades of your life cycle. As most people seem to discount the future and to prefer earlier life decades, these factors seem to favor space-stretching overall.

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

    You’re just plain wrong on this one. In the space-stretched scenario, I only get to see what happens to the world 20 years out. In the time-stretched scenario, I get to see what happens 110 years out. Look at the last 110 years of human history – all the developments in our culture and technology – and tell me who takes the former scenario.

  • http://python3porting.com Lennart Regebro

    I prefer a time stretched life because merging memories seem confusing. Also I’ll be able to see more cool inventions.

    (But the time genocide vs space genocide arguments where interesting).

  • http://danieltarmac.blogspot.com Henry

    There is significant value in seeing events unfold over time. There are new inventions (as Lennart mentioned) but also new arts, fashion, politics, etc. And if you have children or grandchildren, being able to see them grow up is immensely valuable.

  • Ernie

    The value of time spent living isn’t just the sum of the length of time – it’s its relation to the past. The guy who lives for 110 years has a much larger average number of remembered years than the other.

    Take it to the extremes – would you like to have millions of copies that live for just a minute each? or one copy that lives 1/4 of each year but lives for 440 years? I’d prefer the latter.

    • michael vassar

      Agreed, but this may not be so relevant when talking about more than a couple decades. How much identity do I really have with myself of half or double my age? I think surely less than half.

      I think I’d choose to maximize my life-stage adjusted years, which based on my impressions of other people (I massively distrust self-reported happiness) would probably mean years in one’s 30s to 50s. Put me down for 4 27.5 year copies, e.g. making it to 60 and having 4x the impact on the world until then.

      I guess my 59 year old selves might disapprove of my choice, though I honestly don’t think they would, but I’m probably unusual. I think most people’s older selves would.

      I imagine one family oriented, one career oriented, one abstract thought oriented and one rich, diverse experience oriented middle age would be most appealing to most people with tastes similar to mine, and much more appealing than old age.

  • http://www.spaceandgames.com Peter de Blanc

    I expect that having more knowledge leads to a better life. You spend more time with more knowledge in the time-stretched life than in the space-stretched life. This could perhaps be mitigated if you exchanged memories with your copies continuously instead of waiting for 10 years.

  • Bryce

    The time stretched life and the space stretch life don’t seem like a fair comparison (or at least, not as fair of a comparison as the time stretched genocide and space stretched genocide comparison). I suspect most people place a different amount of value on different periods within their own lives, and in the example you created, you can’t parallel process your twenties, thirties, fourties, etc.

    That said, I would imagine many people would favor a scenario in which they could live their twenties or thirties multiple times in exchange for ten years off the end of their lives.

  • Armok

    Yea, your scenario has way to many unrelated facotrs to actualy work, here’s an adjusted version:

    You make 9 copies like described, and then live for 1 year, merging memories each day so each copy gets the advantage of new learned things etc. After that, you freeze yourselves cryonicaly for 9 years, then live all 10 copies syncing memories every 24h for another year and then sleep for another 9 years etc. Until the total is up to 110 years.

    This life I’d probably actually *prefer* to your described time stretched one, for various reasons.

  • Schmauli

    I think the comparison is fair, but that in the end, it boils down to what one thinks ones purpose in life is. Saying that 10 simultaneous lives of 10 years are roughly the same as one life of 100 years, to me seems based on the assumption that life is about collecting memories. So the thing one ought to do is: go out, experience stuff, in order to remember it later. And a useful life has been one with xx years worth of memories in the end?

    If this is so, then when does one stop experiencing, and start remembering? Because if one doesn’t start to remember at a certain point, all the experiencing has been a waste of effort (based on the premise that the time-stretched and space stretched are roughly equal because they yield the same amount of memories).

    Maybe the space stretched life opens possibilities for experiences that aren’t available in a linear life. Just imagine being able to live on both sides of a world conflict. Would it lead to great understanding, or a completely screwed up idea about things as good and wrong?

    Awkward.

  • Alrenous

    The assumption that I accept they’re me disguises the real issue.

    If you prick them, do I feel bleeding?
    Later I gain a memory of having been pricked. It doesn’t hurt anywhere near as much as the actual thing.

    Similarly, would you prefer to experience a very fine movie, or would you be okay with randomly gaining a memory of having experienced such? (Does it even matter whether it actually happened?)

  • http://www.gwern.net gwern

    > While he gets some things wrong, and an economist would get more right, Brin does far better than most science fiction.

    I’m re-reading _Kiln People_ now and it seems very good.

    The only thing I would count as wrong is the underspecialization of workers; he mentions that remote tele-dittoing would instantly collapse many job markets to the replicas of just one em (I mean, ditto), but doesn’t cover that his setup allows for something similar, one person having a fulltime job making their dittos and re-uploading, freezing the dittos cranked out every few minutes, and shipping them world wide.

    But that seems like a pretty small error (at the speeds and longevity given, one master could only replace about 96 workers, if they sleep 8 hours and make one every 10 minutes and dittos last only one day).

  • Michael Kirkland

    Someone who chooses wide gets just that. Someone who chooses long may get both, due to technological improvements, but at worst they come out even.

  • Konkvistador

    “I again attribute this to the future seeming more far:”

    Perhaps this is so. Intuitively it seems to work that way.

    I personally don’t feel any more fuzzies thinking about a time stretcher compared to a space stretcher who chooses really distant locations.

    A. Get extra 30 years of life here.

    B. Get 10 years of life here while simultaneously live on a tropical island and Japan.

    I’m not sure people would admire A more than B.

  • jb

    There won’t be that many advances in technology in 10 years. There will be a lot of advances in technology in 110 years. Time is an input to this process, and you can’t ignore its potency.

    This isn’t the future seeming “far” so much as well-established facts about scientific advancement.

    If you give me a scenario where technology is held constant over those 110 years, so that there’s no technological difference between the two situations, I’d still take the time-stretched one, because I’d get to meet more people (grandchildren, nieces, nephews, etc), and more stuff would happen (volcanoes, etc).

    If you give me a scenario where technology and people and events are held constant, then I become indifferent to the two options. But that seems like a horribly strained scenario.

  • tom

    I don’t like the two alternatives because they’re missing one key thing: in any world in which could copy myself, I would also be able to improve myself.

    If I was short, ugly, unathletic and socially awkward, my first wish would be to improve those traits so that I would rise relative to others and be able to interact with people on better terms. The most direct way to do that would be to improve those traits in myself. It would not be to make multiple weak mes.

    An army of losers? Or one super me? I can see where a weird equilibrium would emerge depending on which types of technology is more advanced. But I think I’d want multiple mes to improve my status (by solving my problems or making me money) much more than I would want them to increase the number of experiences I have at a lower status.

    Bring on the personal Clone Army!

  • Pingback: TheMoneyIllusion » The Good Life: A bleg

  • Hyena

    I think you’d want to come up with a good space/time stretching equilibrium. It might be better to have three of me for 40 years, for example, that 10 of me for 10 years because my newfound manpower may not translate into a linear reduction into the time it takes to achieve my goals.

    I think, in general, I’d want to be many when I need to cover a wide area and few when I need more time. So it would be good for 3 of me to work hard from 20 to 30, for one of me to hang out in Mumbai between 30 and 40, and many of me to spend the accumulated gains from 40 to 60 so that one of me can enjoy the memories in the years before he dies but is too frail to do much (why have 6 copies of the hospital bed experience, after all).

  • Michael Wengler

    Is the memory of living through something really the same as having lived through it? In that Dick/Schwarzennegger movie about Mars, there is a company that sells vacations… or rather the memory of vacations at a tiny fraction of the price of the actual vacation. Given that we know memory is very spotty, that it is a sparse set of clues about the remembered event which the brain then weaves into a richer replayed story when it is remembered, this seems much technically simpler than creating a copy of someone and then merging the copies back into one at the end of the day.

    Perhaps I am biased at the effort I have spent trying to merge the coding of two java programmers (I am one) after 4 hours of parallel development starting from the same code base. The conflicts are astonishing and NOT automatically fixable.

    I suspect I would buy pleasant memories if they were really cheap. I would not value them equally as having lived through the same thing. One can argue that its only because I know (or think I know) they are not “real” memories, but clearly the value we place on things is infinitely tied to what we think we know about it.