Frozen Or Plastic Brain?

My post Monday on donating $5K to a Brain Preservation Prize testing fund has induced commentary (here, here, here, see also here). They’ve raised many issues on the choice between freezing brains or fixing them with chemicals.

Some prefer the term “chemopreservation” over “plastination”, which some artists have used to describe approaches that don’t try to preserve fine spatial detail. But I don’t like to replace clearer short terms with long vague awkward ones, just to avoid a weak association. If needed, I’d clarify by saying “ultra plastination”.

Some worry that we can’t prove plastic brains will last a long time. But brain researchers have looked at samples preserved many decades ago, and see almost no change. Tissues preserved in amber seem to have remain unchanged for forty million years. We have pretty good chemistry reasons to expect these plastics to last a long long time.

Some worry that plastic forgoes the prospect of reviving brain tissue directly after thawing, and relies instead on transferring its info to a new substrate, as with emulation. But direct revival seems extremely difficult given freezing and anti-freeze damage, and I think brain emulation is the future anyway.

Some worry that tests on fresh brains won’t show how well the techniques preserve less that fresh brains. But we could cheaply do tests now on not so fresh brains, after we test fresh ones.

The big issue, I think, is that plastination probably merges and diffuses some relevant chemical densities. If we knew about the minimal sufficient sets of chemicals to track, we could probably design dyes to mark such a set before we sent in the plastic. But since we aren’t sure which chemicals to track, we’ll have to make educated guesses, guesses that could be wrong.

Now many of us expect an awful lot of redundancy in brain cell spatial shape and various chemical densities, such that it will probably be enough to know the cell shapes, connection strengths, and the chemical densities that happen to be preserved in the first otherwise good plastination approach. If we go out of our way to tag a few more chemical densities, this can increase our odds. This is hardly a guaranteed approach though, so you might think freezing is safer, at least if anti-freeze can be shown to preserve more chemical densities.

But the much bigger risk, however, is that cryonics organizations won’t last long enough to keep brains frozen long enough. Most cryonics customers signed up a while ago, and their age distribution is slowly aging. If they can’t restart exponential growth, they’ll have more and more old dying customers relative to young paying supporters, and then may have a declining customer base. In addition, a great many managerial, political, social, etc. surprises could result in patient thaws even in a growing healthy organization.

Thus we unfortunately must choose between two unwanted risks – we must either suffer a plastination risk of not saving enough chemical densities, or suffer a cryonics risk of thaw by organizations with limited long term reliability. Since I judge the info saving risk to be mild, and the organization reliability risk to be severe, I’d choose the former.

GD Star Rating
loading...
Tagged as: ,
Trackback URL:
  • Pingback: Overcoming Bias : Frozen Or Plastic Brain? « Mind Whimsy

  • Robert Koslover

    1.  Does this mean you are having second thoughts about being cryogenically preserved?  (If so, I did not expect that.)

    2.  Re: “If they can’t restart exponential growth, they’ll have more and more old
    dying customers relative to young paying supporters, and then may have a
    declining customer base.”  — That’s known as “Social Security,” right?
    :)

    • Matthew Graves

       “1.  Does this mean you are having second thoughts about being cryogenically preserved?  (If so, I did not expect that.)”

      This is actually one of the more exciting things I’ve seen written by Hanson, since one of my primary reasons for not signing up for cryonics was pessimism about its success chance. (And assuming finite total lifetime utility, so that the chance does actually matter.)

      To see that Hanson’s cryonics support was (probably) because it was the best of bad options, and that he’s willing to jump ship to a better option as soon as it appears, makes it clear that he’s interested in his ends and not his means, which is refreshing.

      • officer_fred

         Why do you say that’s refreshing? I would think most cryo types just want brain preservation, and aren’t too attached to any method. (Really wondering: I’m new here, so maybe there’s history I’m missing).

      • Era

        What you’re missing is that this blog used to be co-written and get comments from members of the lesswrong community who fanatically support cryonics and are blind to its weaknesses. If robin was truly interested in pointing out bias he would point this out, but this post and others recently are a big step in the right direction.

    • Robin Hanson

      I remain a cryonics customer.

  • V V

    Now many of us expect an awful lot of redundancy in brain cell spatial
    shape and various chemical densities, such that it will probably be
    enough to know the cell shapes, connection strengths, and the chemical
    densities that happen to be preserved in the first otherwise good
    plastination approach. If we go out of our way to tag a few more
    chemical densities, this can increase our odds.

    You keep saying that, but it seems that you have in mind a very simplified model of biological neurons, probably inspired by artificial neural networks.
    AFAIK, fine details of the cell surface features seems to be functionally relevant.

    This is hardly a guaranteed approach though, so you might think freezing
    is safer, at least if anti-freeze can be shown to preserve more
    chemical densities.

    The anti-freeze chemicals may also disrupt fine detail due to denaturation and osmotic shock. I never saw that point addressed by cryo-proponents.

    Most cryonics customers signed up a while ago,
    and their age distribution is slowly aging. If they can’t restart
    exponential growth, they’ll have more and more old dying customers
    relative to young paying supporters, and then may have a declining
    customer base.

    Are you saying that cryonics organizations are essentially Ponzi schemes?
    Anyway, I don’t think that plastinated brains would have far more chance of remaining preserved in the event of organizational failure.

    • gwern0

       > Are you saying that cryonics organizations are essentially Ponzi schemes?

      It’s more complex than that. Alcor magazine has a number of articles covering the demographics stuff; the quick summary is that members tend to underestimate inflation and prices haven’t always been actuarially perfect, so there are long term imbalances. A steadily increasing membership would avoid that, but I don’t think Ponzi scheme is the right phrase, unless you want to dilute the meaning of that phrase even more than it already has been…

      • V V

         

        It’s more complex than that. Alcor magazine has a number of articles
        covering the demographics stuff; the quick summary is that members tend
        to underestimate inflation and prices haven’t always been actuarially
        perfect, so there are long term imbalances.

        IIUC, the claimed business model is that part of the life insurance premium gets invested in a low-risk fund, and the revenue from this investment should cover all the expenses for that particular preserved person, in perpetuum.

        I assume that the investment fund didn’t fail, so if its revenue somehow fails to cover the expenses, then either the estimates were grossly off, or something fishy is going on.

        A steadily increasing membership would avoid that, but I don’t think Ponzi scheme is the right phrase

        Relying on an increasing number of members to pay off to old members with new members money fits exactly the definition of Ponzi scheme. In this case there might be possibly no intention to defraud, but the end effect will be the same: these business models are all eventually unsustainable.

      • http://lukeparrish.rationalsites.com Luke Parrish

        The definition of Ponzi scheme includes intent to defraud. You lose at least half the meaning there. You also aren’t considering the possibility that there would be economies of scale that make it legitimately more sustainable with larger numbers of patients. Have you run the numbers on storage of LN2 in large versus small volumes?

      • V V

        The definition of Ponzi scheme includes intent to defraud. You lose
        about half the meaning right there. The fact that it is not an
        investment for direct monetary gain is also different.

        It doesn’t make any difference with respect to the unviability of the business model.

        Moreover, this framing obscures the fact that economies of scale make cryonics legitimately more sustainable at larger volumes.

        I don’t think so. If the number of cryopreserved people becomes any signficant fraction of the human population, they’ll start effectively competing for resources. And I think it’s pretty easy to guess who is going to prevail between hungry living people and frozen corpses.

        Moreover, if a restoration technology is ever developed, it is reasonable to suppose that only a few preserved peoples are restored and reinserted to the society. The more preserved people there are, the less the individual chance of any one of them to be restored.

      • http://lukeparrish.rationalsites.com Luke Parrish

        It doesn’t make any difference with respect to the unviability of the business model.

        Agreed. If this is a serious description of cryonics (and I don’t think it is) the label “Ponzi-like” applies.

        I don’t think so. If the number of cryopreserved people becomes any
        signficant fraction of the human population, they’ll start effectively
        competing for resources.

        Not true, since cryopreserved people consume a very small fraction of what
        living people do. You would have to have significant multiples, not fractions, to seriously compete for resources.

        Furthermore the efficiency gets dramatically better on a larger scale. The
        energy costs to maintain cold temperatures in a given volume are 
        based on the surface area of the container — you only have to insulate
        the outside. You can also add thicker insulation without running up
        against diminishing returns from the increased surface area as quickly.

      • V V

         

        Agreed. If this is a serious description of cryonics (and I don’t think it is) the label “Ponzi-like” applies.

        Why don’t you think it is?

        Not true, since cryopreserved people consume a very small fraction of what
        living people do.

        But they don’t work hence they don’t produce anything, and given that cryonics organizations have some financial difficulties, I’m not sure the maintenance costs are so low.
        Anyway, even if the investment fund was producing more revenue than the expenses, it would act as a drain that removed wealth from the economy.

        Furthermore the efficiency gets dramatically better on a larger scale. The
        energy costs to maintain cold temperatures in a given volume are 
        based on the surface area of the container

        That’s not relevant since there is a maximum size above which making containers becomes impractical.
        Alcor, for instance uses four-bodies six-heads containers, not a single refrigerated warehouse for all its >100 cryopatients.

        Anyway, IIUC, refrigeration expenditures are just a small fraction of the total maintenace cost per person.

    • Robin Hanson

      Yes of course there exist functionally relevant fine surface details. The question is the functional redundancy of those with all the details that are preserved in some process.

      • V V

         I’d expect that there is some redundancy within each class of features. But if you completely discard one class of features, because it is not preserved by your process, it’s  unlikely that all the relevant information it contains is mirrored in the classes of features you preserved.

      • dmytryl

        I would not expect full blown redundancy at microscopic level, across different types of features. Does plastination even preserve the surface proteins (e.g. the ion gates and receptors)? Everything seems pretty vague on this point. If it misses one type of receptor that is e.g. used for communication with glial cells, and if density of that receptor is variable and important for anything, then i wouldn’t expect it to be replicated somewhere else. Similar to how you would not expect e.g. connections alone to be sufficient without the gate and receptor density. Any big blunder and you’re dead. Need to take couple dozen cells, test electrical properties, do several times to have a sense how much it varies, then try to upload this and simulate precisely. Sort of uploading version of viability test.

      • daedalus2u

        The redundancy of the brain is more
        like parallel circuits that are always live and synchronized so that
        circuits that fail can be bypassed in a few milliseconds with no
        interruption in consciousness. That requires automatic and
        continuous “tuning” so that operation out of sync by milliseconds
        doesn’t happen (that is what causes hallucinations and delusional
        thinking).

        Subtle perturbations of brain chemistry
        (for example by a few whiffs of xenon) produce unconsciousness. That
        is from a breakdown in regulation and synchronization of the brain.
        How and why things like xenon cause such a breakdown is completely
        unknown.

        The redundancy of the parallel circuits
        that allow for uninterrupted consciousness does nothing to induce
        resistance to xenobiotic chemicals such as xenon. How will it be
        possible to tell if one part of the brain is being emulated in a
        psychotic state while another part is in a delusional state and
        another is in a hallucinatory state?

  • Ilya Shpitser

    Why don’t we preserve brains in amber?

    • daedalus2u

       The problem with amber is that it can only preserve small things.  The resins in amber take a long time to diffuse into something macroscopic and something like a brain would degrade during that time. 

    • jhertzli

       Forever amber?

  • Jason

    I am disappointed the prize has received so little to date.

    When prizes like this are proposed as a method of government funding the great advantage is that they are only paid out if the desired result is achieved.  But the same benefit doesn’t accrue to individual donations – if I give $5,000 I lose that money regardless of whether the prize is paid out.  

    I think the ideal contribution method would be to sign some kind of conditional contract agreeing to contribute a certain amount of money only if the prize is awarded.  Are such contracts possible?  There would be some logistical issues with bankruptcy, meaning the actual prize would probably be less than the total pledged contributions, but hopefully still a substantial fraction.

    There must be some very rich people who would be willing to make large conditional contributions of this sort.

    • daedalus2u

       Prizes are a gimmick.  They are not going to work to stimulate research and development because the value of the “prize” is tiny compared to the cost of the research.  No one who is rational would spend $10 million in anticipation of “winning” a $100k prize.  Why would anyone rational think that prizes are a way to spur research and innovation?  In my view, it is all wishful thinking.  

      I think that the mindset that thinks that prizes from rich people are a way to spur innovation is a large part of the problem with why there is insufficient innovation.  

      Doing research requires time to understand the field, research equipment and materials.  In our current economic structure, all of these require money.  Money for education, money for scientific literature, money for food, housing, internet access, health care, research space, research equipment, research supplies.  

      To those who think that prizes are an effective way to spur innovation, where does the up front money for all of these things come from?

      One of the reasons that innovation is difficult is that it is expensive, and it is expensive because businesses work to extract as much profit at they can from every transaction.  Why does Elsevier charge ~$30 for a paper they were given for free by the authors 30 years ago?  So Elsevier can have a 30% profit margin.  Why does Elsevier have a 30% profit margin?  Because scientists have to publish and read papers and so Elsevier can charge what the market will bear.  Charging what the market will bear is good business, but it stifles the growth by diverting resources into profits, which preclude low cash-flow enterprises from buying the products.  

      R&D is always going to be a low cash-flow enterprise because there is nothing to sell yet because it is still being developed.  Lowering the cost of R&D by reducing overhead, health care costs, housing, etc, would do more to spur innovation than offering a few prizes for chump-change.  

      But you are right, with the mindset that with a grant you “lose the money”, who is going to front the money needed to do the research?  

      • Robin Hanson

        When choosing between the effectiveness of encouraging research via grants or prizes, we should compare ones with similar financial costs.

      • V V

         Academic circles often organize competitions, usually with no prize or just token prizes, since the real reward is largely social. These competitions do seem to spur innovation.

  • James Andrix

    This isn’t directly related to risk. My gut reaction is that plastication has a lower ick factor than cryonics, but I am uncertain. e.g. Are there more people who find frozen heads creepy, but would like to have grandpa’s brain in a case on the mantle? Or just plain burying the plastinated brain with the body, to my mind it ‘feels’ almost traditional.

  • mjgeddes

     
     Hacker’s Maxim #10: ‘Don’t lose your head’
    http://www.youtube.com/watch?v=LcG5HQnCBU4

    • V V

       Self-interested Singularitarian’s Maxim #1: ‘There can be only one’

  • Dennis Peterson

    Some researchers in Japan seem to have made a significant advance in cryonics. They were trying to preserve sushi. They drop the sample well below freezing while vibrating the water molecules with a magnetic field, then turn off the field for instant freeze. No ice crystals, no chemicals. Now they’re exploring it for organ transplants. Google “sushi freeze magnetic organ.”

    Doesn’t solve organizational failure, but bring a mouse back from deep freeze and the cryonics companies’ financials are likely to improve.

    • V V

       Possibly, but keep in mind that a mouse weights 0.035 kg, while a human brain weights 1.4 kg.

      How does the technique scale?

    • http://lukeparrish.rationalsites.com Luke Parrish

      I seem to remember Brian Wowk took a look at that back when it first came out, his verdict was that it is most likely not really doing what it sounds like. (Possibly something like forming smaller ice crystals. Preventing ice from forming altogether was not considered physically plausible.)

    • daedalus2u

       Their patent is nonsense.

      6,250,087

    • dmytryl

      I did read the paper’s abstract:

      http://www.sciencedirect.com/science/article/pii/S0011224010000854

      The field was 0.01 milliTesla, or 0.1 Gauss. The Earth’s magnetic field is about 1 Gauss. Complete bullshit, no further reading necessary.

      I guess they tested different fields and had many enough tests so that one passed the statistical significance criteria.

  • Peter McCluskey

     I think Alcor would be able to handle a declining customer base by cutting back on the speed and quality of new suspensions.

    My biggest concern is that they will become focused on preservation to the exclusion of reanimation. The conservative approach they need to keep existing suspendees safe is psychologically very different from the attitude needed to pioneer reanimation. I can easily imagine them permanently saying “we need more research to make sure it works well; we don’t yet have enough money for that research”.

    Plastination might avoid that risk if the plastinated brains could be stored somewhere such as a bank safety deposit box where the entity responsible for reanimation is separate for the storage business, and the company responsible for storage wouldn’t depend on this particular customer base to remain in business.

  • Pingback: Overcoming Bias : Brain Prize Eval Fund Near Enough