Masters of Time: Sir Arthur C. Clarke on resurrection technology

I am re-reading for zillionth time my collection of works of the Master, Sir Arthur C. Clarke. This is a selection of his writings on time scanning, resurrection and afterlife.

Sir Arthur was not a believer in any traditional religion, and did not believe in resurrection and afterlife. Yet, he had a “Possibilian” open mind and often said, for example in conversation with my friend José Cordeiro: “I’m always paraphrasing J. B. S. Haldane: “The universe is not only stranger than we imagine, it’s stranger than we can imagine.”

In Profiles of the Future, revised millennium edition, at the end of the chapter “Brain and Body,” Sir Arthur wrote:

“I recently sacrificed some of my few remaining hairs, to be launched into space as part of the AERO Astro Corporation ‘Encounter Project’.

If all goes well, they will leave the Solar System (after a boost from Jupiter) and the hope is that, maybe a million years from now, some super-civilisation will capture this primitive artefact from the past.

Recreating its biological contents might be an amusing exercise for their equivalent of an infants’ class.

Of course, I’ll never know — unless the experimenters are both very considerate — and Masters of Time.”

It is easy to imagine that a super-civilisation may be able to recreate a body from biological samples, but to recreate his personality and memories, thoughts and feelings, they would need to be Masters of Time with at least read-only access to the past. So it seems that here Sir Arthur had in mind resurrection via time scanning technology plus mind uploading, aka “copying to the future” or quantum archaeology. In the chapter “About Time” of the same book, he is more explicit:

“If there is any way in which we can ever observe the past, it must depend upon technologies not only unborn but today unimagined. Yet the idea does not involve any logical contradictions or scientific absurdities…

The reconstruction of the past is an idea even more fantastic than its observation; it includes that, and goes far beyond it. Indeed, it is nothing less than the concept of resurrection, looked at in a scientific rather than religious sense.

Suppose that sometime in the future our descendants acquire the power to observe the past in such detail that they can record the movement of every atom that ever existed. Suppose they reconstruct, on the basis of this information, selected people, animals, and places from the past. So, though you actually dies in the Twenty-First Century, another ‘you’, complete with all memories up to the moment of observation, might suddenly find yourself in the far future, continuing to live a new existence from then onwards.”

In the novel “The Light of Other Days,” Sir Arthur and Stephen Baxter describe a fictional time scanner, the “Wormcam”: a remote viewing device that permits scanning any location at any time, including in the past, by using micro wormholes naturally embedded with high density in the fabric of space-time (every space-time pixel is connected with every other space-time pixel). Soon, engineers are able to resurrect the dead:

“It was possible now to look back into time and read off a complete DNA sequence from any moment in an individual’s life. And it was possible to download a copy of that person’s mind and, by putting the two together, regenerated body and downloaded mind, to restore her…”

  • Mike Perry

    To start, A. C. Clarke, for all his good points, was a particular frustration and saddening disappointment to a cryonicist like me. He actually went on record saying he thought cryonics had a good chance of working, but would not choose it himself. His given reason was that (paraphrasing) “we become a different person after ten years anyway”–that is to say, unless you deanimate soon after cryopreservation it won’t be “you” that is being preserved–or that is brought back if cryonics works. I suppose that was a deep-seated feeling not amenable to logical arguments–I can only guess. But losing anybody this way seems tragic to me, albeit like Giulio himself I have serious hopes in some form of eventual resurrection. But I don’t think there will just be a straightforward “readout” of necessary information from the distant past, that is to say, “quantum archaeology,” in bringing this about. I think QA will prove impossible just as there is no way at present to unmelt an ice cube and put every particle (water molecule etc.) back in its original place. And I realize there might be some counterarguments that seem to suggest it should be possible, based on “conservation of information” and the like. Still, for a number of reasons, I remain a firm skeptic.

    An alternative to QA should be possible, however, based on the property that the main thing you need as starting point for a resurrection is just a suitable database from which to begin. This database would include a complete description (down to atomic level, say) of the person or persons you’d want to resurrect. (Starting with that, and advanced future technology including nanotechnology, you make copies of all the people, which reasonably qualify as resurrections, at least if your outlook is similar to mine.) With finitely many people, the database would be finite too. On the other hand, only a finite number of people (plus other sentient life forms you’d also like to give consideration to) have lived in the earth’s biosphere since our prehistory began. If the universe or some accessible extension of it is “reasonable” you should have enough time and other resources to construct your (again, finite) database. How many bits would it have to contain? Say somewhere in the neighborhood of 10^40 to be conservative. (10^40 would allow for 10^11 total people at 10^18 bits per person, times another 10^11. Big overkill probably, use for sake of argument.) So, if you randomly guess 10^40 bits there is a small but nonzero probability you get them all right. Here then is a way in principle of resurrecting without any QA. Of course you need to do better for reasons of practicality.

    The approach I would advocate for “doing better” is what I will call “multiverse archaeology”–MA in contrast to QA. First, you assume the earth’s biosphere has been mined to recover all historical data that survive. So every fossil, etc. will have been examined along with the more recent remains of our human civilization. This you expect to accomplish through foreseeable technology of the future including mature nanotechnology. (Swarms of burrowing nanites and the like might accomplish this historical survey in a matter of decades or centuries.) You also will recover information from living individuals who by then will hopefully have escaped such limitations as aging in a biological body and will be interested themselves in participating in this grand and morally uplifting project.

    I should mention that some relevant historical information will be contained outside the confines of the earth, such as remains of lunar landings, and there could even be historical information recorded by extraterrestrials using telescopic observation or whatever detection methods they might have. This issue would need to be investigated, but hopefully at some point you will have arrived at a historical record that could be considered complete or very nearly so.

    So now, instead of just a random 10^40 bits your database will be so adjusted as to be consistent with the historical record as recovered. This means that, while there will be much information now available to be added in from the historical record, still, much more will have to be recreated by guesswork to fill out a reasonable historical timeline that includes all sentient life forms of the version of the past you are creating. Possible in principle though, without an exotic data recovery process or “virtual time travel.”

    Every person recreated then will fit expectations. Your deceased ancestors for example will perfectly fit whatever you or anyone else may remember about them, together with meeting reasonable expectations if you should ask them questions whose answers are not found in the historical record.

    In effect, then, this is a resurrection by means other than QA, something I am calling MA. More work is needed, of course, to further rationalize it as a “real” resurrection and not just a creation of fantasy individuals who “very likely did not really exist just as you have created them.” To do this requires some additional argument, based around expectations of the existence of a multiverse. I assume there is in fact a multiverse, a collection of multiple universes or something equivalent, in which histories in all their varieties are happening. Moreover, when information is lost from the historical record in our universe it literally makes the past ambiguous. That means I am saying that history in effect does not have a unique timeline. (Alternatively, if this is not so it means information is never really lost and QA is possible after all.) With the amgiguity comes a kind of authenticity. Your database with the details you had to recreate by guesswork will not be the only authentic past but will definitely be one of them.

    Moreover, the expectation is that, in all the parallel universes, there must be other resurrection projects just like yours that will come into being at the “moment” you start yours (if the time reference is meaningful). These in turn should produce databases just like yours except where guesswork was needed to fill in gaps in the historical record. And the guesswork that is done by those alternates of yourself will exhaust all the logical possibilities so in fact every possible variation of every possible person will be resurrected. If you worry that your “real” lost ancestors could not be the ones that were resurrected in your project, don’t worry, they must be resurrected in some other project that will seem to be “home” to them. However, the ambiguity of the past means that actually the version of ancestors that you resurrected will be just as much the “real” ones to you as the others would have been. You do have to accept the idea that all along you never fully knew who your “real” ancestors were–in fact that ambiguity already existed. So you can bond with your resurrected ancestors in confidence and even go on with the “discovery” process as they become better known to you than they were before.

    And so on. I have thought about the difficulties of all this for a long time and think there are reasonable answers. If QA is impossible we still have MA. Resurrection I think is a real possibility, and actually inevitable in a prolific reality-as-a-whole.

    I haven’t said anything about why I think, given resurrection appears to be inevitable, one should still choose cryonics. Another long story, with a number of arguments, for another time. (Both cryonics and resurrection arguments are covered in my book, *Forever for All*, free online at, though I have some recent refinements.)

    Mike Perry

  • I love it! As you probably know, I subscribe to block-time eternalism and digital physics (everything is fundamentally information and matter is a myth) This makes space-time a type of recording media. Resurrection is probable in this light.

  • Giulio Prisco

    James, I would use terms like “probable” because we just don’t know enough about these things to make any probability estimate. If you only know that in a jar there are some white balls and some black balls, the probability to pick a white ball is totally indeterminate, you can only say that it is between zero and one.

    The idea that in some conditions space-time may act as a type of recording media is, in some cases, trivially true: hit a piece of wood with a hammer, and a record is kept that permits to reconstruct the main aspects of the dynamics of the hit. In other cases, the record is much more difficult to detect. In yet other cases, current science seems to say that the record is irreversibly lost to classical dissipation (irreversible thermodynamics) and/or quantum collapse. Yet, usable “akashic” records may be kept in places that we have not found yet.

  • Giulio Prisco

    Good to see you here Mike. The multiverse has a metric, a notion of distance. The versions of me who did not go to the to the mall yesterday are close to the versions of me who are writing this sentence now, but the versions of me who were taken to live in a country far away at age 3 are not.

    I can accept a version of me from a close branch as a valid resurrection copy, but a version of me from a far branch would be someone else (but still better than a version from a very, very far branch like that child who was abducted by octopus-shaped aliens at age 1;-)

    Re “one should still choose cryonics” – I most certainly agree, because while I hope to be resurrected by a future super-civilization, I have no means to be certain that this will happen, so cryonics increase the chances of surviving to see an interesting future. Also, I wouldn’t want to miss the fun of the next couple of centuries.

  • Mike Perry

    Hi Giulio,

    You say,

    “I can accept a version of me from a close branch as a valid resurrection copy, but a version of me from a far branch would be someone else (but still better than a version from a very, very far branch like that child who was abducted by octopus-shaped aliens at age 1;-)”

    Maybe there is a misunderstanding here (well-intended of course). I posit that in parallel universes that are different overall, nevertheless, there are copies of you that are exact down to the bit level. This should not pose any great difficulty if we accept the usual ideas about the multiverse. Different universes might be expected in fact to have a good many features in common, just not every feature. (If they did, then we just identify these universes, via “identity of indiscernibles.”) “You” are not localized to this one universe (at any rate we think of ourselves as being in just one universe that I call “this” universe) but in fact are distributed over all these other domains too, *right now!*. That is to say, right now *you* have no way of knowing which universe “you” are in–so reasonably you (as an observer, a special type of entity) are really not confined to one universe but are distributed over many. (This might be said to underscore the property that while universes can be called “material” clearly observers are something other than this, call it “informational.” Actually I think the case can be made that “all is really informational,” agreeing with James above, but that’s another effort.)

    At any rate, the non-locality of observers does I think pose one more challenge to QA (but not to MA). That is to say, the very history you would uniquely reconstruct with QA is itself ambiguous. (The ambiguity will be affected by probability weights, but not eliminated.) With MA you accept the different histories on an equal (more correctly, a probability-weighted) footing. You choose one, probabilistically, and all should be well (to a first approximation anyway). Your resurrection if needed would basically reproduce an exact informational copy of an original “you”, or possibly go directly to a more advanced version of that being or “continuer.” There should not be any difficulty with a variant of you which might not really be “you.”

    On cryonics, you give as a reason to choose it (rather than depend on eventual recovery by other means): “I wouldn’t want to miss the fun of the next couple of centuries.”

    Good starting point, and one example of why one should consider cryonics *even if* one is very confident that one would be resurrected eventually regardless. In either case you would be taking a certain path through the future, but the two paths would differ. Call cryonics the “lesser leap” and the alternative (eventual resurrection) the “greater leap.” Why choose the lesser over the greater leap, if both will lead eventually to happy eternal life? I propose that the lesser leap, if properly pursued, will yield the greater reward, before things finally stabilize and from that point the two paths might be more-or-less equally good. In the first place, as you say, the greater leap will cut out interesting historical events you’d like to be part of rather than just learning about it afterward. In addition you place yourself in the role of “heaven builder”–trying along with others to make the future world what it should be for everyone, rather than a “heaven enjoyer” merely. Benefiting others should reap reciprocal benefits. So overall your reward or benefit should be greater, in view of enlightened self interest.

  • Giulio Prisco

    @Mike re “I posit that in parallel universes that are different overall, nevertheless, there are copies of you that are exact down to the bit level.”

    I don’t think I understand this point. Surely in some parallel universes that are _very_ different overall, for example in a universe where life never developed on Earth, there cannot be an exact copy of me.

    Here is an argument for cryonics in a multiverse, which I find quite solid:

    The “quantum immortality” argument posits that after you die you continue to find yourself in a branch where you are still alive. But of course at a certain point the probability of your survival becomes very low: if you jump off a window at the 50th floor every day (“quantum suicide’), there is a growing sheer improbability that you find yourself still alive, so the region of the multiverse where you are still alive becomes extremely thin, because in the vast majority of branches you are dead.

    Now, instead of jumping off a window, just let yourself age without cryonic arrangements. Suppose that a cure for aging and death (for example via mind uploading) is found in year 2150. But then you would have to be so lucky to survive until 200 by some other means. There are some branches of the multiverse where this happens, but they are only an infinitesimally small region of the multiverse.

    But in all branches where you do have cryonic arrangements now, you have a very plausible path to be there in 2150 (you are suspended and successfully revived). So we can assume that you survive in many of these branches.

    In other words, signing up for cryonics is a means to live in a much bigger region of the multiverse.

  • Thanks for this site Giulio,

    I’m trying to track Quantum Archaeology on my site but it’s ideas & scope are already to big to pin down.

    Mike as regards not being able to reform an ice cube that melted back into the exact structure it held…why could we don that? It seems that QA posits that we must or water holds some property that is mystical (I know ur not saying that).

    Nick Bostrom has guessed a simulation of the entire universe could be done in 10^42 operations (Bostrom Simulation argument, University of Oxford).

    That possibility alone would guarantee reversing the ice cube. I’m concerned that people would abandon cryonics after reading QA dismissing the safety that freezing brings and as your suspended friend Bob Ettinger warned ‘It hasn’t been done yet.” I think its possible both areas must advance jointly and recovery techniques will overlap a great deal.

    Much depends on what computing power we can stoke up.

    I’m bias because of writing the site on it, but proof of concept is already here with facial reconstructions.

    By far the great strategy is not to die in the first place: this gives me a say in how the world emerges and some little influence, even to lobby to start groups and adopt new technology as it comes. etc.

    I attended UK Alcor meetings in the 1990’s briefly and it helped form my views, but I couldn’t see how it would be achieved then, only to see it much later.



  • Giulio Prisco

    Good to see you here eldras!

    Nobody needs a perfect replica of the original ice cube, just one good enough to feel like the original ice cube, whatever it is that ice cubes feel.

  • Mike Perry

    Giulio says, “@Mike re ‘I posit that in parallel universes that are different overall, nevertheless, there are copies of you that are exact down to the bit level.’

    “I don’t think I understand this point. Surely in some parallel universes that are _very_ different overall, for example in a universe where life never developed on Earth, there cannot be an exact copy of me.”

    My answer: I am not arguing that *every* universe must have a functioning being that is identical to you, just that *some* universes will(infinitely many, in fact, or infinitely many regions that in effect can be considered different universes.)

    Giulio says, “Nobody needs a perfect replica of the original ice cube, just one good enough to feel like the original ice cube, whatever it is that ice cubes feel.”

    My answer: to resurrect a person you *do* need a pretty good copy of their brain (or representation, informationally speaking). Making them “feel like the original” would be, I think, a challenge comparable to unmelting an ice cube. It’s unlikely you could get this much information directly, I think, without invoking something like my MA idea.

    “Something like”–maybe what we really need is a new term, that would emphasize that both MA and QA are variants of some basic strategy. If one doesn’t work, the other might. What would be a good new term? Would “information physics,” already in use, fit the bill?

  • I think you need a copy good enough for humans to be faithful to what they were which is thought to be 5 nanometres.

    The issue I see is how you compensate for information loss.

    For example is you have this architecture of a man’s brain in 1066 and there’s a bit missing, there are loads of things that it couldn’t be.
    The loads of things that it could be dont exist in a vacuum but will have causal effects on the rest of the brain, and on his actions as the results of his thinking.

    Retrodicting you might be able to eliminate most of the possibilities until you are left with a few alternate contructions: but I have to doubt this could happen as if a module pattern doesn’t affect anything else it can’t exist as everything is interlinked.

    The truly vast cross-referencing we’re going to be able to do in just 20 years is almost unimaginable.

    Another thing is that brains/bodies dont exist in any possible cell combination. They are constructed by laws. We can configure Neanderthal basic DNA good enough to debate resurrecting their species (Harvard this month) and this will be a building block prototype to begin fleshing individual brains as quantum archaeology grid is constructed

    This is not fantasy: components of what will become it are already being assembled in other databases like the biological record.

    So the question of how specific we can do archaeology is answered by everything is related. There are zillions of time lines all leading causally, inevitably, to each required event.

    But I dont KNOW QA will work.

  • (I’ll put your Arthur C Clarke quote above on my site if that OK, as it;s really relevant).

  • Darren Reynolds

    What is dying?

    Taken from a panpsychist viewpoint – I assume it’s OK to start there? – it is impossible for consciousness to be completely lost. But it can still change radically in nature. Consciousness appears to us to have certain features that have a scale from minimal to maximal, and death – whatever it is – appears to drive consciousness down that scale.

    When we are alive, some 60Kg of stuff is bound up in each of us. Some 1.4Kg of that is bound up in a brain that somehow manages to deliver a combined, unified conscious experience. Each of us, most of the time, has a sense of being exactly one person. We can’t see through each others’ eyes, and when we see through our own eyes, we have a single view of the world.

    Not everyone is like that. Tatiana and Krista Hogan are two five year-old Vancouver girls who quite verifiably can, to some extent, see through each others’ eyes. Because they are still young it is still unclear exactly what ability they have or to get any real understanding of what it is like for them. Yet, one girl, with her eyes covered, can tell you which toy is being shown to her sister.

    Going in the opposite direction to Tatiana and Krista, there are individuals whose brains produce more than one view of the world. Those who suffer with dissociative identity disorder alternately become, to any reasonable observer, different people. Those who suffer the unrelated condition (or possibly, a range of conditions), schizophrenia, appear to have difficulty joining up all the aspects of their perception into a single, unified whole.

    Is it possible that the worst aspect of death is the complete dissociation of the mind into fragments? Is it the loss of binding – binding that enables the self-perception of an individual on the scale of the human being? And together with that loss of binding, is it the loss of the organisational features that enable the creation of memories (in the broadest sense to include skills, etc.) and the retrieval of those memories in determining a course of action?

    It seems a reasonable conjecture.

    If that is what death is, then the opposite of death is the opposite of that description. The opposite of death is the creation of that organisation. It is the joining together of matter so as to create an ever more-unified experience, drawing on an ever greater archive of memories. You know, sometimes transhumanists talk about uploading minds and the sorts of features a virtual world would have, to make the world a better place for those minds to live in. Even the august article here talks about resurrecting individuals, even multiple times simultaneously.

    ‘Resurrection’ is described as if it is the re-animation of human beings to either a real or facsimile state of being similar to what we might today call the norm.

    Surely that misses the point?

    It seems far from clear to me that any individual will want to exist independently, once the capability is there to share in a single mind that stretches out into the cosmos.

    The opposite of death is not the restoration of a being to a living state. The opposite of death is to take an existing, living being, in the opposite direction to death.

    That is, to take a being up the scale to ever greater life.

  • Giulio Prisco

    Thanks for joining Darren!

    I have more than 60kg of stuff, I hope my spiritual/astral body will be slimmer.

    Tatiana and Krista Hogan are joined, so I don’t find it too surprising that there are neural pathways from the eyes of one to the brain of the other. Now, if they were separated and still had cross-vision ability… wow!

    • Darren Reynolds

      “Tatiana and Krista Hogan are joined… if they were separated and still had cross-vision ability… wow!”

      Yes, they share a thalamic bridge and it seems highly likely that is where the communications go, according to their neuroscientist doctors. I hope you can see the point? Whether two brains are connected by a thalamic bridge or radio transmitters, if you can share part or even whole of your mind, so that, in effect, one mind has two bodies, then something quite special has happened. Does you see how their story has great communicative potential?

  • eldras

    hi Daren,

    There is more unknown than knowable IMO.

    I had to learn to ruthlessly conform to the scientific method (I often miss it, and I hear from great scientists it is extremely hard).

    I spent a lot of time searching weird reports to find meaning in the world
    I dunno what your background is?

  • eldras


    re Turing Church

    I have worked on a pre-page on Quantum Archaeology and it’d be great if someone read it out as I’m a bit of a recluse?



  • Giulio Prisco

    Thanks eldras, very cool page. Re Turing Church workshop, instead of a single workshop with many speakers, we will do many interactive talks with one or two speakers and a small active audience.

    The reason is that finding a date that suits the agenda of many speakers is difficult and introduces important delays. The videos are posted online very soon after the shows, so those who cannot attend interactively can watch the videos.

    When do you wish to schedule your talk? Best is on a Saturday or Sunday at 6pm UK time, which is convenient for Europeans and US participants on both coasts.

  • eldras

    6pm Saturday 22nd dec?

    Oxford conference on AGI-12 will be over then I think.

    If QA is seen as viable the EU will certainly plough billions into it!