The Obligatory Mind Uploading Blockchain Crossover
Keith B. Wiley
2017-09-24 00:00:00

A classic thought experiment in the philosophy of mind is reduplication, in which a person (or her mind) is duplicated such that two or more descendant people of shared mental ancestry now exist where previously there was one.  The philosophical quandary is to resolve what happened to the original person’s identity.  Did she survive and if so, in which of the resulting people's minds?  Which of the two resulting people is the original and which is a mere copy of denigrated identity status?  Alternatively, is there something fundamentally wrong with the wording of such questions, such that we should we adopt a different perspective on the nature of personal identity that offers alternative solutions to the reduplication quandary?  Reduplication further arises not only in abstract philosophical musings, but also in the futuristic and variously conceivable (depending on the reader's tastes), technology of mind uploading, in which a person's physical brain is emulated via the technology of whole brain emulation.  While mind uploading might produce a single result, such as if the original brain is destroyed by the uploading process and only one upload is created, we can also conceive of either nondestructive scenarios (in which the original brain is not destroyed) or scenarios that produce multiple uploads.  Either case results in multiple descendant minds, each operating in distinct physical systems (brains or cloned brains, or computers of some sort).  The philosophy of personal identity has produced several possible stances on the nature of personal identity.  The most popular are body identity and psychological identity, with other options including closest continuer identity, space-time worm identity and branching identity.  However, there is always room for new theories to enter the discussion.  The way in which blockchains work, and Bitcoin's mining process and protocol for handling orphaned blocks, suggests a new theory of identity along with a new solution to the reduplication problem.  The proposed blockchain solution to personal identity has applications to the handling of the reduplication problem as it may arise during a futuristic mind uploading procedure.



A blockchain holds a hashed transaction ledger, essentially the history of all transactions encoded to prevent any subsequent alteration of the history.  In this way, all transactions back to the beginning of the ledger's history can be confirmed by any interested party.  Deceit, fraud, and other attempts to undermine the history simply don't work, and consequently blockchains enable a variety of interactions with the currently most popular being digital currency.  In addition to more conventional applications, blockchains could also be used to assign identity status (original or copy) to the descendent minds of a mind uploading procedure.  Each descendant could then venture out into the world confident that their identity status will be honored by all third parties thereafter.  Let us call this the blockchain theory of personal identity. Blockchain identity relies on the unpredictable mechanics of blockchain mining to first randomly assign identity status, and second, record the designated identity status in the permanent and unalterable ledger for posterity.



Alice will be our guinea pig.  She submits to a mind uploading procedure, but knows in advance that she does not subscribe to any of the conventional theories of personal identity.  Alice is not a body identist for example, primarily because she is a physicalist, materialist, and functionalist, and therefore holds that the mind is purely the product of the rote physical forces and events within the brain and not a property of any particular set of atoms (she believes that any given atom in her brain could be substituted with any other identical atom without harming her neural function, mental experience, or personal identity). Furthermore, she is well versed in the paradox of the split hemispheres, introduced by Derek Parfit, in which two descendant minds receive equal portions of the original body, thereby preventing the assignment of identity according to body identity principles.  Due to her objection to body identity, she grants no identity primacy to the biological person preceding the mind uploading procedure.  She would just as willingly grant primacy to the computerized mind that the procedure produces.



While Alice believes that memories and personality are the critical indicators of personal identity, she is not a straightforward psychology identist (identity via psychological state such as memories), and surely not a psychological continuity identist (in which the continuity of conscious experience indicates identity) because she is well aware of the paradoxes and conundrums that simpler versions of psychological identity and associated continuity inspire, such as Thomas Reid’s brave officer thought experiment, in which a middle-aged person might be coidentified with his younger or older self due to overlapping memories, while his older self cannot be coidentified with his younger self due to a total loss of childhood memories.  Alice has also studied the problems facing psychological continuity identity, such as resolving the apparent threshold of loss of consciousness below which identity is arbitrarily deemed unpreserved.  In decreasing order of continuity, we willingly grant identity preservation upon awakening each morning, when recovering after fainting, after undergoing modern anesthesia, and even in a few incredible and rare medical cases of recovery from rapid frigid drowning.  Yet psychological continuity identity implies that there is some threshold of continuity below which an awakened person should properly be branded by society with the identity status of a doppelganger arisen in the place of a former person who has otherwise died. This notion is so alien to our sense of personal identity (and alien to how we would actually treat one another in such a scenario) that no one honestly takes it seriously upon close scrutiny.  Alice knows all of this, so she rejects psychological and/or continuity identity.



Alice’s rejection of body identity and psychological continuity identity leads her to conclude that she is also not a space-time worm identist, which claims that identity must follow a contiguous path through the world and time.  Space-time identity imposes a presumed, and seemingly unnecessary and arbitrary, spatiotemporal property on the immaterial metaphysical mind and its associated personal identity. Since Alice sees the mind as entirely nonphysical, she finds space-time worm identity to be uncomfortably close to body identity in its conflation of physical constraints on the brain with unnecessary related constraints on the mind. The two concepts are not the same thing (for why else have two words to refer to the two concepts, brain and mind?) and therefore should not necessarily be subject to the exact same set of constraints.  It is not clear to her what forces or other properties of the universe or of reality would constrain the mind to contiguous spatial locations in the first place.  Such a constraint violates her notion of Occam’s Razor since it imposes additional and seemingly unnecessary and unmotivated constraints on a system otherwise just as easily comprehended with fewer such constraints.  She also finds the requirement for contiguous time problematic for the same reason that she rejects psychological continuity identity.  She sees no fundamental principle of the mind that would preclude it from stopping and starting again since it is already subject to a wide array of partial such disruptions in function that have no apparent effect on personal identity.  For these reasons, Alice cannot be a space-time worm identist.  This conclusion enables her to permit identity to reassociate between various physical systems, such as her brain prior to uploading and the computer into which she uploads.



Alice is also not a closest continuer identist, and so places no primacy on the mind descendent who by some external metric, such as spatial proximity or material composition, is the most similar to the single mind ancestor who enters into the uploading procedure.  She recognizes the problematic conclusions that closest continuer identity prescribes, such as hinging one person's identity status on the presence or absence of another seemingly unrelated person.  She is aware that in a destructive teleportation scenario from Earth to Mars (i.e., teleportation that destroys the original body), closest continuer identity grants the original identity to the person reconstructed on Mars since they are definitively the most similar to the original person (they are, in fact, the only resulting person in the destructive scenario and therefore preserve identity by default according to closest continuer identity).  But closest continuer identity further claims that in a nondestructive teleportation scenario, the person on Mars in fact does not preserve the original identity since the person on Earth is by certain measures the more similar of the two descendents to the ancestor who entered the teleporter.  These competing conclusions are irreconcilable to Alice. There is also the peculiarity that the person on Mars won’t even necessarily know their own identity status if they don't know whether the person on Earth was destroyed.  Short of psychosis, which is irrelevant to this discussion since our subject is presumed sane, how can someone not know their own identity?  And what happens if the original body is destroyed, but only shortly after the Martian body is constructed, or perhaps steadily deconstructed concurrently as the new body is steadily constructed?  What role does the speed of light play here since it takes several minutes for the teleporter transmission to reach Mars? Closest continuer identity is all over the map (or solar system) on this issue.  Alice consequently finds closest continuer identity too problematic to take seriously.



Lastly, Alice also rejects branching identity (putting her in stark disagreement with the author of this article incidentally), and so cannot tolerate an outcome in which both descendants receive equal primacy in their respective claims to the original identity.  Alice's values and beliefs require that, at most, a single descendent mind receive the identity status of the original.  The author would advise Alice to reconsider branching identity (it is the central topic of his book, if the reader is curious), but for the purpose of exploring the implications of blockchain identity, let us assume Alice rejects all currently available theories of personal identity.



Alice decides that she is, in fact, an arbitrary identist.  This is yet another possible theory of personal identity in which during reduplication, personal identity randomly or arbitrarily follows one mind descendant, thereby relegating the other mind to the secondary status of mere copy, to never be considered a true continuation or preservation of the original single identity.  Arbitrary identity raises a big question: how is identity arbitrarily assigned to the resulting people?  Is reality constructed such that identity randomly follows one descendent or the other (perhaps invoking aspects of quantum mechanics and its somewhat abused penchant for randomness)?  If identity is truly randomly associated in a reduplication scenario, then how could anyone actually determine the result?  How would we know which descendent mind had received the original identity and which held some newly born identity?  Since Alice rejects body identity, she can't solve the conundrum by asking which, if any, is the original body.  One solution is to honor the requirement of arbitrary and random identity status assignment, but to take control of the arbitrary process so we can observe and record the random outcome.  Blockchain identity is one way to do this.



We need a way to label the multiple descendant minds of a mind uploading procedure.  We will refer to the two descendants of Alice’s mind uploading procedure as Barbara and Betty, explicitly without making any designation as to which is the biological original and which is the upload (or perhaps two uploads of a destructive procedure).



Before considering the blockchain solution to arbitrary personal identity, let's consider the simplest such solution: a single coin toss.  Alice undergoes mind uploading, thereby producing Barbara and Betty.  A coin is tossed with heads assigned to Barbara.  If the coin lands heads up, then Barbara is granted primacy of identity and goes on to live her life as a continuation of Alice's identity, both in terms of societal recognition and presumption of true metaphysical identity, and also in terms of lawful and contractual obligations.  Betty will then go out to make her own way in the world too, a free and new person (perhaps with half of Alice's finances, assuming Alice made such preparations in advance).  Betty will, of course, feel very strongly that she is the true continuation of Alice and that something dreadfully wrong has occurred during the coin toss, but since Alice rejected branching identity in advance of the mind uploading procedure, and since the memory of that commitment carried forward into the values and beliefs of both Barbara and Betty, we can hope that Betty will honor the outcome of arbitrary identity.  Betty will accept her fate as a mere copy for the rest of her days, attempting to reject the deep inner conflict posed by her conviction that she really feels like the one true Alice.  She may need substantial therapy.  Hopefully Barbara will sympathize and assist in this regard.  Of course, if the coin lands tails up, then the opposite result is enacted. The problem with this solution is enabling each person to prove her respective identity status in future interactions.  A formal registry is required, saying that Barbara is the continuation of Alice and Betty is just a copy.  Such registries concentrate record-keeping and power that may be subject to error, data corruption or loss, or even outright fraud.  They also create single points of failure as targets of such error or fraud. These problems are precisely what blockchains aim to alleviate.



The blockchain solution solves the logistical problems associated with the coin toss solution.  We first produce two possible reports, one for each possible outcome: Barbara-Wins and Betty-Wins.  Equal numbers of copies of both reports are then distributed to the blockchain cloud and its block miners.  Each miner incorporates their received report into the block they are currently mining, so various miners hold one or the other of the possible reports, but no miner holds both reports.  The first report associated with a block that is successfully hash-mined (i.e., randomly seeking a hash of a block of data that satisfies a proof-of-work requirement, such as a hash with a prefix consisting of a certain number of zeros) will be rapidly distributed throughout the blockchain network, thereby locking the resulting identity status into the distributed hashed ledger, unalterable for all time.  Either Barbara or Betty will win, but due to the unpredictable nature of hash-mining, there is no practical way to predict or influence the outcome in advance (since some miners are more powerful than others, it may be necessary to distribute the reports not randomly, but rather according to a fair balance of total distributed mining power).  In the event that two miners with differing reports successfully mine their associated blocks at the same time, the blockchain protocol is already designed such that only one of the mined blocks and its contained reports will eventually win over the distributed network.  In this way, arbitrary identity is enacted, identity status is assigned to Barbara and Betty, and the result is recorded unalterably by a distributed network that is resistant to both error and fraud.



As a brief addendum, let's consider the matter of societal and legal identification: how does anyone interacting with Barbara or Betty know if that person is the surviving original or the lowly copy?  How does each of them prove they are who they claim to be?  One simple solution is to generate two public-private encryption key pairs.  Attach the public keys to the reports, assigning one public key to each person.  The resulting blockchain will then contain in its ledger not only the designation of identity status of Barbara and Betty, but also their respective public keys.  They will each only hold their own private key of course, and therefore can confirm their identity whenever required by simply decoding any brief message encoded with the relevant public key.  In this way, not only does the blockchain encode identity status, but also the means to prove one's own identity in any future interaction.



Of course, as alluded to above, the author strongly recommends branching personal identity, which obviates the need for blockchain identity.  But since so many people are strongly resistant to branching identity (for reasons the author believes should be vigorously reconsidered), perhaps there is still room to consider solutions such as blockchain personal identity for those readers who cannot accommodate the other available models of personal identity.



Keith Wiley has written numerous articles on mind uploading and its implications for personal identity.  His book, A Taxonomy and Metaphysics of Mind-Uploading, is available on Amazon.  His personal website is http://keithwiley.com .



References:

https://www.amazon.com/Taxonomy-Metaphysics-Mind-Uploading-Keith-Wiley/dp/0692279849/

http://www.npr.org/sections/thetwo-way/2017/08/14/543417269/bitcoin-streaks-to-4-300-mark-continuing-meteoric-rise

http://www.huffingtonpost.com/ameer-rosic-/5-blockchain-applications_b_13279010.html

https://blockgeeks.com/guides/blockchain-applications/

http://www.nytimes.com/1988/07/26/science/the-doctor-s-world-ingenuity-and-a-miraculous-revival.html

https://www.carboncopies.org/

https://www.amazon.com/Reasons-Persons-Derek-Parfit/dp/019824908X