Thought experiments as knowledge

One of the most remarkable aspects about the ongoing debates concerning the technical feasibility of mind uploading is the excessive confidence that some people have that these issues can be resolved without further experimental validation. The (implicit) assumption seems to be that our current understanding of the neuroscience of consciousness is sufficient to demonstrate the technical feasibility of mind uploading by logical deduction from these findings alone. This is a mysterious claim for at least two reasons. The most fundamental reason is that the scientific study of consciousness has not nearly evolved to a stage that allows for making bold claims about the subject, let alone its far-reaching consequences. The other reason is that, in the absence of empirical examples of substrate-independent life in general, it cannot be argued that such logical arguments are just innocent or inescapable conclusions from what we already do know.

It should not be surprising that such arguments fail to convince some of the participants in the debate. After all, many “mind uploaders” also believe that the case for cryonics is just a straightforward exercise of Pascal’s Wager and the technical feasibility of molecular nanotechnology can be settled by arguing that the idea does not “contradict the laws of physics.” As I have argued in a more detailed article about this tendency, the common denominator in all of  this is the excessive role that is assigned to logical arguments (or rationalism). But there is an important difference between, let’s say, predicting the unmeasured viscosity of a specific aqueous solution from a formula that has been derived from numerous measurements on the one hand and drawing far-reaching conclusions from general scientific observations or even philosophical premises (materialism, reductionism) on the other hand.  This does not mean that one should completely refrain from speculation about future technologies, but it should induce a habit of having less confidence in your conclusions as the chain of assumptions and logical arguments lengthens, let alone if your conclusions are highly controversial.

One could object that since advocates of mind uploading are generally strong advocates of cryonics, that even treating their arguments with skepticism risks alienating prospective supporters of cryonics. The fact of the matter is, however, that presenting cryonics as just one element in a larger futurist framework strongly weakens the point that cryonics is an experimental medical procedure, not an ideology or life-style. It is not possible to present cryonics in a fashion that does not alienate anyone at all. But presenting cryonics as an experimental medical procedure without additional ideological, philosophical, or sociological add-ons  has the important merit of reducing this amount of alienation to the greatest possible degree. It also has the distinct advantage that it facilitates the recruitment of people who can move the field forward; experimental scientists and medical professionals.

Mind uploading, falsifiability and cryonics

On the cryonics discussion list Cryonet cryobiologist Brian Wowk weighed in on the topic of mind uploading in a post that merits quoting in its entirety:

I read with interest Bob Ettinger’s recent remarks about Mark Gubrud’s piece in The New Atlantis.

http://futurisms.thenewatlantis.com/2010/06/why-transhumanism-wont-work.html

Although I have not been around as long Bob, I have nevertheless observed arguments about uploading, identity duplication, and related subjects for decades.  In all that time there are two things I’ve never seen: (a) A truly new argument, and (b) Someone change their mind.  What is seen are people who passionately believe they are correct, and who believe that they have just the argument to finally convince the other side that they are right.  They never do.

I have come to believe that the question of whether a computationally equivalent duplicate of a human mind (assuming equivalence in this context is even definable) constitutes a continuation of the original person may be objectively unanswerable.  It’s almost a matter of taste, like alternative interpretations of quantum mechanics that assume different underlying realities that give exactly the same measurable results.

Eventually the distant day will come when the computational processes of a human brain are duplicated in an electronic computer, or even in another identical organic substrate.  When that day comes, we can be certain of this: If the person who was “duplicated” believed before duplication that duplication constitutes survival of the self, then, by definition, the duplicated entity will insist vociferously that indeed they did survive.  This has ethical implications.  Conversely,
an entity derived from a person who did not believe in this form of survival might be quite unhappy to be told that they were the product of a destructive scan of somebody.  This too has ethical implications.

Philosophical truth aside, evolution selects against humans who spend time worrying about whether sleep, anesthesia, or biostasis endangers personal identity.  Similarly, it is easy to predict which side of the uploading and duplication debates will win in the long term.  There is no entity more invulnerable or fecund than one that believes it consists of information.

Recent discussions of the topic of mind uploading on the Cryonics Institute members mailing-list contradict Wowk’s claim about people changing their mind about mind uploading. Robert Ettinger posted an itemized list with objections against the idea of mind uploading as a strategy for personal survival and I weighed in on the (current) lack of experimental evidence to settle the matter. The effect was that some people changed their mind or became more agnostic about mind uploading.

Wowk may be correct that the question whether a “computationally equivalent duplicate of a human mind…constitutes a continuation of the original person may be objectively unanswerable.” The discussion about mind uploading and persistence of the original person has distinct similarities with discussions about solipsism, consciousness, and the existence of the external world. It is not inconceivable that in a world where mind uploading has become routine the debates will still continue because the hard problem of persistence of the person is not falsifiable in a meaningful manner.

There are mind uploaders and there are Mind Uploaders. The Mind Uploaders are a small but vocal minority who display little patience for the argument that the technical feasibility of mind uploading requires empirical verification and cannot be completely settled by logical deduction or thought experiments. As cryonics activist and ex-Alcor Board member David Pizer says, “Having existed with Uploading Lovers for many years now, I believe they are as firmly entrenched in their beliefs as  traditional religious persons believe that their souls are going to Heaven after death here on Earth.”

Cryonics is often associated with ideas like mind uploading and transhumanism. One negative consequence of this (un)intentional association is that some people who are considering cryonics feel that they have to embrace a much larger set of controversial ideas than what they are actually being asked to consider. As a result, there is a real risk that people reject cryonics for reasons that have little to do with the proposal of cryonics itself. Advocates of cryonics do not do themselves a favor by promoting the idea of human cryopreservation as part of a larger set of futurist ideas instead of just promoting cryonics as an experimental medical procedure to extend life. There is too much at stake to alienate people by piling more controversial ideas on top of what is already considered to be a radical idea. Such a low-key attitude will also produce a more consistent message because it extends the element of uncertainty that is inherent in cryonics to other areas of life as well.


David J. Chalmers on the Singularity, mind uploading and cryonics

If I would make an argument in favor of mind uploading (or substrate independent minds) it would not be a logical deduction from what we know about neuroscience but from what we don’t know.  As one of the leading philosophers of mind David J. Chalmers has argued in this insightful paper about the Singularity and mind uploading:

Can an upload be conscious? The issue here is complicated by the fact that our understanding of consciousness is so poor. No-one knows just why or how brain processes give rise to consciousness. Neuroscience is gradually discovering various neural correlates of consciousness, but this research program largely takes the existence of consciousness for granted. There is nothing even approaching an orthodox theory of why there is consciousness in the first place. Correspondingly, there is nothing even approaching an orthodox theory of what sorts of systems can be conscious and what systems cannot be….

It is true that we have no idea how a nonbiological system, such as a silicon computational system, could be conscious. But the fact is that we also have no idea how a biological system, such as a neural system, could be conscious. The gap is just as wide in both cases. And we do not know of any principled di differences between biological and nonbiological systems that suggest that the former can be conscious and the latter cannot. In the absence of such principled di differences, I think the default attitude should be that both biological and nonbiological systems can be conscious

One can argue with this derivation of what the “default position” should be, but his more skeptical approach has a degree of modesty in its favor that is often lacking in transhumanist circles.

David J. Chalmers also discusses cryonics in a favorable context:

Cryonic technology off ers the possibility of preserving our brains in a low-temperature state shortly after death, until such time as the technology is available to reactivate the brain or perhaps to upload the information in it. Of course much information may be lost in death, and at the moment, we do not know whether cryonics preserves information sufficient to reactivate or reconstruct anything akin to a functional isomorph of the original. But one can at least hope that after an intelligence explosion, extraordinary technology might be possible here

On his blog he also writes that “for the last couple of weeks I have been in Oxford giving the John Locke Lectures on Constructing the World.  The title is an homage to Rudolf Carnap’s 1928 book Der Logische Aufbau Der Welt. The lectures are based on a book I have been writing for the last couple of years, trying to execute a project that is reminiscent of Carnap’s in certain respects.”

A person who discusses mind uploading in a meaningful context, gives cryonics a fair hearing, and has a work in progress that is inspired by Rudolf Carnap’s The Logical Structure of the World should not be ignored, let alone be ridiculed.

Free will versus determinism as it relates to cryonics

Excerpt from “Ben Best – A Case for Free Will AND Determinism”

Determinism implies materialism — implies that consciousness is material. Cryonics is based on the premise that the preservation of the fine structure of the brain at low temperature will preserve the self — ie, that the self is entirely determined-by and contained-in the physical brain. Determinism would imply that preservation of the material basis of mind/self is theoretically possible. (For an exploration of how the self is encoded in the brain, see my series The Anatomical Basis of Mind. Development of the anatomical argument to explain the functioning of mind is best summarized in Chapter 8, Neurophysiology and Mental Function.)

Defenders of “free will” who say that the self has a spiritual basis independent of the brain often reject cryonics as being unnecessary. There are a few “spiritually” oriented people (like the Fyodorovians) who think that “resurrection of the body” is essential due to an intimate connection between the body and the “soul”, but these are in the minority. The majority of cryonicists do not accept spiritual beliefs, but there are notable exceptions, namely people who regard cryonics as a form of medicine. If cryonics can extend life, it is no more an affront to spiritual belief than other life-extending practices such as exercise and the avoidance of tobacco.

What about anti-determinist materialists who believe in “free will”? Those, like Roger Penrose, who claim that the mind is ultimately rooted in quantum uncertainty might not accept the possibility of biostasis, but Penrose has made no explicit statement about this subject. Penrose writes of the non-computability of mind, but acknowledges that non-predictability does not equate with “free will”.

Predictability is really at the heart of what is required for cryonics. If the mechanical operation of billions of neurons and trillions of synapses result in the phenomena known as the mind, the Self and the Will, then preservation & restoration of this machinery by cryonicists & nanotechnologists is possible in principle. But this also means that human beings are machines whose future actions are, in principle, entirely predictable. The positive side of this is that understanding the machinery in sufficient detail could provide the basis for reconstructing those aspects of the mind (parts of the brain) that were destroyed beyond recognition or repair. The negative side is that many people find it “dehumanizing” to believe that we are nothing but machines.

The proposition that the self/mind has a complete material basis in the mind has practical implications for cryonics, but also raised baffling questions. If it is possible to use a cryopreserved brain as a template for atom-by-atom reconstruction of a new brain, the identity of the person whose brain was cryopreserved would presumably be restored. But if such reconstruction could be done once, there is no reason why it could not be done hundreds of times. Would each reconstruction have the same personal identity (the same self) as the original? (For more detail on this question, see my essay The Duplicates Paradox).

Chemical preservation and cryonics research

In the 2009-4 issue of Alcor’s Cryonics magazine I review the technical and practical feasibility of chemical preservation. One of the most interesting aspects of chemopreservation is that it could play a useful role in the cryopreservation of ischemic patients.

There is accumulating evidence that vitrification agents cannot prevent ice formation in ischemic patients. This raises the question whether some cryonics patients could benefit from chemical fixation prior to transport and cryoprotective perfusion.

Such protocols raise a number of obvious concerns but the question is not so much whether these procedures are inferior to vitrification of non-ischemic patients, but whether fixatives can improve the situation of some ischemic patients compared to the prospect of substantial ice formation, or even straight freezing (cooling without cryoprotection). This is an empirical question which needs to be settled by experimental research.

Chemopreservation: The Good, The Bad and the Ugly

Down with uploading

Over the last couple of years, cryonics pioneer Robert Ettinger has been a vocal critic of simplistic defenses of the idea of mind uploading as a survival strategy. He has worked out his reservations in detail in his latest book Youniverse: Toward a Self-Centered Philosophy of Immortalism and Cryonics. In a recent CryoNet message he reiterates some of his basic arguments:

“Identity of indiscernibles” is a  common tenet. Often attributed to Leibniz, one  version is that if two physical objects or systems cannot be distinguished from  each other by any criterion, then they  must be considered the “same” or  identical. First, this assertion actually asserts nothing except a  certain preference in use  of language. It has no consequences. It is also useless because if the question arises, are A and B distinguishable, the answer  is always yes.

It is hard to see how anyone can claim complete certainty  on the topic of mind uploading. Nevertheless, to some of its more dogmatic advocates the case for mind uploading is simply an exercise in deductive reasoning. There are major objections to such an attitude. The most obvious point is general; why should mind uploading be an exception to the rule that we can have no certain knowledge? One might object that absolute certainty is possible in logic. But in that case one would need to defend the thesis that the feasibility of mind uploading (and its associated views about identity) is a purely logical matter and exempt from empirical testing. This is not a credible position.

This does not mean that questions about identity will be easily answered when such technologies are available. For all we know, mind uploading will be technically feasible and the debates about identity continue.  There is a lot of merit to discussions about mind uploading and identity, especially for those interested in cryonics and life extension. But there is also a lot to say for being modest in making bold claims before such technologies have materialized.

Brain preservation

Mind uploading advocate Kenneth Hayworth has launched an interesting website devoted to the science of brain preservation. Of particular interest is his Proposal for a Brain Preservation Technology Prize (PDF). This document includes one of the most comprehensive discussions of chemopreservation as a strategy for personal survival. For example, one of the most common objections to chemopreservation is that fixatives like formaldehyde and glutaraldehyde do a poor job of fixing lipids. In this document, Hayworth reviews a number of papers where a fixative that can stabilize lipids, osmium tetroxide, is perfused (!) through the circulatory system.   For human sized brains such a step would be necessary to avoid the ischemic damage and autolysis that would occur in the case of the time-consuming alternative of diffusion fixation.  He also speculates that such a fixed brain can be perfused with a high viscosity plastic resin for long term preservation.

One of the limitations of this approach, as the author concedes, is that the procedure needs to be started before death. In reality, the situation is even more challenging  than that because the procedure would have to be started before ischemia-induced brain perfusion abnormalities associated with terminal disease and the agonal phase will manifest themselves. This is a problem where “old fashioned” cryonics has a clear advantage. Perfusion impairment may interfere with complete distribution and equilibration of the cryoprotectant in the brain but the unperfused tissues will still be stabilized (although in a damaged form) through low temperatures. In the case of chemical fixation such a “second chance” is absent. This is not just a theoretical problem. Cryonics researchers have become painfully aware of the adverse effects of even the slightest perfusion artifacts on the quality of fixation and the resulting electron micrographs.

As a consequence, this kind of “high quality” chemopreservation can only be a credible alternative for cryonics if the medical establishment would permit the procedure for those who are diagnosed as terminally ill. If the acceptance of cryonics is any guidance, there is little chance that this will happen any time soon.

Chemopreservation has another major obstacle to deal with. As the cryobiologist Brain Wowk has stated on numerous occasions, chemical fixation is a dead end in terms of reversibility with contemporary technologies. This aspect of chemical fixation limits the demonstration of its technical feasibility to a demonstration of ultrastructural preservation.  In the case of cryonics, evidence of excellent ultrastructural preservation has produced little excitement among the scientific establishment and the general public. Linking chemopreservation exclusively to mind uploading may present another obstacle to its acceptance.

In his essay Killed by Bad Philosophy: Why brain preservation followed by mind uploading is a cure for death [PDF] Kenneth Hayworth attempts a defense of mind uploading by identifying the philosophical errors that those who reject the concept, and those who argue that “a copy is not you” in particular, engage in. The author shows little doubt about his position although one might object that the central example that is used to make the case could also be used to  argue against mind uploading. One might even object  that the whole debate involves a pseudo-problem if any kind of empirical observation can be made consistent with the case for and the case against mind uploading.

Aside from these complexities, this is an admirable effort to raises interest in high quality brain fixation. Initial funding for more experimental research should be encouraged.

Revival of cryonics patients literature

There is a growing literature that discusses the technical aspects of revival of cryonics patients. The following list of the published literature was compiled by Ralph Merkle and Robert Freitas and published as an appendix of their article on molecular nanotechnology in Cryonics Magazine 2008-4:

Robert C.W. Ettinger, The Prospect of Immortality, Doubleday, NY, 1964

Jerome B. White, “Viral Induced Repair of Damaged Neurons with Preservation of Long-Term Information Content,” Second Annual Cryonics Conference, Ann Arbor MI, 11 April 1969

Michael G. Darwin, “The Anabolocyte:  A Biological Approach to Repairing Cryoinjury,” Life Extension Magazine (July-August 1977):80-83

Thomas Donaldson, “How Will They Bring Us Back, 200 Years From Now?” The Immortalist 12 (March 1981):5-10

K. Eric Drexler, Engines of Creation:  The Coming Era of Nanotechnology, Anchor Press/Doubleday, New York, 1986, pp. 133-138

Brian Wowk, “Cell Repair Technology,” Cryonics 9(July 1988)

Mike Darwin, “Resuscitation: A Speculative Scenario for Recovery,” Cryonics 9(July 1988):33-37

Thomas Donaldson, “24th Century Medicine,” Analog 108(September 1988):64-80 and Cryonics 9(December 1988)

Ralph C. Merkle, “Molecular Repair of the Brain,” Cryonics 10(October 1989):21-44

Gregory M. Fahy, “Molecular Repair Of The Brain: A Scientific Critique, with a Response from Dr. Merkle,” Cryonics 12(February 1991):8-11 & Cryonics 12(May 1991);  “Appendix B. A ‘Realistic’ Scenario for Nanotechnological Repair of the Frozen Human Brain,” in Brian Wowk, Michael Darwin, eds., Cryonics: Reaching for Tommorow, Alcor Life Extension Foundation, 1991

Ralph C. Merkle, “The Technical Feasibility of Cryonics,” Medical Hypotheses 39(1992):6-16

Ralph C. Merkle, “The Molecular Repair of the Brain,” Cryonics 15(January 1994):16-31 (Part I) & Cryonics 15(April 1994):20-32 (Part II)

Ralph C. Merkle, “Cryonics, Cryptography, and Maximum Likelihood Estimation,” First Extropy Institute Conference, Sunnyvale CA, 1994

Ralph Merkle, “Algorithmic Feasibility of Molecular Repair of the Brain,” Cryonics 16(First Quarter 1995):15-16

Michael V. Soloviev, “SCRAM Reanimation,” Cryonics 17(First Quarter 1996):16-18

Mikhail V. Soloviev, “A Cell Repair Algorithm,” Cryonics 19(First Quarter 1998):22-27

Robert A. Freitas Jr., “Section 10.5 Temperature Effects on Medical Nanorobots,” in Nanomedicine, Volume I: Basic Capabilities, Landes Bioscience, Georgetown, TX, 1999, pp. 372-375

Ralph C. Merkle, Robert A. Freitas Jr., “A Cryopreservation Revival Scenario using MNT,” Cryonics 30(Fourth Quarter 2008).

Robert White on brain death

Robert J. White is most known, or perhaps most notorious, for his work on primate head transplants. Less known, but more relevant to the practice of human cryopreservation, is his work in cerebral ischemia, hypothermia, and brain preservation. Most of White’s innovative work was published in the 1960s and 1970s. White also published a substantial number of opinion pieces on a variety of topics. One of these topics is brain death.

In an 1972 editorial for the publication Hospital Progress, “The Scientific Limitation of Brain Death,” White notes that:

…we have to acknowledge the probability that eventually all of the major cellular complexes of the human body will be replaceable either by transplanted organs (man or animal) or by sophisticated engineering modules.

As a consequence, the clinical definition of death is shifting from cardiopulmonary criteria to the central nervous system. But unlike other organs,

…this system is not replicatable, representing as it does the repository of the highest functions of man…when this elite cellular system fails it would seem reasonable to assume that what is characteristically ‘human’ is also being lost from the body.

But just as the cardio-respiratory definition of death has evolved and changed with the clinical practice of cardiopulmonary resuscitation,  a similar fate may be in store for the definition of brain death. The clinical use of general anesthesia and hypothermic circulatory arrest, in which the brain can be put “on pause,” emphasize how important the aspect of “irreversibility” is.

As presently defined, the definition of brain death puts much emphasis on brain function upon physical examination. A major limitation of this definition is that it categorically ignores the prospect that brain function could be restored in the future by technologies more advanced than practiced today, provided the material basis of brain function is preserved.

Another challenge is that the science of cryobiology has advanced to such a state where brain slices can be preserved at subzero temperatures and recovered without loss of viability through vitrification. When recovery of organized electrical activity can be demonstrated in vitrified mammalian whole brains, the prevailing definition of brain death will need to be challenged again because it will open the practical possibility to maintain critically ill people in a state of low temperature circulatory arrest without producing one of the indicators of irreversible brain death. Such advances would be an extension of the  experiments Robert White did on isolated hypothermic brains.

As White stresses in the final paragraph of his paper:

…like all biological activity, life and death merge into one another representing a continuum and the neuro-scientist can only in the final analysis determine the point of irreversibility of this highly complex system at which the possibility of organized activity that characterizes human behaviour  has been exceeded.

Promoting cerebral blood flow in cryonics patients

It has been shown that perfusability of the brain is significantly compromised after long-term (>5 min) ischemic events (the “no reflow” phenomenon). Improving cerebral blood flow after circulatory arrest is one of the fundamental objectives of human cryopreservation stabilization protocol.  To that end, cryonics organizations administer the resuscitation fluid Dextran-40 and the drug Streptokinase to dilute the blood (and inhibit  red cell aggregation / cold aggulination) and  break up blood clots, thereby improving macro and microvascular circulation. Research by Fischer and Ames, who investigated the effects of perfusion pressure, hemodilution, and anticoagulation (i.e., the use of heparin) on post-ischemic brain perfusion, indicated that hemodilution is the most effective component of the post-ischemic perfusion protocol for enhancing brain perfusability. However, a later study by Lin, et al. (1978) reported significant improvement of cerebral function and blood flow with combined dextran and Streptokinase administration after cardiac arrest in dogs.

In their study, the researchers measured regional cerebral blood flow and cardiac output as well as EEG (i.e., brain wave activity) during five hours of post-resuscitation physiological maintenance following 12-16 minutes of cardiac arrest. Animals were divided into three groups as follows:

Group I:   no treatment

Group II: 1 g/kg dextran 40 in 10% saline following arrest and 10 mg/kg/minute during the five hour maintenance period

Group III: combined therapy of dextran-40 and Streptokinase — same dose of dextran as Group II and 5,000 u/kg rapid infusion and 25 u/kg/minute during the five hour maintenance period

The duration of flat EEG was significantly shorter in Group III animals (20 to 45 minutes with a mean of 28.8 +/- 2.8) than in Groups I (20 to 120 minutes with a mean of 59.5 =/- 10.8) or II (20 to 62 minutes with a mean of 46.9 +/- 4.8) and showed a faster recovery pattern than in Group I (significant difference was reached at three hours). Group II also showed a faster EEG recovery than Group I, reaching significance at five hours.

Cerebral blood flow, particularly in the hippocampus and grey matter (the areas most detrimentally affected by ischemia) in Group III was significantly improved as compared to Group I as early as three hours post-arrest, and was greater than that in Group II (significantly better only in the hippocampus). There was no difference in cardiac output found between the treated and untreated groups. All groups suffered a decrease in cardiac output of nearly 50% of baseline level (measured at 3 and 5 hours post-arrest).

Hematocrit — the proportion of blood volume occupied by red blood cells — was measured in each group and was found to be significantly increased during the post-arrest period in Group I, decreased to 25% of the baseline measurement in Group III (at both 3 and 5 hours post-arrest), and unchanged in Group II.

The authors speculate that “the improvement in cerebral circulation at the microvascular level after infusion of low molecular weight dextran was thought to be 1) related to the rapid increase in plasma volume with resultant lowering of hematocrit and reduction in blood viscosity, 2) a direct effect on the RBC [red blood cell] which increases its negativity and reduces the tendency to cellular aggregation.” They also note that though some doubt had been cast by the Fischer and Ames paper on the hypothesis of vascular endothelial cell swelling as a cause of no reflow, they did observe a higher proportion of smaller diameter capillaries in ischemic brains as compared to controls, and that “if capillary narrowing does play a role in microvascular deterioration, then hemodilution and prevention of cellular aggregates such as occurs with dextran would be beneficial in minimizing poor flow in narrow capillaries.”

Taken together, these findings indicate that combined dextran-40 and Streptokinase therapy improve brain perfusion after cardiac arrest — at least for arrest periods of up to 16 minutes.– supporting the choice for these agents in cryonics. One limitation of this study, however, is that the experiments did not include a group which received only Streptokinase. Including a Streptokinase group would have given more  precise data about the individual effects of the two agents in improving post-ischemic cerebral blood flow. Recent clinical trials with clot busting agents in cardiac arrest have failed and some contemporary authors question the phenomenon of post-arrest blood clotting. Perhaps streptokinase is useful in the treatment of circulatory arrest but its efficacy is dependent upon other blood flow improving interventions such as hemodilution. The case for post-ischemic hemodilution (and interventions to reduce RBC aggregation) is strong but the case for antithrombotic therapy in cryonics (and resusctation medicine) remains to be made.